Towards anegative

understandingof

psychotherapy

Arno L. Goudsmit

RIJKSUNIVERSITEIT GRONINGEN

Towards a negative understanding ofpsychotherapy

Proefschrift

ter verkrijging van het doctoraat in dePsychologische, Pedagogische en Sociologische Wetenschappen

aan de Rijksuniversiteit Groningenop gezag van de

Rector Magnificus, dr. D.F.J. Bosscher,in het openbaar te verdedigen op

maandag 26 oktober 1998om 14.45 uur

door

Arno Leo Goudsmit

geboren op 28 maart 1955te Utrecht

promotores:prof. dr. P.J. van Strienprof. dr. P.L.C. van Geert

this electronic document can be found at:http://www.ub.rug.nl/eldoc/dis/ppsw/a.l.goudsmit

the author’s address:Capucijnenstraat 926211 RT MaastrichtNetherlandsemail: goudsmit@mailbox.unimaas.nl

to the memory of my father,

Jacques Goudsmit (1916-1960)

"... man in his ignorance makes himself the rule of the universe (...). So that, as rational metaphysicsteaches that man becomes all things by understanding them (homo intelligendo fit omnia), this imaginativemetaphysics shows that man becomes all things bynot understanding them (homo non intelligendo fitomnia); and perhaps the latter proposition is truer than the former, for when man understands he extendshis mind and takes in the things, but when he does not understand he makes the things out of himself andbecomes them by transforming himself into them" (Giambattista Vico,The New Science, 1744, §405/1986, p. 129-30)

Table of contents

Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

1Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

2Autonomy and language in Maturana’s theory of living systems 19

2.1. Autopoiesis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192.1.1. Structure and organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202.1.2. Perturbations and interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212.1.3. Distinctions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232.1.4. Description versus operation of autopoietic systems. . . . . . . . . . . . . . . . . . . 242.1.5. Structural coupling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

2.2. First order consensuality. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262.2.1. Consensual domains. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262.2.2. Description versus operation of consensuality . . . . . . . . . . . . . . . . . . . . . . . 272.2.3. Quasi-objects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

2.3. Second order consensuality. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302.3.1. The emergence of objects. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312.3.2. Some other examples of second order coordination. . . . . . . . . . . . . . . . . . . 332.3.3. "Objects as warrants for eigenvalues"? . . . . . . . . . . . . . . . . . . . . . . . . . . . . 352.3.4. Objectivity and solipsism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 362.3.5. Description versus operation of second order consensuality. . . . . . . . . . . . . 39

2.4. How will these ideas be used in this book?. . . . . . . . . . . . . . . . . . . . . . . . 39

3Experiences and naive objects. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40

3.1. Experiences as first order distinctions, and Merleau-Ponty’s "flesh" . 40

3.2. Naive objects and the coordination of disagreement. . . . . . . . . . . . . . . . 42

3.3. Naive objects: Vico and poetic language. . . . . . . . . . . . . . . . . . . . . . . . . . 46

3.4. The absence of the naive object’s form. . . . . . . . . . . . . . . . . . . . . . . . . . . 50

4Critical objects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54

4.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54

4.2. Gestalt formation and incorporation: the naive perception of form . . . 55

4.3. Epistemology verified: two ways of allocating ’substance’ and’accident’ to ’form’ and ’content’ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 604.3.1. The failure to attain critical objects: the manifestation of power . . . . . . . . . . 634.3.2. The realist discourse. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 644.3.3. The nominalist discourse . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 654.3.4. Expression and impression, explanation and interpretation. . . . . . . . . . . . . . 68

4.4. Recapitulation: naive and critical discourses. . . . . . . . . . . . . . . . . . . . . . . 70

5Internal objects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73

9

5.1. Internal objects and the two critical discourses . . . . . . . . . . . . . . . . . . . . 73

5.2. Internal objects in the realist discourse . . . . . . . . . . . . . . . . . . . . . . . . . . . 755.2.1. The construction of ’self’ in the realist discourse: Descartes. . . . . . . . . . . . . 75

5.3. Internal objects in the nominalist discourse . . . . . . . . . . . . . . . . . . . . . . . 795.3.1. The construction of ’self’ in the nominalist discourse: James and Sartre. . . . 81

6Classical and non-classical science. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86

6.1. Classical science: the integration of the two discourses. . . . . . . . . . . . . . 876.1.1. Content and form in classical science . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 876.1.2. The modeling relation in classical science . . . . . . . . . . . . . . . . . . . . . . . . . . 916.1.3. Galileo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94

6.2. Non-classical science and self-reference in the researcher’soperations: the realist-nominalist controversy revisited . . . . . . . . . . . . . 966.2.1. Gödel’s formal undecidability: the proof process mixes up with the

contents of the proof . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 976.2.2. Bohr’s understanding of complementarity in quantum mechanics: the

measurement process mixes up with the contents of the measurement. . . . . . 996.2.3. Löfgren’s ’linguistic complementarity’ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104

6.2.3.1. Remarks on terminology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1066.3. Psychotherapy and self-reference. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108

7Self-reference, the one-sided boundary and thenegative way . . . . 111

7.1. ’Self-accidence’ and ’self-subsistence’. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1117.1.1. ’Self-accidence’: the self-appearing essence. . . . . . . . . . . . . . . . . . . . . . . . . 1157.1.2. ’Self-subsistence’: the self-describing meaning. . . . . . . . . . . . . . . . . . . . . . . 117

7.2. The interface as the limit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1197.2.1. The ’doctor’s fallacy’ and the concept of ’health’ . . . . . . . . . . . . . . . . . . . . . 122

7.3. A one-sided boundary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1237.3.1. Ambiguities in the definition of the autopoietic system’s boundary . . . . . . . . 1237.3.2. The idea of a one-sided boundary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1237.3.3. The interface and the one-sided boundary. . . . . . . . . . . . . . . . . . . . . . . . . . 125

7.4. A via negativafor understanding psychotherapy . . . . . . . . . . . . . . . . . . . 1267.4.1. Definition of the negative way . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1267.4.2. The use of the negative way . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127

8Psychiatric diagnosis between realist and nominalist discourses. 129

8.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129

8.2. The realist discourse in psychiatric diagnosis: ’neo-Kraepelinianism’. 131

8.3. The nominalist discourse in psychiatric diagnosis: ’DSM and itsvicissitudes’ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133

8.4. Four implicit contaminations of the nominalist diagnostic project . . . . 1388.4.1. Contamination: diseases as gestalts. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1388.4.2. Contamination: conceptual elegance as facticity. . . . . . . . . . . . . . . . . . . . . . 1398.4.3. Contamination: statistical clusters as disease entities. . . . . . . . . . . . . . . . . . 1398.4.4. Contamination: meanings as essences, symbols as appearances. . . . . . . . . . . 139

8.5. The desire for etiology: theory and practice . . . . . . . . . . . . . . . . . . . . . . . 1418.5.1. The theoretical desire for etiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141

10

8.5.2. The practical desire for etiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1438.5.3. Using the patient’s background for explanation versus personal

understanding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1438.5.4. A misplaced preference for realist approaches. . . . . . . . . . . . . . . . . . . . . . . 1458.5.5. A process oriented argument against the DSM?. . . . . . . . . . . . . . . . . . . . . . 146

8.6. A scientific position of non-detachment. . . . . . . . . . . . . . . . . . . . . . . . . . . 150

8.7. In conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150

9Against behavioral technology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152

9.1. The context of discovery and the context of justification. . . . . . . . . . . . 1529.1.1. Psychotherapy as applied science. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1549.1.2. Psychotherapy as exploration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1549.1.3. An asymmetry between explorative and procedural modes. . . . . . . . . . . . . . 156

9.2. Exploration as a self-referential activity . . . . . . . . . . . . . . . . . . . . . . . . . . 1579.2.1. Uniqueness. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1579.2.2. Self-reference and the archimedean point . . . . . . . . . . . . . . . . . . . . . . . . . . 1589.2.3. Improvisation and rules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161

9.3. The problem of psychotherapeutic procedures. . . . . . . . . . . . . . . . . . . . . 163

9.4. Gadamer: phronesis and the aesthetic non-distinction. . . . . . . . . . . . . . 166

9.5. The ’hermeneutic identity’ of psychotherapy . . . . . . . . . . . . . . . . . . . . . . 171

10Observing therapeutic events through parallel processes. . . . . . . . . 173

10.1. Utilizing the parallel process for the passage of one-sidedboundaries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174

10.1.1. The passage from inside to outside. . . . . . . . . . . . . . . . . . . . . . . . . . . 17510.1.1.1. Case. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17510.1.1.2. Elaboration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176

10.1.2. The passage from outside to inside. . . . . . . . . . . . . . . . . . . . . . . . . . . 17810.1.2.1. Case. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17810.1.2.2. Elaboration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185

10.2. The parallel process and the unclarity of therapeuticperformance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186

10.2.1. The central event of the therapy session. . . . . . . . . . . . . . . . . . . . . . . 18710.2.2. The presumed perceptibility of the therapist’s own

countertransference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18810.2.3. Confusion in the interviewer during the interview . . . . . . . . . . . . . . . 19010.2.4. The interview’s confusion unraveled: the second interview . . . . . . . . . 191

11Some notes on projective identification. . . . . . . . . . . . . . . . . . . . . . . . . . . . 194

11.1. Not thinking as an empirical phenomenon. . . . . . . . . . . . . . . . . . . . 194

11.2. The idea of ’projective identification’ . . . . . . . . . . . . . . . . . . . . . . . . 19611.2.1. Klein’s original concept . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19611.2.2. The spectator’s engagement: a benign instance of projective

identification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19711.2.3. Interpersonal conceptualizations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198

11.3. The significance of the concept of ’unconscious phantasy’. . . . . . 19911.3.1. ’Unconscious phantasy’ as a theoretical construct. . . . . . . . . . . . . . . . 19911.3.2. ’Unconscious phantasy’ as a quasi-object: the ’as-if-proviso’. . . . . . . . 200

11

11.4. Some implications of Bion’s ’container-contained’ model. . . . . . . 201

11.5. Projective identification as incorporation . . . . . . . . . . . . . . . . . . . . . 204

11.6. Projective identification and the one-sided boundary. . . . . . . . . . . 206

11.7. Beyond the modeling relation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208

12Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210

References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211

Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224

Samenvatting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226

PrefaceIn 1980 I had written a master’s thesis in clinical psychology at Utrecht University on what was believedto be the cognitive schemas of psychotherapists. It had been a dissatisfying enterprise for me, as I had toassume the existence of such cognitive schemas somewhere inside the therapists, whereas their empiricalobservation was troublesome, a matter of interpreting words and expressions, to the extent that it becamefully unclearwhosecognitive schemas were studied: those of the therapist or those of the researcher whowas so eager to discover them. I did not manage to come with an adequate account.

In the mean time I had found a job in a truly kafkaesque House of Horrors, the Central Bureau ofStatistics of the Netherlands. This was a rich source of frustration, and at times also a source ofexperience with self-sustaining systems that, unlike the persons that inhabit them, are devoid of anycommitment to empirical inputs or intellectual outputs. Here, in 1983, my interests for self-organizingsystems took shape, mainly out of an acute awareness that I should find my way out of the moloch thatemployed me, by developing some expertise in a subject matter far beyond its reach. Mr Jan de Beij,before graduating in Dutch Law, worked as a library assistant at the University of Utrecht. During theyears 1983 and 1984 he sent me the three-monthly microfiche updates of the Utrecht University librarycatalogue. In a pre-internet era this turned out to be a major life-jacket for me, as it enabled me tomaintain access to the literature of my interest. I am still most grateful to him for his generous postings,that have been of so much help to escape from the conditions of my work site.

Out of a fascination that remained unclear to me for a long time, I decided to study Maturana’s theory ofautopoietic systems. In 1983 I organized a study group that focused on Maturana’s writings, and on thewritings of related authors in the field of second order cybernetics. This group, the ’study groupAutopoiesis and Cognition’, was to become for nearly 10 years a two-monthly forum ofSpielerei, debatesand exchanges between participants from several disciplines, where guest speakers were often received toshare a variety of perspectives and ideas. It is to this group as a whole that I am very indebted, and alsoto some of its active members, such as Max van Trommel, Leo van Dijk, Rob Zuijderhoudt, Marcel vander Horst, Maarten van Rhijn (who most generously hosted the group’s meetings for a long time) andJoachim Mowitz.

In 1985 I came to work at the University of Amsterdam, where I made acquaintance to a group of personsthat became of even more importance for my project. This group was at that time involved in finding aformal model of self-organizing processes within bureaucracies. Joachim Mowitz hosted it every fridaynight at his office, as a no-absenteeism full-commitment gang. This group is still active on a two-weeklyschedule. We soon found a shared interest in paradoxes, self-organizing systems, dynamic geometry andother things. This group developed as an intellectual pressure cooker, where wild speculation mixed upfreely with austere argumentation. I am grateful first of all to Joachim Mowitz, with whom I co-authoredseveral papers on autonomous processes, as well as to Jan Hendriks, Jan-Gerrit Schuurman and Paul deGreef, for their long term partnership in the exploration of autonomous systems, and for their unrelentingsupport and their willingness to stumble on every possible block that could be found in the things I wrote.I am also indebted to Ulrike Bercher, who participated in the group only a for few years. The group hasbeen a mutual sounding board; a forum of puritanist and recalcitrant querulantism, where people werehard to stop before having tried all possible deviations and all trivial side-ways; a forum also of cordialencouragement and intense commitment to get the gist out of one’s immature ideas and to help searchingfor more adequate articulations.

In 1986 I had the fortune of being invited by professor Pieter van Strien of Groningen University to workwith him and arrange my ideas into a doctor’s thesis. I am grateful to Trudy Dehue whose original idea ithad been to arrange the initial encounter. I am most indebted to professor van Strien for his support of mywork, in several ways. He has arranged a variety of means to keep me paid by the University for severalyears, and he has made me profit enormously from his vast erudition and his prudence, during all thestages of my writing activities. I am grateful to professor van Strien also for his patience with me, his

13

regard for the various delays I experienced in writing this thesis, and for his tolerance with respect of mychoice of subject matters. For though I still share with him some basic convictions (see also footnote 447on page 152) concerning scientific performance in psychological practice, he also indulged my departurefrom many of his original insights.

I am also most grateful to professor Paul van Geert, also of Groningen University, for co-supervising thepresent thesis. Though his time investment with my work was smaller, the depth of his critical (andalways gentle) remarks was inversely related to the amount of time he had spent on them, or so it seemed.Some of his brief outbursts of clarity even kept me searching for years, others were immediately helpfulin clarifying my own unclarities. I have always appreciated his availability as well as his encouragingsupport, and his many elaborate comments on the manuscript have been of enormous value to itsimprovement.

In 1987 and 1988 professors van Geert and van Strien supported and submitted a request for a researchgrant from the Dutch research funding organization NWO. ’My’ proposals on self-organization inpsychotherapy were nominated twice as the best or the near-best in their category, which under normalcircumstances would have meant a sure position on the rank order of accepted versus rejected researchproposals. For some reasons, however, in the years of my grant requests the entire category was notreceiving any money, due to considerations made at a higher level of the NWO. So in a certain way I cannow express my gratitude even to this organization. For had one of my research proposals been accepted,then this thesis would not have got the time to develop its crisis that is so typical of the negativeunderstanding I am presenting here. It is for this reason that Buber’s (1957) description of the doctor’scrisis, in the fragment quoted below on page 160, has become of a special significance to me. I amconvinced that the freedom I have experienced to phrase and rephrase the topic of my study, the freedomto depart from my original intentions and to return to them in a way unforeseen, and most particularly thefreedom to investigate my own errors and blunders as a function of the objects studied; this freedomwould not have been possible if one is bound to a research path that had to be specified in advance, thatis, at the time the grant request is submitted.

In 1991 I left Groningen for Maastricht. I am grateful to my colleagues at the School of General Practiceof the University of Maastricht, where I have worked since 1994, especially to its current and previousdirectors Yvonne van Leeuwen and Paul Höppener, for their encouraging attitude with respect to thiswork.

There are other persons to whom I am grateful, for commenting on parts of the text, or for literaturesuggestions. Others have spent much energy on interviews or on discussing various issues with me andsaved me from many blind alleys. For these reasons I am grateful to Hilde Bakker, Pietro Barbetta, Lexvan Bemmel, Tom Berk, Ad Boerwinkel, Tom Brinckmann, Mauro Ceruti, Yara Dekking, Arnout Esser,Hans Ettema, Heinz von Foerster, Maarten van Gils, Jaap van Ginneken, Jos van Groeningen and othermembers of the Riagg Groningen psychotherapy study group, Eef Hogervorst, Richard Jung, VincentKenny, Wim van Kessel, Floor Kolk, G.A.S. Koster van Groos, Regina Kursch, Annie Mattheeuws,Mathieu Matthijssen, Timon Meynen, Henk Milius, the late Gordon Pask, Frenk Peeters, Hans Pols,Herman van Praag, Han Sarolea, John Shotter, Frank Smulders, Jaap Terhorst, Coby Tromp, WiesVerheul, Rienk Vermij, Paul Voestermans, and Gertrudis van de Vijver. Furthermore, I am grateful to anumber of (other) psychotherapists who remain anonymous, with whom I conducted interviews and/orwho permitted me to observe their therapeutic performances.

I would like to thank my wife, Rosemarie Samaritter, for the love and respect she has been capable ofmaintaining for me during the years of my involvement with this book, and for enduring me in its variousstages. In the mean time, her clear ideas in the field of dance therapy did not cease to be a major sourceof inspiration and wonder.

Our three children Marthe, Ben and Eva have taught and shown me some of the essentials of life, andthey have provided a wealth (and at times an overflow) of experiences in a domain that is mostinaccessible to critical discourse. Though the care of our kids did take some time, the crisis from whichthe writing of this book suffered between 1995 and 1997 was in fact due to my incapacity to relate the

14

things I had written thus far on self-reference and the construction of objects (roughly the chapters 2-6, 8and 9), to the phenomenon of projective identification, which I felt had always played a major role in mythinking, but a role I found myself unable to specify.

My attendance in 1997 to a conference in Turin on the work of Wilfred Bion, the great psychoanalyst, hasbeen most helpful and even releasing in this respect, especially Grotstein’s (1998b) contribution to theconference. Furthermore, the internet discussion list "Bion97", moderated by Silvio Merciai, which hadbeen set up as a preparatory forum for the conference, and which is still actively discussing a variety ofhighly interesting topics related to Bion’s work, has been for me a rich source of information and betterunderstanding.

Only after I understood more of Bion’s theory of projective identification did I feel capable of containingmy own fears to write chapter 11, undoubtedly the most postponed chapter of this book. In the mean time,chapter 7 had come into existence as the theoretical blueprint of the drama.

Projective identification was a concept and a phenomenon of which no unified understanding andinterpretation was available. The intense controversies on it pertained to the ontological status of theconcept. What I came to believe was that the ontological problems (of the variety: "how can the patient’sunconscious phantasy get into the therapist’s mind?"; and: "how is the parallel process possible?") couldbe reframed with the help of Merleau-Ponty’s concept of ’flesh’, as a convergence of object and subject,beyondthe distinction of being from knowing, and, hence, beyond the scope of critical discourse. This, infact, was the connection I had been looking for, and that in vain I had tried to catch in more positiveterms.

1Introduction

"Fail again, fail better"(S. Beckett,Worstward ho, 1983)

The field of psychotherapy research is a festival of clarity. The interested reader is treated on a wealth ofstudies on the properties of particular disorders, as well as comparative studies on their specific treatmentprocedures. How effective is a particular psychotherapy? How can therapy outcomes be measured? Howcan they be improved? Which variables are relevant? Research questions of these types have led to anincreasing search for concrete and clearly definable properties of the therapeutic procedures and theireffects upon patients’ behaviors. The leading question is always: "what treatment, by whom, is mosteffective for this individual with that specific problem, and under which set of circumstances?" (Paul,1967, p. 111). Furthermore, therapists, therapeutic institutions and therapeutic schools that wished to showtheir effectiveness found a natural ally in those researchers who were capable and eager to measure clearoutcomes. If anywhere in the history of the empirical sciences there has been a spirit of optimism andprogress, it must be here.

What about the unclear outcomes of vague therapies? What about the failing measures, the invalidoperationalizations, the almost metaphysical assumptions concerning unconscious, preconscious, intuitive,experiential, existential, and perhaps even religious aspects of psychotherapeutic encounters?Psychotherapists have written extensively in these terms, and have accounted for their therapeuticactivities by taking recourse to theoretical concepts that didnot stem from any empirical researchtradition. Given the metaphysical constitution of these concepts, and given the difficulties of translatingthem into empirically measurable variables, does it mean that these therapies are of a minor quality? Tosay the least, their vagueness did not fit in well with a research tradition of clearly defined andmeasurable variables.

The patient’s failing recognition that his therapist is different from his father, the patient’s incapacity toexpress some of his own experiences, the therapist’s intuitions about those inexpressed experiences, thetherapist’s incapacity to come to grips with a particular problem, the patient’s experience of his ownautonomy, or his incapacity to have this experience: these are all examples of frequently occurring eventsin psychotherapeutic sessions, and the only way in which modern empirical research is capable to handlethem is totranslatethem into observation terms, i.e. behavioral measures! Accordingly, both therapist andpatient behaviors can be measured and correlated, but the quality of the things they are talking about anddealing with does not re-appear in these behavioral measures.

Only a few have dared to speak, be it in optimistic terms, about a crisis in the field of psychotherapy (e.g.Erwin, 1997). Could it be that the difficulties to make appropriate and valid translations are more than amatter of research technique? Could there be some property of the therapy itself that precludes or hindersa proper translation of the relevant events into observation terms? The present study is meant to contributeto such an answer. If indeed there is a feature that remains hidden, in spite of, and perhaps even due to,our very attempts to unveil it, then it may be time to rethink psychotherapeutic situations in terms of theirlogic, not their mechanics. If this idea makes any sense, then what should such a feature look like andhow should we study it empirically?

It is here that my search for an understanding of autonomous systems, entities in which a certain self-reference seemed to be built in, has started. In particular, I have investigated how self-reference can beunderstood both as a property of life and lived experience, and how it hides itself from being observed inconcrete operational terms. This provides a ground to reappreciate the vagueness of many psychotherapies.

15

16

For if this vagueness is due to a self-reference inside the therapy, and if we can show where this self-reference resides in the therapy and how it can affect our attempts to observe it, then the disappointingoperationalizations, far from being pointless, can be acknowledged as manifestations of the things we areinterested in. That is to say: if we fail, let’s see how the failures came to be! The way in which theresearch processsuffersfrom the self-reference encountered can be reappreciated now as itself a researchtool. As a consequence, the type of psychotherapy research that I have ventured in this study had someresemblance to what is usually done in supervision sessions.

The course of my argument will be as follows. I will introduce some ideas on autopoietic systems and thecapacity of such systems to make distinctions, and eventually to perceive objects. I will elaborate theseideas in my own way, presenting various modes in which people are capable to coordinate their actionsand moderate their conflicts by means of language and the ensuing construction of objects. My main pointwill be that, in doing so, self-reference can only be understood as something beyond the domain of theseconstructs. This holds also for the moments of self-referential exploration that occur in psychotherapeuticencounters: these cannot be described in terms of step by step procedures performed by the therapist, norin terms of concrete behaviors on the part of the patient. It holds also for the moments of projectiveidentification between therapist and patient, when the therapist’s capacity to coordinate his ownexperiences is severely inhibited. Nevertheless, a therapist can indicate and demarcate these moments onlythrough self-referential movements that necessarily become less clear as they approach the limits of hisdiscourse. It is at these limits that the phenomenon of ’not understanding’ becomes a token for that whichcannot be caught in clear terms. We are making here small steps in the tradition of thevia negativa.

Some of the chapters of this book have been reworked, others are new. Chapter 2 is an adaptation frommy 1991 paper. The chapters 3 to 5, as well as parts of chapter 7, are an extensive elaboration of my1992a paper. Chapter 9 is an elaboration of my 1992b paper. Chapter 10 contains parts from my papers1993 and Goudsmit & Mowitz (1987).

In chapter 2 I will introduce some concepts from Maturana’s theory of autopoietic systems. Of interest ishere the notion of autopoiesis, the ’organizational closure’ of living beings. This is a concept at the edgeof metaphysics, in that it is defined in a way that disables immediate empirical confirmation. One cannotsee the closure closed; one has to force an opening in order to see anything. There is an interestingparallel with Merleau-Ponty’s ideas on silence: "how every philosophy is language and nonethelessconsists in rediscovering silence" (1968a, p. 213)«1».

I will pay attention to the idea that autopoietic systems are capable of making distinctions as a necessaryeffect of their organizations. Thus, according to Maturana, making distinctions is essential to the processof staying alive. The concept of distinction will be a major issue throughout this book.

Maturana’s ideas on how it is possible that objects are perceived by language users is taken here as apoint of departure for the elaboration in chapter 3 of some concepts beyond the theory of Maturana. Inparticular, I will relate Maturana’s concept of distinction to what Merleau-Ponty wrote on the ’flesh’,which is a kind of ontological substratum for both subject and object; it is that from which both objectand subject develop. Furthermore, I will relate Maturana’s ideas on the perception of objects to the ideasof Giambattista Vico on the poetic language of the earliest peoples. The objects thus perceived arise as acommon construct, a social artifact for the coordination of diverging experiences between persons. Theyserve as an elementary mode of avoiding conflict. I will call these objects ’naive’: they do not distinguishbetween that which is perceived and how it is perceived.

This latter distinction is put into the focus in chapter 4. I will call the usage of language that has thisdistinction ’critical’. I specify two critical discourses: both are used to temper conflict and find agreementbetween persons by constructing some stable ground, to which diverging qualities can be attributed. In thefirst discourse, which I will call realist, the stable ground is some abstract essence, to which concrete, butdiverging, appearances are attributed; in the second discourse, which I will call nominalist, the stable

1. "comment toute [philosophie] est langage et consiste cependant à retrouver le silence" (1964a, p. 267)

17

ground is a concrete symbol, to which abstract, but diverging meanings are assigned.

Both discourses can be used, in radically different ways, to speak about mental entities (called ’internalobjects’), both in other persons and in oneself. This will be the subject matter of chapter 5. Theseradically different ways will be elaborated further in chapter 7.

But both discourses, when applied to the world outside oneself, could become integrated into what I willcall ’classical science’ (chapter 6). However, this was possible only as long as the issue of self-referencedid not receive the attention of scientists. I will also argue in that chapter that, with self-reference havingbecome an issue of interest (first in mathematics and physics), the integration between the two criticaldiscourses did no longer hold.

In chapter 7, then, I will elaborate how, with respect to self-reference, the two critical discourses takeentirely different positions, from which self-reference also looks entirely different. Nevertheless, eachdiscourse shows a serious lacuna; neither can fully express the concept of self-reference, and I will arguethat both lacunae in fact converge upon one single ’clearing in the woods’«2». In that chapter I willintroduce my concept of ’one-sided boundary’, which is to be understood as a boundary to the domain ofits own definition: beyond that domain the boundary does not exist! This is precisely the matter with the’vagueness’ of psychotherapy: it brings the external observer (and also the internal one, the therapist) intoa position of not (or no longer) understanding. Hence, I will maintain that we may appreciate our failuresto understand as properties of that which we are trying to grasp. This is thenegative way, and I proposeto apply it here in the tradition of Cusanus and of Merleau-Ponty, whose concept of ’flesh’ returns here interms of a ’true negative’: existing as an absence to what we can perceive, not as some transcendentalentity that is positively present elsewhere. At this place in the book we are at the vertex of my argument.The chapters that follow are meant to utilize the ideas built up thus far.

First, chapter 8 is to illustrate how in the field of psychiatric diagnosis the opposition between realist andnominalist discourses has led to different conceptions of mental disorder, one more interested inexplanations and past developments, and one more interested in future developments and treatment policy.It is maintained how incompatible these conceptions are if the observer’s (diagnostician’s) own personalcommitment is not involved. It is through making naive objects that the diagnostician is capable tointegrate the two critical discourses. Nevertheless, naive objects are not always sufficiently powerful. Inparticular, they fail when the practitioner’s very capacity to think is affected. When this happens, we arein the realm of projective identification, which will be the stuff of chapter 11. The chapters that follow areto prepare that.

Next, in chapter 9, I will present the construction of naive objects in terms of self-referential explorationprocesses. It will be maintained that the task of exploration in psychotherapeutic sessions cannot beperformed as a technological procedure. This is argued by means of Gadamer’s concept of ’phronesis’(prudent action), which is to be seen as a kind of intermediary between general rules, or generalknowledge, and unique individual situations; and by means of Gadamer’s ideas on ’playing’. According toGadamer, in the performance of a piece of art (e.g. theatre, music) there is a particular collapse of (what Icalled) critical discourse; the observer becomes naive, and can no longer distinguish the piece of art fromthe way in which it is performed, nor from the way the composer had meant it to be. Gadamer calls thisthe aesthetic non-distinction. Transposing this idea to the field of psychotherapy, we find that theprocesses of exploration are qualified by a similar kind of aesthetic non-distinction. Exploration goes bymeans of naive objects, and is self-referential. This cannot be replaced by a procedure, applied by a’behavior technologist’ or ’behavior engineer’. It would eliminate the very aspect of exploration.

In chapter 10, then, I proceed by giving some case material from my own experiences with psychotherapyresearch that attempts to do justice to the self-referential and closed qualities of therapeutic systems. Thischapter is meant to clarify and illustrate some of the phenomena of getting trapped in a closed system, aswell how the observer’s or researcher’sabsence of understandingand its exploration can be made useful.

2. I owe this beautiful metaphor to professor Pieter van Strien

18

In chapter 11, finally, I discuss the concept of ’projective identification’. Unlike the self-referencediscussed in chapter 9, here we encounter situations in which a decrease of the therapist’s (as well as theexternal observer’s) capacity to clearly think is a property of the thing to be observed! Whereas inprocesses of exploration critical perception was impossible, in favor of naive perception that couldflourish in the therapist and the patient, matters are different here. In processes of projective identificationnaive perception and, with it, exploration are halted and can be resumed only to the extent the therapist(or observer) is capable of enduring (’containing’) some of the negative experiences that the patient isincapable to endure. These negative experiences are of an ’undigested’ quality, they are ’raw experiences’,like radical atoms ungraspable without changing and binding them. It is here that the negative way, theexploration of one’s incapacity to fully understand, shows to be most fertile.

The gist of this book is that the discipline of psychotherapy, currently being the subject matter of animmense amount of empirical investigations, does contain some features that are sensibly real to thoseinvolved, but beyond the reach of those who wish to investigate, observe, judge and improve it. Thereason for this divergence, I think, is, that some things belong to a domain about which one cannot speak;but instead of being entirely silent about it, as Wittgenstein recommended, it may still be possible at leastto experience something of it, and to say at least something afterward; and to discover how much thesewords are besides the mark, or perhaps not.

In short, this book maintains that some things that happen in psychotherapeutic situations can beapproached if the investigator is at least willing to pay attention to the various ways in which hisdescriptions turn out to be inadequate. This is the road towards a negative understanding.

2Autonomy and language in Maturana’s theory ofliving systemsVery little has been written about the linguistic interaction processes that are supposed to take placebetween autopoietic systems. Unlike most texts on autopoiesis«3», I will lay emphasis in this chapteron the notions of language and consensuality in Maturana’s theory of autopoietic systems, because thesenotions are quite useful as a conceptual platform from which to develop some thoughts of my own aboutprocesses that take place in psychotherapies, as well as the constraints put upon us in our attempts toknow them.

2.1. Autopoiesis

The theory of autopoietic systems, developed by Maturana and Varela (1980) intends to explain theautonomy which characterizes living beings. Anautopoietic systemis to be understood as a productionprocess that permanently coincides with its own product. Such a production process is not comparable toany artificial production process developed by modern technology. For in the latter the products alwaysdiffer from the ways in which they are produced. Not so in the former. Here production process andproduct are of the same type. This is called by Maturanaorganizational closure«4». Anorganizationally closed system is not directed towards a particular goal, even not towards survival, thoughit appears as such. Below I will pay some more attention to the observer’s temptation to attribute goals,functions, meanings and information (processing) to the performance of an autopoietic system. Anautopoietic system does not operate purposively, but simply continues operating until it halts. All eventsthat it undergoes (i.e, all events that affect the ways the system operates) are either incorporated into theongoing self-producing process, or they are not. In the latter case the production process stops and theorganism dies.

Autopoietic systems exist as networks of processes, that continuously coincide with the processes bywhich they are generated. The processes that take place in the autopoietic system are the processes bywhich the system is kept alive, i.e. the processes by which it continues to produce itself.

The autopoietic system is defined such as to produce itself as a self-producing and self-preserving networkof relations between processes and components. This definition does justice to a fundamental property ofliving systems, namely that they are capable of conserving their capacity to survive,by surviving. Livingsystems succeed in this by their capacity to adapt to environmental changes. Adaptation is boththe way inwhich they survive, andthat whichthey preserve in surviving«5». Adaptation, therefore, is not a mere’feature’ of autopoietic systems, but a kind of ’essence’ that is situated at the system’s interactionsurfaces«6».

3. Even Mingers (1995), in his extensive treatment of autopoietic systems, only spends a few pages on Maturana’s ideas onlanguage. The best exception can be found in Verden-Zöller’s (1989/1993) writings on mother-child play (see also Verden-Zöller &Maturana, 1989).

4. Varela (e.g. 1983, p. 163) prefers the term ’operational closure’.

5. cf. Varela, 1984, p. 161

6. I will return to this important peculiarity when discussing the fundamentalinsufficiencyof the essence-appearance dichotomy todescribe autopoiesis (cf. section 7.1.1).

19

20

2.1.1. Structure and organization

Maturana uses the terms ’structure’ and ’organization’ for two things that are on first sight nearlyidentical. A system’sstructure is the way it is materially realized. Itsorganizationis the way in whichthrough this realization the system’sidentity is preserved«7». For example, the structure of a car (i.e.,a particular specimen from the class of all cars) consists of all specific components of this car plus theirspecific interrelations. When the car’s color is changed, so is its structure. On the other hand, the car’sorganization does not change with its color (this does not hold entirely for fire engines). Rather, the car’sorganization is that which makes this car a specimen of its class. Take for example a particular type ofcar of a particular brand. If one were to change the color of its back lights, e.g. red into green, then thiswould entail a change of the car’s identity. For such a car would not function properly in modern traffic.The way in which the car’s identity (i.c. its brand and type) is defined makes up its organization. The wayin which a particular specimen of this type and this brand is constituted, makes up its structure.

Part of the properties of a system’s structure is not essential for the organization. Interactions that involvejust these structural properties are calledorthogonal interactions(Maturana, 1987a, p. 348). Thus, not allfeatures of the structure of a particular specimen of a car are essential for the type of which this care is aspecimen. The car’s color is normally not included in its definition (for fire engines different rules hold).Thus, the car’s organization can be preserved when its structure undergoes certain changes. Therefore, itis possible that a similar organization is realized in an entirely different structure. However, at somestructural changes the organization changes radically (as in the case of turning the car’s back lights intogreen, or of changing its wheels into square blocks)«8».

It is also possible that one and the same compound of constituent parts (same structure) possesses morethan one organization. For example, an old car’s market value may be assessed qua means of transport,but also qua scrap. On a hot summer’s day its roof may even be used to fry an egg. Then the organization’frying pan’ and the organization ’car’ are both realized in one compound of car/pan constituents. Clearlysuch a specimen of a frying pan has a lot of inessential structure parts.

Living beings are qualified by a very particular organization, one that we denoted above as autopoiesis.The autopoietic organization is ’closed’. The system as a product coincides with the process by which it isgenerated. Put in circular termswe might characterize«9» such an organizational closure as thecoinciding of the organization with the way in which this organization is materially realized in thestructure. Indeed: a recursive process«10». The organizational closure means that the system’sorganization pertains to the way in which the organization itself is preserved during the structural changesthe autopoietic system continuously undergoes. In still other words: the system’s structure has the doublerole of process and product; its organization is the way in which this double role preserves itself.

One might wish to understand the relation between organization and structure as one between an abstractessence (organization) and a concrete appearance (structure), but this does not hold for organizationallyclosed systems, for in the latterit is the very relation between organization and structure that makes partof the organization. I will discuss this in section 2.1.2 in terms of a system’s domain of possible

7. The usage of the terms ’organization’ and ’realization’, for many people rather peculiar, seems to be derived from the work ofRashevsky (1954) (cf. Rosen, 1985, p. 172).

8. A well known case of doubt was a design by the British inventor Sir Clive Sinclair: His design for an electric ’car’ lacked somefeatures that the competent authorities considered as crucial for a car. For example the steering wheel was situated behind the driver’sback, not in the normal frontal position. Was such a thing a car? (I owe this example to Vincent Kenny.)

9. Though Maturana does not formulate it thus, it seems an adequate way of putting it.

10. One might be inclined to describe the organization metaphorically as the organism’s blueprint, its ’computer program’, but thiswould immediately reveal one of this metaphor’s weaknesses: it would require a computer program that coincides with the rules of itsown implementation! Cf. also: "... to caution the reader against the force of many years of dominance of computationalism, and theconsequent tendency to identify the cognitive self with some computer program or high level computational description. This will notdo. The cognitive selfis its own implementation: its history and its action are of one piece." (Varela, 1991, p. 188; italics added)

21

interactions.

2.1.2. Perturbations and interactions

Each interaction the system undergoes, yields a change in its structure, which is called ’perturbation’. Theterm ’perturbation’ could give the impression that something goes wrong, but this is not what is meant.To the contrary, an autopoietic system does nothing but continuously compensating these perturbations.These processes of compensation of perturbations do not so much consist of adequate counteractions, as inclassical negative feedback systems, but rather are they processes of incorporating these perturbations intothe system’s network of self-producing processes. Thus the perturbation loses its perturbing effect andbecomes part of the system’s modes of functioning.

This adaptive capacity to compensate perturbations is what Maturana callscognition. It is because of thiscapacity that the autopoietic system has its autonomy and is able to keep it. In this way ’cognition’ isdefined in a really wide sense, viz. the very skill of living and surviving. "... conservation of living (...)constitutes adequate action (...) and, hence, knowledge:living systems are cognitive systems, and to live isto know" (Maturana, 1987a, p. 357). Though this definition often leads to misunderstandings, it should betaken for what it is: a matter of definition«11».

Notice that the process of compensation of a perturbing event does not show a dichotomy that we areused to from our dualist philosophical tradition. It is the dichotomy between impression and expression. Inthe process of compensation of perturbations these two coincide«12». The process of compensation isat the same time both the impression that the perturbing event triggers in the system, and the expressionof the system’s individuality.

The impacts of events upon the autopoietic system are called ’perturbations’, and not ’inputs’. For it is theway in which the system deals with perturbing events, the way in which the perturbation is incorporatedin the system’s modes of functioning that is decisive for what the perturbing events mean to the system;there are no ways to determine what these events mean to the system, but to look how the system dealswith the event.

There are no ’an sich’ qualities of a perturbing event that could be specified in termsother than the wayin which the autopoietic system deals with them. And this way in which the system deals withperturbations is determined solely by the system’s structure at that moment, and not by any alleged ’ansich’ property of the perturbing event. This is what Maturana callsstructure determination. A rubber ballat room temperature will jump up. However, when first frozen until close to the absolute zero, it willshatter into splinters, as if it were made of glass. Therefore, the collision that I impose upon the ball isnot determining its behavior, but rather is its state determining the way in which it deals with thecollision. In the present example this state is easily summarized as in terms of temperature. But, one willbe inclined to object, this temperature was also imposed upon the ball beforehand? The answer is: no, theway in which, as well as the fact that the ball undergoes the freezing, are properties of the ball, not of thetreatment it receives. In order to use the ball for demonstration purposes, we should prepare it in a waythat is in agreement with its properties, irrespective of whether we are concerned with collisions, changesof temperature or other interactions. Properties ’an sich’ of a treatment that we give to an object neverdetermine how it behaves. In Maturana’s terms: ’instructive interactions’ are not possible. In this respect

11. Cf. "Cognition as a biological phenomenon takes place in a living system as it operates in its domain of perturbations, and assuch it has no content and is not about anything. Therefore, when we say that we knowsome-thingwe are not connoting what happensin the mechanism of the phenomenon of cognition as a biological phenomenon, we are reflecting in language upon what we do."(Maturana, 1987a, p. 361)See also Varela (1991, p. 181).

12. This is in contrast to the dualistic position about mind-body interconnections (see section 4.3.4). An example of the coincidenceof impression and expression can be found in a field study by Kaye (1979) on mother-infant interactions. He describes (p. 199) how byusing imitation, the mother can mold and change the infant’s expressions (e.g. away from crying).

22

autopoietic systems do not differ from non-autopoietic (’allopoietic’) systems. Structure determination isuniversal, and is not unique for a living system’s autopoiesis. Also a corpse is structure determined. Inregard of structure determination of systems Maturana writes: "The interactions [these systems] undergowill only trigger changes in them; they will not specify what happens to them." (1987b, p. 73).

The qualities the autopoietic system assigns to perturbing eventsexist asthe process of compensatingthese events. Hence the way in which an organism deals with the perturbation becomes a definition of thatperturbation. Maturana uses in this context the metaphor of the (old times) pharmacologist (1983, p. 268).Only by paying attention to the way in which a laboratory animal responded to a particular chemicalmaterial could something be concluded about the properties of that material. Thus these properties weredetermined and judged by the way in which the animal behaved while compensating«13». We areentering here the cognitive world of the sophists who maintained that man is the measure of all things.Maturana in fact turns this idea somewhat more radical: each organism is the measure of its ownmedium«14». Whether or not something has a meaning for an organism depends only of theorganism’s interactions. Only then and only in that way does the organism come to make distinctions (seesection 2.1.3).

A domain of interactionsis the set of all interactions that a system can have at a particular momentwithout disintegrating. Events that do not have an impact upon the system’s modes of functioning do notbelong to the domain of interactions. They do not exist for the system. The notion of disintegration isimportant, because it entails the existence of limits to the influences a system can undergo. Furthermore,each domain of interactions has its restrictions. Interactions beyond the system’s domain of possibleperturbations’«15»will, when perturbing the system, effectuate a change of organization or adisintegration of the system«16».

More specifically, the domain of interactions of an autopoietic system at a particular moment is thetotality of interactions that the system at that moment can have with its medium, such that it canincorporate the perturbations into its autopoietic self-producing network. As long as these compensations(incorporations) can continue the system stays alive. The domain of interactions of a system, whetherautopoietic or not, continually changes as a result of the system’s interactions with itsenvironment«17». Thus, the domain of interactions is a feature of the system’s structure at aparticular moment. This structure changes continually. Since the autopoietic organization is qualified bythe way in which the structure is both process and product, and since the structure continually changes asa result of the system’s interactions with its medium, the autopoietic organization is to be regarded astheway the system deals with its interactions. This way of dealing remains constant during the structuralchanges. It is the achievement of incorporating the perturbations that the system undergoes. Theautopoietic organization cannot be conceived apart of these continuous interactions and the correspondingchanges of structure that it undergoes.

Maturana (e.g. 1987a, p. 359) distinguishes ’environment’ from ’medium’. Theenvironmentis the milieuin which the observed individual is observed to make his distinctions. This environment exists for theexternal observer. On the other hand, the milieu as it exists for the individual himself, i.e. the milieu inwhich the system is realized by interacting with it, is called (Maturana 1987a, pp. 338ff) the

13. This example is not discounted by the fact that, later on, it became possible to analyze biochemical materials without the use ofanimals. For in no way would that mean that the objective properties of the material could be established. For even now the data foundthrough natural science investigations exist as properties of the behavior (the ’praxis of living’) of individuals, viz. the community ofobservers (cf. Maturana, 1987a, p. 327).

14. As yet there is no talk of perceivable objects, which will be introduced in section 2.3.

15. Maturana, 1983, p. 259

16. Maturana (1983, p. 259) speaks of a ’domain of possible destructive interactions’.

17. Therefore, it cannot be determined how at a particular moment a system’s domain of interactions exactly looks like.

23

’medium’«18»The locus of the interaction boundary is called theniche (Maturana, 1987a, p. 341).At this site the continuous adaptation to the system’s medium, takes place, enabling the system to survive.An autopoietic system’s niche is not a thing that affords being perceived or measured, but rather acontinuously changing part of its medium, where the system meets its medium.

The niche is determined by the nature of the interactions. For example, when confining ourself to theinteraction ’breathing’, we can conceive of the niche as the site where lungs and air meet. For other kindsof interactions (e.g.: walking) the locus of the niche differs accordingly. Problems to empirically observethe niche will return in section 2.1.5 and in section 7.1.1.

The autopoietic system, then, exists as long as it succeeds in using the effects of its ongoing interactionsas parts of its own internal dynamics. Precisely at the locus of its interactions with the medium in whichthe system exists, its dynamics is decisive for its adaptation. Its state at a particular momentis the way inwhich it manages to deal with perturbations. Thus, the state of the system at a particular moment is theparticular network of processes at that moment that take place as a result of the interactions that thesystem is having at that moment.

2.1.3. Distinctions

By the acts of compensating the effects of perturbing events, an autopoietic system assigns particularqualities to these events. These assignments are calleddistinctions. One would expect that a distinctionpertains to something that pre-exists its being distinguished, as in "I am distinguishing the figure from itsbackground", or "the Pavlov dog distinguishes the circle from the ellipse". In the work of Maturana andVarela«19», however, the drawing of a distinction is not equivalent to the perception of objects, butinstead is only a necessary condition of it. By drawing a particular distinction, this distinction is actuallybeing drawn at a locus where it did not exist thus far«20».

The drawing of a distinction is an elementary operation, with which the autopoietic system compensatesperturbations. I will use the term ’first order distinction’ or ’primary distinction’, in contrast to ’secondorder/secondary distinctions’. Below (section 2.3) we will see that the coordination of such operationsplays a central role in the bringing forth of perceivable objects. We may conceive of the elementaryoperations as the colorings that a naive individual (e.g. a baby) assigns to its world. There is no talk ofperceiving objects, only of basic assignments of valuations (pleasant/unpleasant, etc.). These assignmentsarise during the innumerable processes of compensation of perturbations, for instance while taking anddigesting food. During these processes the body makes the transition from ’hunger’ to ’no hunger’, andthus draws the corresponding distinction.First order distinctions, therefore, exists in these perturbationcompensating activities and come into existence through them«21».

The idea of first order distinctions«22»might suggest the existence of a kind of substratum«23»

18. We are reminded of von Üxküll’s (1909) term ’Umwelt’ (cf. Maturana, 1980a, p. 51; 1980b, p. 28).

19. primarily 1980

20. Distinctions, thus conceived, have been presented and elaborated by Spencer Brown (1969). Notice that such distinctions evadethe traditional opposition between nominalism and realism (cf. also sections 4.3ff).

21. Cf. Lingis, who explains Merleau-Ponty’s (1964a) concept of ’perception’ in remarkably similar terms:"What we perceive then is not a positive term existing in itself and supporting its own "properties"; what we perceive is a

contrast, a tension - not an adequation with our substance, but a difference from us, marked out in the continuous fabric of being, of"flesh," of which we too are a part. (...) For Merleau-Ponty the terminterrogationwas intended to name perception as an operation ofmaking differences and distances appear." (Lingis, 1973, pp. 86/7).

See also footnote 51 on page 32, where I propose that Merleau-Ponty’s notion of ’voir de voir’ resembles that of ’second orderdistinction’. I will return to the concept of ’flesh’ in section 3.1.

22. that are basically identical to experiences, as I will maintain in section 3.1.

23. Below in section 4.3 and following I will extensively use the notion of ’substance’, which comes close to it.

24

that underlies all real objects (1987a, p. 371). Maturana recognizes the attractiveness of such an idea ofsubstratum, though he underscores that ontologically a substratum cannot have an existence as an object."Nothing exists in the substratum" (Maturana, 1987a, p. 371). The source of the perturbations is thereforenecessarily concealed in vagueness. Distinctions are not imposed upon an already existent substance, but,instead, they lay the ground for its existence. I will return upon this in section 2.3.4.

When an observer sees a system interacting with its environment, then it is appealing to believe that thesystem draws a distinction in its environment, a distinction of ’something’ the observer can denote as aparticular object. Below in section 2.2.2 we will see that such a ’something’ does not exist also as adistinguishable object for the autopoietic system itself, but that it is instead a feature of the observer’sdescriptions. Furthermore, in section 2.2.3 I will present the notion of ’quasi-objects’ in order to be moreprecise on this subject matter.

2.1.4. Description versus operation of autopoietic systems

For an observer of an autopoietic system it may be tempting to recognize sense and meaning in thebehaviors of the system, that in a strict sense do not make part of those behaviors. The observer may thuspostulate that the system’s behaviors are purposive. The system is apparently directed at its own survival.A related postulate of an observer pertains to the functionality of the system’s behaviors. An observermay believe the system not to behave arbitrarily, but in a way that is functional to the attainment ofparticular ends. Or the observer may attribute particular meanings or communicative signals to thesystem’s behaviors, which are seen to ’denote’ something«24».

However, neither the purposiveness nor the functionality, nor the semantics of the observed behaviors arefeatures of these behaviors. Maturana and Varela call them properties of the descriptions made by anobserver of these behaviors. They belong to thedomain of descriptions«25». This domain should notbe confused with the organizationally closed processes of the individual. Descriptions in terms of goal,function or meaning are to be conceived as in themselves most legitimate ways of symbolizing theseprocesses. It would be considered illegitimate, however, to presume that the autopoietic individual actuallydoes operate according to the descriptions made by the observer«26». Its organizational closure takesplace in a domain to which the observer’s concepts (function, goal, meaning, etc.) do not pertain. I willreturn to this in chapter 7.

Other elements from the observer’s domain of descriptions are descriptions of the observed behaviors in

24. A beautiful example of the assignment of meaning to behavior can be found in Bateson’s description of his cat. This cat layasleep at his working desk while he was dictating into his tape recorder the sentences of his book about the complexity of relationsbetween logical levels of nonverbal signals. By way of example he made a remark about the cat’s ears and its bodily posture. Batesoncontinues:

"While I was dictating the last hundred words, the cat changed her position. She was sleeping on her right side, her headpointing more or less away from me, her ears in a position that did not suggest to me alertness, eyes closed, front feet curled up - afamiliar arrangement of the body of a cat. While I spoke and, indeed, was watching the cat for behavior, the head turned toward me, theeyes remained closed, respiration changed a little, the ears moved into a half alert position; and it appeared, rightly or wrongly, that thecat was now still asleep but aware of my existence and aware, perhaps, that she was a part of the dictated material. This increase ofattention happenedbeforethe cat was mentioned, that is, before I began to dictate the present paragraph.Now, with the cat fullymentioned, the head has gone down, the nose is between the front legs, the ears have stopped being alert. She has decided that herinvolvement in the conversation does not matter." (Bateson, 1979, p. 130/1).

25. Maturana & Varela, 1980, p. 75, pp. 85ff; Varela, 1979, pp. 11/12

26. Cf.: "We as observers can project all cellular processes upon a system of three orthogonal coordinates, and legitimately say, asvalid in the projection, that specification is mostly produced by nucleic acids, constitution by proteins, and order (regulation) bymetabolites. The autopoietic space, however, is curved and closed in the sense that it is entirely specified by itself, and such a projectionrepresents our cognitive relation with it, but does not reproduce it." (Maturana & Varela, 1980, p. 92).

Cf. also Strasser (1985, p. 137): "Furthermore my body maintains a peculiar relationship to itself; for instance, I can touch myleft hand with my right hand. Thus a body is demarcated that is ’auf sich selbst bezogen’ [’related to itself’; Strasser refers here toCartesianische Meditationen & Pariser Vorträge, Den Haag, 1973 (2nd edition), §44, p. 128]. Moreover, Husserl rejects the scientisticopinion according to which one’s own body were situated in a geometrical space."

25

terms of the processing and transmission of information. The information that, especially in the presentera of computers and information sciences, we are eager to recognize in the performance of autopoieticsystems is not a property of their behaviors. We are pushed here to the limitations of the usage ofmetaphors. For we are accustomed to understand the behavior of machines, such as computers, in terms ofinput-output functions. Something is entered and something is returned. Due to a system’s structuredetermination an input-output description can only have a mere symbolic status: it is often an elegant anduseful way of putting into words a system’s behaviors. Inputs and outputs are auxiliaries for a describer(observer) to summarize a system’s performances. They exist in the observer’s domain of descriptions, notin a domain of operations«27».

The difference between these two domains is also apparent from the way in which the concepts ’input’and ’perturbation’ are used. Whereas a perturbation is defined by the individual system’s way ofcompensating it, an input, conversely, is defined by an external observer in order to describe the system.The input exists in the observer’s domain of descriptions, and as such it is a constituent of the definitionof the system (cf. Varela, 1983, p. 150; Varela, 1987). It is meant by the observer as a signal, physical ornot (e.g. sunlight), which exists in itself as an entity independent of the system, by means of which thesystem’s behaviors can be described. It is because of this independence that an input belongs to theobserver’s domain of descriptions. We already noticed that the properties of events exist as compensationprocesses in an individual. Properties of inputs, which are defined beyond these compensation processes,only exist for the observer who describes them. Inputs do not exist in a system’s domain ofoperations«28». Neither does an autopoietic system perceive an input«29».

A major principal point is not made in the literature, which isthat the very idea of autopoiesis (or ratherorganizational closure) entails an impossibility to describe its very closure in exact observational terms. Iwill postpone my discussion of this issue until section 7.1.1, after I have introduced my concept of’critical discourse’. I will discuss there how critical discourse, as a major tool of empirical science, isincapable to catch autopoiesis, and self-reference in general, in observational terms.

2.1.5. Structural coupling

When two systems, living or non-living, enter into an interaction, then this interaction leads to acontinuous change of states in each system. Each system perturbs the states of the other system, and eachcompensates for these perturbations by means of a change of its states. Maturana uses also the terms’coontogenic structural drift’ (1988, p. 47) and ’ontogenic structural drift’ (1987a, p. 344). A simpleexample of this is a boat floating on the water. At each moment the waves and the boat determine eachother’s position and movement. There is a mutual and simultaneous influence«30», without theidentities of the interacting systems being changed (cf. Maturana, 1982, p. 290). An other example is thegrinding of two round slices of glass against one another. In the course of the grinding one slice willbecome concave, the other convex. The structures of both systems mutually adapt. They come to fittogether in a way that was not yet the case before. Maturana calls this the emergence of a structuralisomorphism (Maturana, 1982, p. 290)«31». Such a development is spontaneous, and it takes placeuntil the nature of the interaction changes in a way that at least one of the interacting systemsdisintegrates (e.g. the concave slice becomes too thin and breaks). Finally, an established mutuality of

27. This is a vital difference from Rosen’s ideas on the organization of the living organism. According to Rosen (e.g. 1991), it is thepresence of functionalityinsidea system’s organization, that makes it differ from mechanisms.

28. cf. Varela, 1979, pp. 11/12

29. William James (1890) introduced the notion of "psychologist’s fallacy" precisely in order to emphasize this difference betweenwhat an observer can see, and what he can impute to an observed individual (cf. Goudsmit, 1989a). See footnote 112, page 51.

30. Other terms for this which can be found in Maturana are ’coontogenic drifting’ and ’coontogeny’.

31. In the more established terminology of Watzlawick, Beavin and Jackson (1967) we would call it the occurrence of a’complementary relationship’.

26

state changes can be dissolved, e.g. when the two interacting systems drift apart. Thus two interactingsystems yield changes of states that are mutually influencing.

This type of interaction process which takes place at the interface of a system and its medium, or betweentwo systems, is calledstructural coupling. It is characterized by a reciprocal influence between states andinteraction processes. The way in which a structural coupling between two systems develops in time isdetermined from moment to moment by what we may call the ’states’ of the coupled systems. In theirturn, these states are determined from moment to moment by the nature of the interactions that the twosystems are having.

An observer cannot look around freely in the interaction without also disturbing what he sees.Furthermore, there where a system touches its medium, the observer cannot perceive whathappens«32». He cannot come into the in-between, without the interface ceasing to be an interface.For this reason a system’s niche is not empirically observable.

Like structure determination, structural coupling pertains both to living and non-living systems.Interactions between systems always take place in a way that either a structural coupling is established, orthat the identity of one or more systems is changed. This is true for interactions between living systems,between non-living systems, and between living and non-living systems. The structural coupling is notpresent from the outset of the interaction. There is always a simultaneous mutual influence that can bringa structural coupling within more or less time. It is impossible to perceive, or even conceive, the system’sstate as something beyond the system’s interactions with other systems«33». So far for both livingand non-living systems.

When two living systems structurally couple, then their interactions lead to structural perturbations ineach, that can be compensated by further behaviors of the interactants. This active capacity for adaptationis essential to life, as has been hinted upon already in section 2.1 (esp. footnote 6, page 19).

Like non-living systems, the interactants develop a structural isomorphism at their sites of interaction, sothat a particular structural fit arises between them. It is at this interface that the living system differs fromthe non-living: by the incorporation of perturbations into the living system’s current dynamics of statechanges.

If the perturbations triggered by the interactions between the interactants cannot be compensated by one ofthem, then that damaged system loses its capacity to survive and disintegrates. Then there is no longer astructural coupling between the two systems. Violence is also to be understood in these terms. But there isan alternative to violence: the construction of consensual domains.

2.2. First order consensuality

2.2.1. Consensual domains

It is possible to consider the structural coupling between two living systems at the level of theparticipating individuals and at the level of the interaction itself. Between these two a difference of logicaltype exists, like between the parts and the whole of a Gestalt«34».

When considered at the level of the individual, the multiplicity of forms of structural coupling can beformulated as the individual’s domain of interactions at a particular moment.

32. "... the observer cannot see the encounter of a living system and its niche, which is where conservation of adaptation takesplace" (Maturana, 1987a, p. 346).

33. among which the system’s environment

34. Maturana (e.g. 1978, p. 37; 1987a, p. 340) calls this a difference ofphenomenal domains.

27

On the other hand, the way a perturbation is compensated by an individual may entail a change ofbehavior in the individual. If the individual is in interaction with another individual, then his new behaviorwill affect the other as a perturbation that, let us assume, can be compensated. Then both individualsbecome involved in an ongoing process of compensating perturbations triggered by the ongoingperturbation compensating behaviors of the other. It is considered at this level that a structural couplingmay yield a whole repertoire of recurring interaction patterns. This repertoire is called theconsensualdomainof the interactants. This is the set of shared behaviors that they are able to generate together at aparticular moment.

A different term Maturana uses for such a shared behavior is ’coordination of actions’. Here we maydistinguish between the actions that are coordinated, and the consensual behavioral pattern that theyconstitute. The latter is part of the consensual domain of the two interacting individuals. In section 2.3 wewill see that the consensual behavior in its turn may become an item of coordination.

Above in section 2.1.3 I defined distinctions in terms of elementary operations. A consensual coordinationof actions therefore can also be understood as a consensual coordination of distinctions (cf. Maturana,1987a, p. 359). The interacting persons coordinate the actions through which they make the distinctions.

Events in the consensual domain can be understood as a dance of two (or more) individuals, as in thecourting dances of birds. Such a dance unfolds as a pattern that develops from moment to moment, not assomething planned and prepared in advance. The involved individuals, while compensating perturbations,act momentaneously, thus bringing forth their dance as a consensual pattern.

2.2.2. Description versus operation of consensuality

This consensual pattern is not planned or intended by those involved in it. It emerges as an unexpectedgestalt, constituted of the individual contributions of the participants. Thus, no goal-orientedness in theparticipants’ actions is required, and even less an awareness of what is being performed. Nor is there anexchange of signals.

Any alleged exchange of signals between structurally coupled systems is not something taking place in thedomain of their interactions, but is only a mode of describing this interaction by the observer. Structuralcoupling (cf. section 2.1.5) is a simultaneous mutual influence. The idea of a signal, on the other hand,presupposes that the mutual influence can be taken as a causal process that goes back and forth betweenthe two parts, as in the well known interpunction diagrams of social interactions. Those descriptions aremostly possible and useful, but they do not accord with the simultaneous character of structuralcoupling«35».

To an external observer the structural coupling between two systems may look like a consensus betweenthe participants. The two seem to agree, that is, they have found a modus vivendi, however temporal thismay be. Just as in the description of individual (perturbation compensating) behavior, an observer maydescribe the consensual behavior of two individuals in terms of a particular signification, a particular goal,etc. The signification and the goal, however, belong to the domain of descriptions of the observer, not tothe domain of operations of the observed systems.

The observer may even come to interpret the consensual behaviors as if the participants agreeaboutproperties of their environment. Their consensual behaviors are taken as a shared distinctionof somethingby the participants (cf. Maturana, 1987a, p. 359). But such an object only exists in the domain ofdescriptions of the observer (cf. section 2.1.3), and the coordination of actions by the participants does notentail any perception of such an object. This would be different when the observer describes theconsensual behaviors of individuals that are themselves observers, i.e. persons who make descriptions (cf.section 2.3.5). In section 2.2.3 I will suggest the term ’quasi-object’ in order to denote these objects that

35. Indeed, the very idea of causality is not compatible to the simultaneous mutual influence of structural coupling.

28

exist only for the external observer and not for the observed individual.

But, the reader may object, don’t we see birds sitting on trees, dogs playing with sticks, or cats chasingafter mice? Do the above ideas on objects mean that these trees, sticks and mice do not exist for theanimals that are sitting on them, playing with them or chasing after them? And if they don’t exist forthem, then how should that be possible? The answer to this objection is to be given in terms of the kindsof interactions that are involved. For the human observer, the birds as well as the trees they are sitting onare distinguishable objects. Their objectivity, however, is not a property per se, but an outcome of asecond order consensual interaction between participants of a community of observers.

The observed birds, dogs and cats themselves do not partake in this community. For them the objects thatthey are observed to handle, do not existas objects. The ways in which they are handling the trees, sticksor mice, is a feature of their structural coupling to their respective media. It shows howadequatetheirperformance is (cf. Maturana’s definition of ’cognition’ in section 2.1.2).

Only as soon as the animals were to start to coordinate recursively their ways of handling the trees, sticksand mice respectively, would the latter become objects for the individuals that are involved in thisrecursive coordination of actions«36». Most animals, however, do not attain this level of social andintellectual performance. To describe their behavior in terms of the objects they are dealing with, is thenentirely a matter of conception by the observer, not of their own modes of operation. When describinglower animals in their environments, it may seem a rather innocent and innocuous research style todescribe the animal and the objects it is observed to interact with. For example, we may observe that aparticular fly sets itself on a particular painting by Rembrandt. This way of describing we may call ’naiverealism’. To maintain that the fly interacts with the Rembrandt painting, would confuse the non-linguisticdomain in which the fly operates and the linguistic domain in which the human observer describes the flyand the painting:

"A fly seen walking on a painting by Rembrandt does not interact with the painting by Rembrandt.The painting by Rembrandt exists only in a cultural space of human aesthetics, and its properties,as they define this cultural space, cannot interplay with the properties of the walking fly that definea space of chemical distinctions«37». In other words, we see the fly to be blind to the paintingby Rembrandt because the painting does not exist in the domain of the fly’s distinctions (see Jakobvon Uexküll 1909)." (Maturana, 1980a, p. 28).

Likewise, the gaze of neonates and infants into a particular direction may induce adult observers tobelieve that the child actually sees the mother, its own thumb, or whatever it may be gazing at. However,not until the child has in some basic way been introduced into its social environment’s language system,can one reasonably maintain that the objects the child isseento watch or to handle, do coincide with, oreven resemble the objects the child isactually (if at all) aware of.

Generally this issue is dealt with by Varela in terms of ’logical bookkeeping’ (Varela, 1979, p. 12) or’logical accounting’ (Maturana & Varela, 1987, p. 136). This is to say that the observer should notconfuse his own (legitimately used) domain of descriptions with the domain of operations of the systemshe observes.

An example of consensuality (Maturana & Varela, 1987) is the behavior of animals in flocks, such asdeer. While the flock moves into the valley, one of the animals remains on top of the hill, and onlyfollows when the others have passed the valley. It is then as if the animals agree about the dangers thatloom as long as the flock is in the valley. Birds that fly up in fright thus alarm the other birds.Immediately it appears as if the animals agree about the existence of danger, and as if they are making ashared distinction of the source of the danger. The observer calls the flying up of the first bird a ’sign’ forthe other birds. Such a judgment, however, can only be given by a person who disposes of language and

36. Cf. footnote 56 on p. 33.

37. At least, an external observer would qualify these distinctions in chemical terms.

29

who is able to point at the object of consensus. The question ’what else could they be perceiving?’ is putwrongly: there is no perception, that is: the consensual distinction by the participants cannot be taken asthe distinction of an object that appears to the participants; the consensual distinction does not take placein a domain of second order consensuality; the participants do not create a recursion in their coordinationof actions. I will discuss this in section 2.3.

2.2.3. Quasi-objects

The terms ’quasi-object’ and ’quasi-observer’ are not used by Maturana. I am introducing them here,being convinced that they are useful for a proper understanding of Maturana’s ideas on language.

Let us return, for a moment, to Maturana’s example of the fly and the Rembrandt painting. When the flydoes the things it does as he is seen to sit on the painting, it is making distinctions, but that is not to saythat it discerns by these distinctions some figure against some background. Rather, it means that there issomething that an external observer could see as an object in the fly’s environment, which the latter doesnot see, though it is seen to interact with it.

In fact, when the fly is seen to interact with the painting, it is making distinctionswithout seeing it as anobject. It ismakingthe distinction through its actions, but it is nothaving the result of the distinction as akind of object. Examples of this are easily found in the behavior of animals and young pre-linguisticchildren.

For instance, the ease with which a young child recognizes persons or things does not allow the externalobserver to conclude that the child is able to perceive the recognized person/thing as an object in the realworld (the object and the real world as conceived by the external observer), but only that the child hasacquired a particular capacity to deal with his environment (as the external observer sees thisenvironment).

For instance, by making first order distinctions, an individual doesnot distinguish between:’map’ versus ’territory’«38»’me’ versus ’you’«39»’form’ versus ’content’’subject’ versus ’object’’process of distinction’ versus ’outcome of distinction’«40»’impression’ versus ’expression’’inside’ versus ’outside’’that what is experienced’ versus ’the way in which it is experienced’«41»

All these things arenot meant when an individual is said to make a first order distinction, but they maybe irresistably evident to an external observer.

Let us use the term ’quasi-object’ as different from ’object’: ’object’ refers to that what is observed inlanguage by second order distinctions; ’quasi-object’ refers to that what is distinguished by a first orderdistinction. A quasi-object, therefore, does not exist for the individual (to be called thequasi-observer) asan object«42»; it can only be described by an external observer as the object that the latter ascribesto a non-observing individual as if the latter were also observing it and having it as an object. This is the

38. cf. Korzybski’s dictum ’the map is not the territory’ (see Bateson, 1979, p. 32)

39. cf. section 11.6

40. Cf. Kauffman & Varela, 1980, p. 205: the ’act of distinguishing’ as indistinguishable from the ’distinction’.

41. cf. footnote 81, page 41

42. The notions of ’quasi-object’ and ’quasi-observer’ resemble Husserl’s idea of ’fungierende Intentionalität’, as well as Sartre’s(1943, pp. 19/20) notion of ’non-thetical’ or ’non-positional’ consciousness. I will return to this in section 5.3.1.

30

case for example when we observe a cat chasing a mouse. Then the cat is ascribed the perception of thismouse, as if the cat had made a second order distinction, and had observed the mouse as an object.Rather, in fact, the cat can be observed to act in such a way that by his actions he distinguishes themouse. The mouse can be said to be a quasi-object for the cat. It isas if the cat sees the mouse. Thisdoes not imply that the mouse will be quasi-eaten (as one might suggest). Instead, it implies that the’chasing’ and the ’eating’ are an external observer’s descriptions about which the cat will not agree ordisagree, simply because the cat doesn’t make descriptions at all«43». But not only with respect toanimal behavior does the notion of quasi-object make sense. It is also useful for describing first ordercoordination phenomena occurring between humans«44».

A similar difference between the positions of observer and observed has been met already in Maturana’sconcepts ’environment’ and ’medium’ (cf. 2.1.2). We may formulate it thus: the environment in which anindividual is observed is by definition a quasi-object for this individual. Thus, it is the observer whodescribes the quasi-observer’s usage of quasi-objects. When it is said that an individual incorporates aparticular object into his modes of functioning (as will be done in section 4.2), this is to say that theindividual is observed to use the object of which he is a quasi-observer.

2.3. Second order consensuality

Let us introduce the concept ofsecond order consensuality. This idea is used in order to explain languageand perception as special types of structural coupling. Whereas first order consensuality could beunderstood as a contact between two systems, second order consensuality is a refinement of this, one thatenables the interacting individuals to use their first order consensus as actions that are connotative ofsomething else. An observer may not only assign a meaning to the consensual behavior of animals (cf.section 2.2.2), but also to his own consensual interactions. Like the old-fashioned pharmacologist (cf.section 2.1.2) took the performance of laboratory animals as indicators of something else, so may thelanguage user with respect to his own consensual performances. If he succeeds in agreeing with hisinteraction partners about their current consensual behaviors, thenlanguagearises, as a second orderprocess«45».

Like first order consensus pertains to the coordination of actions, so second order consensus pertains tothe coordination of the very coordination activities. There is not only collaboration, but also a secondorder collaboration in establishing the collaboration. This is according to Maturana the basis of language,and he coins this activity ’languaging’.

In language, the consensual pattern of actions itself becomes (what I called) a quasi-object for theinteracting individuals in regard of which they can coordinate their actions. That is, for an externalobserver it may seem that the interacting individuals coordinate their actions in respect of their ownprevious interaction patterns.

43. Cf. ter Hark (1990, p. 238) on the difference between men and cats in respect of their ’intentions’ to do something (e.g. catch abird).

44. It is in this way that we have to understand Bion’s (1961) notion of ’basic assumption’, which according to him pervadespsychodynamic therapy groups. Rioch (1972) explains these as: "the assumption that is basic to the behavior. It is an "as if" term. Onebehaves as if such and such were the case. (...) [O]n many different levels, by observing the behavior of individuals and of groups, onecan tease out the basic assumptions on which they operate. Bion uses the term to refer to thetacit assumptions that are prevalent ingroups, not to those which are overtly expressed. The basic assumptions of the basic assumption groups are usually outside ofawareness. Nevertheless, they are the basis for behavior. They are deducible from the emotional state of the group." (p. 21)

45. cf.: "What an observer sees (...) is that the participants of a consensual domain of interactions operate in their consensualbehaviour making consensual distinctions of their consensual distinctions, in a process that recursively makes a consensual action aconsensual token for a consensual distinction that it obscures. Indeed, this is indistinguishable from what takes place in our languagingin the praxis of living." (Maturana, 1987a, pp. 359-60)

31

The coordination of actions is itself an action that can be coordinated. Language is to be considered as anongoing process in which a recursion of coordinations takes place. What was at the previous moment anact of coordination, at the next moment is itself the item to be coordinated«46». Like in plain (firstorder) consensuality two dogs may chase after a rabbit and coordinate their behaviors thus, so also two’languaging’ individuals may coordinate their actions in regard of their preceding coordinations anddevelop a shared position in regard of these. Thus a ’something’ comes into being, towards which theymutually orient each other, and in respect of which their coordinated actions become meaningful«47»(Maturana speaks of ’tokens’). Thus, the languaging individuals distinguish their coordinated behaviors,take them as a token for something that in fact comes into existence due to this token. It is only for theparticipants of this domain of second order consensuality that their coordinated actions become a token.Furthermore, the act of coordination and that which is coordinated are of the same kind. For humanspeech, to take the predominant case, they are all in the domain of sounds and movements.

Two persons start laughing together. Originally there may have been something that triggered an initiallaugh in one or both of them, but the true laughing occurs as soon as the participants notice each otherlaughing. It is then the common behavior (shared laughter) that serves as a token for something which isfunny. Only then does this ’something’ attain its dimensions. It is constructed through the shared laughter.Furthermore, each participant ’notices’ this shared laughter and at the same time contributes to it! This’noticing’, however, is not a matter of perceiving it as an object, but of distinguishing it.

The shared laughter is taking place between two participants, and, by laughing, each person notices theongoing (previous) shared laughter. The (new) laughing is a second order distinction of the alreadyongoing laughter. The shared laughter is a quasi-object for those involved in it. They (second order)distinguish it (from the point of view of an external observer), but it does not exist for the participants asan object. Instead, through it, as through an optical instrument, the object of laughter becomesdemarcated.

The notion of a quasi-object is useful for us here in order to describe an individual’s acts of distinction.Then ’the thing that is distinguished’ can be indicated by the external observer as a particular quasi-object, such as the shared laughter, or a person’s individually experienced sensations. This allows anexternal observer to speak of second order distinctions as distinctions of (first order) distinctions, withoutthe latter being observable for the observed individual himself. For the external observer cannot everknow these individually experienced sensations, nor can he know that a person uses these sensations forthe construction of particular objects. He cannot, but he may legitimately postulate them for the purposesof his discourse, as long as he does not claim that the observed person knows them as well«48».

2.3.1. The emergence of objects

What happens when two persons recursively coordinate their preceding coordinations in a continuousprocess? We may conceive of this as an interaction process in which new sounds and movements aremade as ways to agree about the meaning of preceding sounds and movements. Notice that this meaningis a pragmatic one«49», not one put in explicit terms. Through the recursion of coordinations, thecoordinated behaviors become tokens for objects that are brought forth simultaneously with their tokens.

46. Cf.:"... human beings as living systems operating in language constitute observing, and become observers, by bringing forthobjects as primary consensual co-ordinations of actions distinguished through secondary consensual co-ordinations of actions in aprocess that obscures the actions that they [*] co-ordinate." (Maturana, 1987a, p. 362)

([*] ’they’, i.e. the ’secondary consensual co-ordinations’)See also footnote 66 on page 36.

47. The notions of second order distinction or second order coordination are comparable to what Husserl called ’thematischeIntentionalität’ and to what Sartre (1943, e.g. p. 19) called ’thetical’ or ’positional consciousness’.

48. That would be a case of the "psychologist’s fallacy" (cf. footnote 112, page 51).

49. The objects that arise thus are what I will call below (chapter 3)naive objects.

32

We come to perceive the subject matter of our language through our sounds or movements. The metaphorof a slide projection is useful here: our words (movements) of the moment are like a spotlight by meansof which our preceding words or movements are projected as a slide, thus enabling us the perception ofthe object of which the slide becomes the token. Likewise: our current coordinations of our precedingcoordinations are the ways in which these preceding coordinations become the tokens of an object that weare bringing forth«50».

The recursion that happens in consensual coordination of actions between interactants can also beformulated as a recursion in consensual distinctions (cf. Maturana, 1987a, p. 364). The interactantsdistinguish together their coordinated actions. This shared distinction is in its turn a coordinated actionthat can be distinguished consensually. Thus the recursive coordination of actions can be seen as arecursive«51»process of consensual distinctions of consensual distinctions.

An example of this can be found in conversations between people at the start of which it is not yet clearwhat they are to talk about. A talking is commenced, and while conversing the subject matter crystallizes.The talking itself is the way in which the participants orient one another towards something. Byperforming sounds or movements they may succeed in taking preceding behaviors as tokens for an object.They do so by coordinating new sounds or movements, which in turn will be situated as tokens byfollowing coordinated utterances.

I experience something comparable when writing a new text. The first few pages are trusted to the paperonly after many painstaking days or months. Only by the time that the text seems to me to be about a’something’, it becomes easier to add new words to it. Only then it seems as if the text is aboutsomething preexistent«52», and as if everything newly added to it serves merely to accentuate thewords already written as signifiers of something. Thus this ’something’ becomes a thing, and even can betaken as a criterion for the descriptive qualities of the text«53». For a text that describes a topic wellenough it is virtually impossiblenot to believe in the text independent existence of that topic. Throughoutthe text is has received objective presence. Likewise, Maturana maintains that we construct reality in such

50. Cf. Shotter (1987a): "For usually, we ’see through’ the language we use... (...) .. our words are a transparent means throughwhich we achieve a feeling for, and sensible contact with, those around us."

Intentionality as a social activity between persons takes shape, according to Shotter (1989), as a way of actinginto a socialsituation, instead of acting out of a plan or a scheme:

"... in joint action, people create between themselves an "organized setting" (Bartlett, 1932), which can then function as analready partially structured contextinto which to direct their next actions. Thus rather than acting ’out of’ an inner schema or plan, theymay act ’into’ the opportunities and invitations offered them by their own current situation." (1989, p. 9)

51. This strongly resembles Merleau-Ponty’s conception of visual perception, as developed in his last work. He understands (visual)perception as a ’seeing of seeing’. At this place we can recognize the correspondence with the idea of ’distinction of distinction’. Kwant(1989, p. 673) cites the following passage of Merleau-Ponty: "[Mon corps] se voit, il est un visible, - mais il se voit voyant, mon regardqui se trouvelà sait qu’il est ici, de son coté à lui". (Merleau-Ponty, 1964a, p. 324) ["[My body] sees itself, it is a visible - but it seesitself seeing, my look which finds it there knows that it is here, at its own side" (1968a, p. 271)/working note december 1960]

Kwant also mentions the use by Merleau-Ponty of the expression "voir de voir". This also resembles the idea of ’distinction ofdistinction’. There is (what Merleau-Ponty calls) the perception of a perception. This may be initially a vague perception, but theperceiver searches for a position and a perspective from which his vision improves. This, Kwant maintains (p. 673), would beimpossible if vision would not be also a vision of the (imperfections of the) performed vision. Thus, perception is understood as a wayto stabilize and optimize our percepts. The stabilization activities can be recognized as second order coordinations.

Cf. also: "Thus since the seer is caught up in what he sees, it is still himself he sees: there is a fundamental narcissism of allvision. And thus, for the same reason, the vision he exercises, he also undergoes from the things" (Merleau-Ponty, 1968a, p. 139). ["Desorte que le voyant étant pris dans cela qu’il voit, c’est encore lui-même qu’il voit: il y a un narcissisme fondamental de toute vision; etque, pour la même raison, la vision qu’il exerce, il la subit aussi de la part des choses" (1964a, p. 183)]

Flynn (1973, p. 124) adds to this: "The openness of vision upon the visible is not achieved across a distance, but by a doublingof a universal flesh, the visible, upon itself."

Thus, we recognize a parallel to second order distinctions (cf. also footnote 21 on page 23, where I suggested that Merleau-Ponty’s concept of perception as ’interrogation’ is comparable to ’first order distinction’).

52. Cf.: "Nothing pre-exists its distinction" (Maturana, 1987a, p. 370).

53. A comparable reversal is mentioned in footnote 61 on page 34, focussing upon the presumed pre-existence of ’reality’ to’computer model of reality’.

33

a way that the world constructed obscures the process of construction«54».

We all know the situation in which a person, let us prosaically call him A, acts as a speaker and twoothers, named B and C, listen to him. During his talk A may be inclined to address his words to B, andnot to C. We label this: "A finds in B a better audience". What does this mean? We are inclined to say:"A feels that B can follow him better, joins him better in thinking and feeling, sees better before him whatA is talking about, and the like". And we may wish to add: "B emits signals to A, that notify A aboutthis". Such a common sense description usually suffices, but using Maturana’s ideas we had better say: Aand B together make an object in the course of their conversation. The way in which B behaves is acontribution to the construction of this object, more than the way in which C behaves. B’s behaviorenables A better to see ’in’ the eyes of B (or in B’s other responses) that which he is intending to put intowords. B’s listening behavior is constructive, not so much because it contains ’signals’ of the type "Iunderstand", as because it is a substantial contribution to the construction of an object«55».

This listening process cannot be described in terms of information transfer. Nor is there a state of affairsbeing communicated. Rather, the listener is enabled to learn how to join in with the speaker’s perception,thus supporting it and arriving at a similar perception. To the extent the listener sees better what thespeaker is describing, to that extent he is more facilitating to the former to formulate his descriptions.Nay, that what is the matter even comes into existence as an observable entity through the activity of ashared description of it. This is also true for topics discussed in psychotherapeutic encounters, whetherthey are concerned with patients’ unconscious motives or with other issues into which ’insight’ is sought.

2.3.2. Some other examples of second order coordination

Maturana gives the example of the dog and the man, playing a game in the park. The man throws a stick,the dog chases it and brings it back. We may understand this as a basic pattern of interactions, a firstorder coordination of actions that goes together with a shared experience (a consensual distinction). Thenan interruption happens. The man starts a conversation with an acquaintance. The previous coordination ofactions with the dog is no longer warranted. Then the dog barks. Now this barking, according toMaturana, is a case of ’languaging’«56».

The barking is a way to deal with the disrupted first order coordination of actions (the game). If the manis to respond to the barking, then a second order coordination may arise between them that leads to aconsensual distinction of the disrupted game. This can be compared to the consensual distinction of aninterference pattern of two tones«57». Since the game was a coordination of actions, this new ’game’is a second order coordination, i.e. a coordination of the previous game. Notice that the discoordination ofactions (the disrupted game) is a circumstance due to which a second order coordination becomes possibleat all. Notice also that what the barking exactly means is a matter of interpretation to the observer. One isinclined to interpret it as something like ’please throw the stick’. Clearly, it is the observer, not the dog,

54. cf. footnote 66 on page 36

55. In this respect it is illuminating to read how Mahrer describes himself being oriented by a client to that which is the matter forthe client: "The therapist hears the words the client says: "The window was open." The therapist receives the way in which these wordsare said, the quality of the voice, the tone and amplitude. (....) As the behavior and words of the client are received by the therapist, theyconstruct something in the general attentional field of the therapist. (....) One way of saying these words may construct a whining,petulant child (....) Another way of saying these words may construct a fearful scene in which something ominous and menacing isoutside the open window." (Mahrer, 1983, p. 242).

56. The reason for denying the so called ’animal languages’ and also the human ’body languages’ the status of a genuine language,is that it is only in the eyes of an external observer that the performed consensual behaviors appear as tokens for some object. For theindividuals involved in the ’nonverbal language’, however, this symbolic status is not there. Only some higher mammals (e.g. the greenmonkey) have been observed to use particular and specific sounds or movements to denote a particular danger, such as snails or eagles,and only for those restricted situations their behaviors seem to have a consensually assigned meaning. But even then, this consensusitself is not something that can be played with, as in human language usage.

57. I will return to this in section 3.2 (cf. footnote 84, page 43).

34

who is able to put it into words like these. To say that the stick is a quasi-object, however, would missthe point: precisely because the dog comments upon the disrupted game (and only to the extent that this isthe case), the stick does exist for the dog as an object, however different the dog’s perception of it maybe. If the barking would not refer to the stick, this would become apparent at the moment of throwing it:the dog would not go for it, or with some delay. On the other hand, the game itself exists for the dog as aquasi-object: this is something the dog only (first order) distinguishes, viz.by playing it and by barkingwhen it is disrupted.

Other examples of second order distinctions/coordinations are situations where a person distinguishes aconcurrence of several experiences. In infants this can be recognized by the onset of laughing«58».These experiences do not exist for the child as perceivable objects, though the child seems to be able torecognize them and integrate them into patterns about which fun can be made.The child’s distinction ofthe concurrence of two or more experiences is equivalent to a coordination of these twoexperiences«59».

We may end this section with a metaphor. A favorite figure in 18th century music was that of a themeand variations. After the theme had been exposed, several variations of it were given, in each of which theoriginal theme could be recognized. The experience of recognizing the common thread throughout thevarious variations, in fact, is a matter of coordinating them into an object, using the variations as thetokens through which it can be perceived.

However, distinctions (of distinctions) are not made with respect to pre-existing objects, but they arise ascoordinations of experiences (the latter being the processes of compensation of perturbations). Anobserved object therefore is to be understood as a kind of focal point, a point of condensation«60»offirst order distinctions which (by a second order distinction) are taken as tokens for this object«61».

58. For example, little Marthe (at the age of 22 weeks and 2 days) started to laugh when apparently distinguishing the followingconcurrence. When watching herself (and her father) in the mirror, she looked but did not seem to enjoy it; also, when being hopped shedid not seem to experience anything particular (being accustomed to being hopped very often). However, looking in the mirrorwhilebeing hopped made her roar with joy and apparent recognition of herself in the mirror. On the other hand, when dressed in her wintercoat with woollen hat, so that only her eyes and nose were visible, she did not laugh when being hopped in front of the mirror: didn’tshe recognize herself in these cloths? This conclusion is hardly resistible for the observing (and hopping) parent.

Likewise, noticing other concurrences of experiences had a comparable effect, for example when she recognized regularities insound patterns exchanged by her and her parents, or regularities in patterns of sound and movement. Each regularity, thus recognized, isa quasi-object for her, as is each more elementary experience. What she perceives is never the regularity (or the experiences thatconstitute it), as an observer would recognize it. Rather, she laughs contingently to the observer’s perception of these regularities andpatterns, and it is this contingency that makes the observer believe it is such a pattern or regularity that she distinguishes when shelaughs. Furthermore, it is evident for the observer that such a pattern consists of more elementary visual, auditory and/or kinestheticsensations, even though these do not exist for the child as objects. Finally, it is not unlikely that the parent-observer’s ownstyleofpresenting to the child the variety of contingent experiences, is also part (and not in the least!) of the pattern presented to the child. It ishere that the consensuality is part of the pattern, even if the parent-observer does not recognize this consensuality as an object (as mightdo a meta-observer).

59. Notice the resemblance to the idea of convergent measurement as proposed by Campbell & Fiske (1959). The coordination ofdistinctions is here comparable to the convergence of various measures of the same object, except for the fact that the coordination ofdistinctions does not presume a pre-existing object!

60. I am using the term ’condensation’ here in a way not entirely unlike Freud’s classical idea of a dream image into which avariety of unconscious motives are condensed. Likewise, a variety of distinctions can be condensed (i.e.: coordinated) into an object. Itis remarkable that we find a comparable notion (called ’fixation’) in the writings of Vico (cf. section 3.3) on the primitive perception ofobjects from sensations.

61. A computer simulation of perception, on the other hand, is supposed to act upon a particular concrete object that issues sensorydata, and to use these as its input. The fact that such an input is to be fed to the computer in a format dictated by the program, isusually not felt by AI researchers to count as evidence against the idea of a pre-existent reality. Thus, unlike perception in autonomoussystems, the idea of simulating perception in computers does require a pre-existing object, that is thought to serve as an input. This isthe reverse order, though most of current theories of perception take it as the only legitimate one (cf. Varela, 1991, p. 187)! They are infamous company, for Quine too, in his paper ’Epistemology naturalized’ (1969, p. 85) maintains:

"A is epistemologically prior toB if A is causally nearer thanB to the sensory receptors. Or, what is in some ways better, justtalk explicitly in terms of causal proximity to sensory receptors and drop the talk of epistemological priority."

But what if the core of the problem is that we, as external observers, cannot decide what the words ’causal proximity to sensory

35

2.3.3. "Objects as warrants for eigenvalues"?

This focal point can be understood as the indicator that an eigenvalue has occurred. Generally, vonFoerster’s concept of ’eigenvalue’ concerns the value of a function after a repeated (iterative) applicationof a particular operation«62».

Likewise, the manipulation of an object (i.e.: an object as perceived by an external observer, and thereforea quasi-object for the observed individual), seeing, hearing, touching, tasting and smelling it, is understoodby von Foerster as a complex chain of operations, each of which is performed repetitively upon the outputof a previous operation. This may eventually result in a stable performance, which is an eigenvalue of theobserver’s behavior. The emerging objects are warrants«63»of the existence of these eigenvalues.

Maturana would agree with von Foerster that for object perception some stabilized performance isrequired. But for Maturana distinctions are no calculatory operations. Characteristic for a distinction is, asI stated in section 2.2.3, that its process is not distinguished from its outcome. It is not an operator thatstabilizes into a particular value that is distinct«64» to this operator. Only when an external observerinvents, as I called it, a quasi-object that is supposed to be the subject matter ’of’ the distinction, the’operation’ of distinction is taken apart from its outcome or ’value’.

Furthermore, contrary to von Foerster, Maturana considers the consensuality of distinctions as necessaryfor the bringing forth of objects. It is through the attainment of consensual distinctions that individuals areable to create objects in language. Only after an individual has attained some familiarity to the use oflanguage, he may be able to perceive new objects without consensus with others.

Thus, a stable pattern of interactions between individuals is generally understood by Maturana in terms ofa first order consensual domain. A second order consensual domain now can be understood as a (secondorder) stabilized pattern of dealing with fluctuations and irregularities of (first order) stabilized patterns.

By modulating a previously attained stable interaction pattern, the interacting individuals make thisprevious interaction pattern (consensual distinction) into a token for an object. And so on: the interactingindividuals may modulate each newly attained pattern, in an infinite progression.

It is precisely the constraining effect of these second order patterns that psychotherapists often try tomitigate when working with a patient. Though the language is entirely different (e.g.: ’how to accept yourown feelings’) the basic issue is often the change of dealing with one’s first order distinctions. Forinstance, these first order distinctions may have been used for the perception of objects that in fact impededifferent use for the construction of other objects. The objects perceived at moments of frustration, for

receptors’ are to denote? What if the object that is radiating its physical influences into the sensory receptors of the experiencingorganism is the outcome of a consensual construction? If we are interested in what happensbeforeany second order consensus has beenattained, then we should not presuppose the physical objects, as we perceive them, to be the causal agents of any sensory impressions!

62. For example, a repeated application of the operation ’square root’ approximates the value 1 for any positive initial input value.

63. See von Foerster, 1981. Van de Vijver (1989, p. 252) points at the title of the French translation of this paper (von Foerster,1977) in which ’tokens’ are translated as ’gages’. She notices that the term ’token’ is used here by von Foerster as ’warrant’, viz. awarrant of the existence of the eigenvalue in the observer’s performance.

We find in Maturana’s text traces of von Foerster’s influence:Cf.: "Or, in other words, I claim that such recursive consensual coordination of consensual coordinations of actions or

distinctions in any domain, is the phenomenon of language. Furthermore, I claim that objects arise in language as consensualcoordinations of actions that operationally obscure for further recursive consensual coordinations of actions by the observers theconsensual coordinations of actions (distinctions) that they coordinate. Objects are in the process of languaging consensual coordinationsof actions that operate as tokens for the consensual coordinations of actions that they [*] coordinate." (Maturana 1987a, p. 360)

([*] my own reading: ’they’, i.e. ’objects as consensual coordinations of actions’])

64. Von Foerster calls the eigenvalues complementary to the operators that generate them (cf. von Foerster, 1981, p. 278).

36

instance, are often bearing properties related to some warded off or intolerable experiences. In a therapy,on the other hand, the objects that a patient (or configuration of patients, e.g. a family) has constructedbecome less pertinent and less rigid during therapy. I will return to this topic when discussing thecessation of perception (of objects) due to self-reference (see 4.2 and 7.3.3).

2.3.4. Objectivity and solipsism

Objects, as understood by Maturana, do not ontologically precede the coordinating actions of the personswho construct them in language. Nor does the signification of words precede the things to which theyapply. There is a simultaneity, which is, however, not a symmetrical one. There is no kantian world ’ansich’ on the one hand, and on the other hand a domain in which that world is symbolically represented.Rather, the relation between words and objects can be understood best as an instrumental one: objectsarise through our significations. We will return in section 7.3.2 to this asymmetry when dealing with theconcept of a one-sided boundary.

Language is a game with words, in which words become the instruments to bring forth theobjects«65». The perception of objects hinders the participants to see that these objects are theoutcomes of coordinations, as well as that the things coordinated are themselves coordinations ofactions«66». This is not given in the perception of an object; only when an observer pays special

65. This can be compared to the usage of a stick by a blind man. The sensations that the stick triggers in the man’s hand areintegrated into a whole. As a result, the blind man does not perceive the sensations in the palm of his hand, but rather the object that issituated at the far end of the stick. He constructs this object by incorporating the sensations of the palm of his hand, as if he learns toperceive through them. This idea can be found in literature both on phenomenology (Merleau-Ponty, Polanyi) and on quantum physics(Bohr). I will return to this in section 6.2.2

An interesting parallel can also be found in Heidegger, of whom I follow IJsseling’s (1990) reading:"Heidegger even comes to maintain that the moment at which the word is no longer understood as a showing but instead as a

sign, is one of the most important moments in the history of truth and of being. For it is from that moment on that truth is understoodas correspondence and being as being on hand. Originally the word is not a sign of something already existent, but it realizes thatsomething can appear as something. Heidegger writes: ’Only in the word, in language things become and are’ [footnote: ’Im Wort, inder Sprache, werden und sind erst die Dinge.’ (Einführung in die Metaphysik. Niemeyer, Tübingen, 1953, p. 11)] and ’if our beingwould not be in the power of language, each being would be inaccessible to us, both the being that we ourselves are and the being thatwe ourselves are not’[footnote: ’Selbst wenn wir tausend Augen und tausend Ohren, tausend Hände und viele andere Sinne und Organehätten, stünde unser Wesen nicht in der Macht der Sprache, dann bliebe uns alles Seiende verschlossen: das Seiende, das wir selbst sind,nicht minder als das Seiende das wir selbst nicht sind.’ (Einführung in die Metaphysik, p. 63.)]. In the (real) speaking something isshown and a certain clarity comes, but simultaneously in this showing and in this clarity something retreats and hides itself. ’Languageis the clearing-and-concealing advent of Being itself’, is written in the letter on humanism" [footnote: "Sprache ist lichtend-verbergendeAnkunft des Seins selbst.’ (Brief über den Humanismus. Klostermann, Frankfurt a/M 1947, p. 16)]. (my translation of IJsseling’s textfrom the Dutch; Heidegger, 1962)

{"Heidegger gaat zelfs zo ver te beweren dat het moment waarop het woord niet langer begrepen wordt als een tonen maar alsteken, één van de belangrijkste momenten is in de geschiedenis van de waarheid en van het zijn. Vanaf dat ogenblik wordt namelijk dewaarheid begrepen als overeenkomst en het zijn als voorhanden zijn. Oorsponkelijk is het woord geen teken van iets dat er reeds is,maar het bewerkstelligt dat iets als iets kan verschijnen. Heidegger schrijft: ’In het woord, in de taal worden en zijn de dingen pas’[voetnoot] en ’zou ons wezen niet in de macht van de taal staan, dan zou elk zijnde voor ons ontoegankelijk zijn, zowel het zijnde datwijzelf zijn als het zijnde dat wij zelf niet zijn’ [voetnoot]. In het (echte) spreken wordt er iets getoond en komt er een zekere klaarheid,maar in dit tonen en in deze klaarheid onttrekt en verbergt zich tegelijkertijd iets. ’Taal is lichtend-verbergende aankomst van het zijnzelf’, staat er in de brief over het humanisme [voetnoot]."} (IJsseling, 1990, pp. 51/2)

66. Maturana (1987a, pp. 359/360, 362) describes this in terms of an obscuration of the (coordinated actions that constitute the)earlier distinctions by the later distinctions. E.g.: ".... objects arise in language as consensual co-ordinations of actions that in a domainof consensual distinctions are tokens for more basic co-ordinations of actions that they [*] obscure." (Maturana, 1987a, p. 362)

([*]’they’, i.e. the ’consensual co-ordinations of actions that are tokens’)and: "Language is a system of recursive consensual co-ordinations of actions in which every consensual co-ordination of actionsbecomes an object through a recursion in the consensual co-ordination of actions, in a process that becomes the operation of distinctionthat distinguishes it [**] and constitutes the observer." (Maturana, 1987a, p. 364)

([**] ’it’, i.e. ’every co-ordination of actions’)

37

(critical«67») attention to the way in which the perception process takes place, may it becomeapparent. It is precisely due to this absence of attention to the perception process that the things aboutwhich we are speaking can appear to us as objects, and therefore as independent of our languagingactivities«68».

An important implication of this is that an absolute and objective reality is impossible, one that existbeyond the language that we speak with one another, even if it mayseemso when we make each otheradvertent of something. Eventually we always need other people for the confirmation of our impressions.Solitarily we cannot bring forth objects. This brings us to wonder whether we are facing here a peculiarkind of solipsism. Why is there no room for a ’rock bottom’ of hard facts?

According to Maturana reality is strictly related to the way in which it is constituted in language. Asmany language communities and sub-cultures exist, so many realities are possible. We may speak of apluralism. Maturana calls it a ’multiversum’. He emphasizes that together we inhabit such a multiversum.Because of this we should have consideration for the reality that other persons may constitute together,and nobody may claim a privileged access to ’the only real reality’, simply because there is no such athing. Objectivity is not possible but through the awareness that reality is constituted in our language.Maturana calls such an idea of objectivity ’objectivity in parentheses’.

He maintains a) that our perception is relative to our biological constitution«69»; and b) that the onlyworld we can talk about is a world of objects (cf. Maturana, 1983, p. 257; 1987a, p. 331).

Those readers of Maturana who perceive here an element of solipsism do not distinguish between thecompensations a system makes in order to maintain its autopoietic organization invariant on the one hand,and on the other hand the linguistic interpretations and names a person may assign to the medium inwhich he lives. It is then understood as if the construction of reality is made by an individual in thecourse of his autopoiesis, as if the autopoietic functioning of the individual were his domain of realityconstruction. Those interpreters who believe so, are bound to conclude a particular solipsism. They are ledto believe that the autonomous individual, according to Maturana, has his own world about which otherindividuals are not entitled to formulate propositions. The outsiders are believed not to have as directaccess to it as he does himself.

The grounds for this reproach can easily be recognized. It departs from Maturana’s idea that eachorganism interacts with its medium in a way that is state determined, i.e., a way that is only determinedby the way the organism is perturbed by a particular event, and not by any purported ’an sich’ qualities ofthis event (cf. section 2.1.2). The distinctions the organism makes in its medium are related only to itsown functioning.

However, Maturana’s idea of objectivity does in no sense amount to a solipsist position. One might cometo believe this, not in the least due to readings of Maturana such as by Dell (1983), who disputes thissolipsism in a way that is more of a hidden support than its opposite. The only compelling argumentagainst the solipsist interpretation is that according to Maturana objects can only arise in second orderconsensual domains, i.e.betweenindividuals, and never in a private domain of an isolatedindividual«70», as would be required for a solipsist interpretation (e.g. Johnson, 1991). I will return

67. I am alluding here to my discussion of critical objects and critical discourses in chapter 4.

68. cf: "... objectivity arises in language as a manner of operating with objects without distinguishing the actions that they obscure.In this operation, descriptions arise as concatenations of consensual co-ordinations of actions that result in further consensual co-ordinations of actions which, if done without distinction of how objects arise, can be distinguished as manners of languaging that takeplace as if objects existed outside of language." (Maturana, 1987a, p. 363)

69. which seems to be related to, if not derived from Merleau-Ponty’s (1945) idea that our bodily existence is the basis ofperception

70. If there is any ’ism’ that we wish to assign to the idea of objectivity between parentheses, then it would be called, analogouslyto the term ’solipsism’, something like ’socialism’ or ’communism’; terms which have already their own history of (changing)significations. Perhaps ’consensualism’ might do.

38

in section 11.6 on this issue, when discussing the phenomenon of projective identification and the limitsof observing it from an external perspective.

Though the constitution of objects is not a solitary affair, this does not mean that the mere act of namingan object in a conversation, and agreeing on it, would bring it into existence. Though objects do ariseduring the recursive coordination of actions, each of the interacting individuals is having his own bodilyexistence in the first place. And it is in the course of each person’s "praxis of living", as Maturana (e.g.1987a, p. 326) calls it, that new perturbations may pop up that are to be dealt with and compensated.

It is even possible, and indeed often the case, that a person is caught in a language game to such anextent that he does no longer use his original primary distinctions. In common language: this person isdisconnected from his own feelings. It may be desirable then that this person returns to his domain of firstorder distinctions, from where he may make a fresh start to perceive objects«71».

First order distinctions constitute the experiences due to which our perceptions seem to be ’objective’.However, it is not an independent ’an sich’ reality that we dash against, which makes upon us theimpression of something ’there’ which we did not invent or construct ourselves. Indeed, if there isanything that we are not able to steer, it is the course of our own embodied praxis of living. We do notchoose our experiences, nor do we control them. Their inevitability, when coordinated with theexperiences (first order distinctions) of others, gives rise to the perception of objects. Objects are’objective’ not due to some independent existence, but due to the impossibility for us«72» to controlour experiences (first order distinctions).

Maturana does not take the step to consider physical reality as the ’true’ basis, the rock bottom uponwhich reality is made. Like Gadamer«73», he does not believe that a particular animal’s biologicaluniverse is a transformation of a physical substratum upon which it is based as an ’an sich’ bottom.Instead, according to him physical reality is as consensually constructed as other realities (1987a, p. 373).

Finally, Maturana’s notion of ’illusion’ is of our interest. According to him, illusion is not distinguishablefrom ’correct’ perception "because a closed neuronal network cannot discriminate between internally andexternally triggered changes in relative neuronal activity. This distinction pertains exclusively to thedomain of descriptions in which the observer defines an inside and an outside for the nervous system andthe organism." (Maturana, 1978, p. 46). Therefore, the discovery of an illusion is not so much thereplacement of a perception (hence to be considered as ’erroneous’) by a more correct perception, butrather the decision by an observer to discard the previous perception in favor of the current. When, lateron, the observer again decides that his current perception is also an illusion, it will be a matter of justreplacing it, not a matter of coming more close to reality ’as it really is’«74».

71. cf. sections 4.2 and 7.3.3

72. Cf.: "For Merleau-Ponty (...) to think perception does not mean to discover in perception a logical and a priori principle whichrenders it intelligible. To think perception as the unreflected life of consciousness means to recognize that perception is an absolutelyfundamental knowledge for which there is no explanation and no justification, other than its own de facto existence. (.. ..) It is theprogram not of a phenomenology of the Hegelian sort, which attempts to transform the prereflective life of the mind into an intelligiblemeaning, but of a philosophy for which the last word is not the basic intelligibility but the unsurpassable contingency of everything."(Madison, 1981, pp. 161-2)

73. WM (1960), p. 428; TM (1968), p. 410

74. Cf.: "For when an illusion dissipates, when an appearance suddenly breaks up, it is always for the profit of a new appearancewhich takes up again for its own account the ontological function of the first. (...) The dis-illusion is the loss of one evidence onlybecause it is the acquisition ofanother evidence. If, out of prudence, I decide to say that this new evidence is "in itself" doubtful or onlyprobable ( in itself - that is: for me, in a moment, when I will have gotten a little closer to it or looked more closely), the fact remainsthat at the moment I speak it incontestably gives itself as "real" and not as "very possible" or probably; and if subsequently it breaks upin its turn, it will do so only under the pressure of a new "reality"." (Merleau-Ponty, 1968a, p. 40).

Cf. also: "The perceived makes its appearance as a configuration of the world which awaits its confirmation from thesubsequent flux of world apparition, and which contains in germ in itself a sort of intuitive potentiality for its own replacement byanother configuration which would "cross it out" or "cancel" it." (Lingis, 1973, p. 84)

39

2.3.5. Description versus operation of second order consensuality

With respect to the barking of the dog (see Maturana’s example in section 2.3.2, page 33), a humanobserver would be inclined to state that the dog’s barking is a comment to the man’s new behavior, andthat the dog means by it that he should throw the stick again; that it is as if the dog is saying "come,throw it!". Such a qualification takes place in the domain of descriptions of an observer. It consists ofattributions about goal, function and meaning of the interactive behavior of individuals.

Indeed, we could confine our descriptions of the operation of second order consensuality to thecoordinations of coordinations. However, a description in terms of goals, functions or meanings wouldseem to be here more adequate than in respect of first order consensual domains. This is because theparticipants are now themselves observers who make descriptions of objects. Their domain of secondorder consensuality is a domain of descriptions, in which our initial observer, whom we might perhapsbetter indicate now as a meta-observer, may also participate. The meta-observer’s descriptions may alsopertain to the objects which the other (observed) observers together describe«75». These objects maythen become shared by the meta-observer and the observers whom he observes. Such a situation was notpossible in the description of first order consensuality (section 2.2.2), because there the observedindividuals did not (yet) create a domain of descriptions; there were no (not yet) observers.

On the other hand, an observer of first order consensual behavior is very easily tempted to perceive thebehaviors of the individuals involved as tokens, and therefore assign functions, purposes and meanings tothem in a way the interactants themselves do not. Such an assignment would presuppose that the involvedindividuals themselves wouldin principle be able to join the observer’s position and agree with him aboutthe objects (i.c.: the meanings, purposes, etc.) that he is observing«76».

2.4. How will these ideas be used in this book?

Now that I have presented some of Maturana’s ideas on autopoiesis, (second order) consensuality, andobjects, I will finish with some remarks on their position in this book. My line of argument will be toextend Maturana’s ideas about second order distinctions/coordinations and describe how the occurrence ofconsensus, as well as its disruptions, are constitutive for the development of various kinds of discoursebetween people. In particular, I will discuss naive objects and critical objects, and I will describe twomajor ways of using critical objects.

After that, there will be one problem that will draw our attention: how can an observer perceive the closedorganization of a particular autonomous system that he is studying? To that purpose a conceptualframework will be presented to provide some explicit thoughts about what I consider as a particularomission in the theory of autopoietic systems: an affirmation of the impossibility to empirically observethe internal states of an autopoietic system. It will turn out, then, that critical objects, being the principallanguage vehicles for scientific research, are inadequate to make observable the very notions ofautopoiesis and organizational closure. However, this does not mean that we will have to abandon thetheory of autopoietic systems, but rather that it is to be complemented with a delineation of its own limits,and that it is by knowing these limits that we are able to find our way to the ’closure’ of autonomoussystems. We thus attain a formal limit of our linguistic capacities, and it is there that a ’negative way’ tounderstanding psychotherapy will be recognized. I will return to these matters in chapters 7 and 11.

75. As a method of investigation, this kind of description can be found as qualitative descriptions of the ’life world’ of persons. Theinvestigator arrives at an image of these persons’ life world by describing the things that they themselves are describing (theirconsensual everyday life objects). To the extent he succeeds in this, his descriptions are called ’adequate’ (cf. Schutz’ (1932) so called’postulate of adequacy’)

76. Such a presupposition would amount to what William James (1890) called apsychologist’s fallacy! Cf. footnote 112, page 51. Infact, the major difference between the psychologist’s fallacy and Schutz’ (1932) postulate of adequacy (cf. footnote 212 on page 80) isthat the latter is a matter of comparison between two descriptions, whereas the former is a matter of comparison between a first orderdistinction and (the description of) a second order distincton made of it (cf. also section 5.3.1).

3Experiences and naive objects

In this chapter I will introduce some new concepts, which are to be seen as interpretations and extensionsof the concepts already discussed in chapter 2. In the first place, this chapter describes the notion of’experience’ as a proper understanding of ’first order distinction’. In the second place, it is maintainedthat the kind of objects as constructed by autopoietic systems in the course of their languaging, is first ofall of a poetic ornaive type. The ’naivety’ is defined as an absence of a form-content distinction. Thediscourse in which this type occurs, will be contrasted to what is considered the regular scientificdiscourse, which will be calledcritical and which will be described in the next chapter.

3.1. Experiences as first order distinctions, and Merleau-Ponty’s "flesh"

In this section I want to relate the notion of ’distinction’ to Merleau-Ponty’s (1964a)«77»concept’chair’). Merleau-Ponty uses this term to denote the substance, which is itself beyond and prior to thesubject-object distinction, that is the source out of which this distinction evolves. The flesh itself thenremains beyond our scope of the visible. The flesh is the ontological substratum out of which a subject-object distinction may evolve, but which is itself beyond this distinction. Merleau-Ponty (1968a, p. 254)calls it an untouchable, situated at the junction of the touching and the touched. According to Merleau-Ponty, the flesh is to be understood as ’brute Being’, i.e., that out of which both object and subjectcoevolve simultaneously. Hence, it is the basis upon which intentional relations develop, and which, forthat reason, cannot become an object of attention itself. It is in-between the conscious person and theobject that he is aware of, to the extent that the flesh evades becoming itself a subject matter of anintentional act. Hence, it is considered more basic than the subject-object split itself.

We may consider that what Merleau-Ponty calls ’flesh’, as the domain in which experiences exist.Experiences are the mode of functioning by which we, inevitably, participate in the flesh. It is in ourexperiences that we exist before and beyond the subject-object dichotomy and that we are able to havedirect, i.e. immediate contact with others and the world. Our experiences do not constitute a domain offeelings and thoughts, desires and fears, that are all inaccessible, in contrast to a domain in which thesematters are conscious, but rather a domain that gives consciousness its foundation«78».

It is to this conception of experience that we can relate the concept of (first order) distinction: not the’world out there’ versus the ’me here inside’ is at stake, but the immediate contact of (what Merleau-Ponty calls) seer and seen, who both are made from the same stuff, which Merleau-Ponty names ’flesh’.Experiences, then, exist as properties of the interactions that an autonomous individual has with hismedium. They exist as pre-reflexive modes of ’being in the world’«79». No objects are observed orconstructed. I will maintain a strict distinction between the verb ’experience’ on the one hand, and on the

77. I will mostly refer in this book to the English translation of 1968, and accordingly use the term ’flesh’.

78. Kwant (1989, p. 674) refers to Merleau-Ponty’s ideas on the unconscious:"l’Inconscient est le sentir lui-même, puisque le sentir n’est pas la possession intellectuelle de "ce qui" est senti, mais dépossession denous-mêmes à son profit, ouverture à ce que nous n’avons pas besoin de penser pour le reconnaître." (Merleau-Ponty, 1968b, p. 179)["The unconscious is feeling itself, since feeling is not the intellectual possession of ’that which’ is felt, but dispossession from ourselvesfor its sake, an opening upon that which we do not need to think in order to recognize it" (my translation)]

79. We cannot do justice to the wide range of possible experiences individuals can have. According to my use of terms in this text,the whole gamut from the bending of plants towards the light, to the sexual attraction between persons is characterized in terms of ’firstorder distinctions’.

40

41

other hand the verbs ’observe’ and ’perceive’. Only the latter two involve the (coming into) existence ofobjects. ’Experience’ does not involve the perception of an object. It is the act of first order distinction

Though experiences do not specify objects in the environment, an external observer might describe objectswith which he sees the observed individual interact. Then these objects have the status of quasi-object forthe individual that is having the experience. Experiences are not ’internal’ or ’mental’ representations of aquasi-object. However, an external observer may stick to the conviction that the experiences of theobserved individual are the internal or mental correlates of a particular object in the ’real’ world. Theobserved individual is then said ’to experience this particular object’, and the observer ignores thedifference that exists between his own point of view and that of the observed individual«80».

However, experiences do not exist as internal representations of external objects. They exist in a domainwhere the experiencing individual does not distinguish between his own inside and outside. Therefore, the’inside’ in which a person’s experiences are presumed to dwell according to an external observer (e.g. aresearcher or a therapist), does not exist for the observed individual himself, at least as far as hisexperiences are concerned. The same holds for the ’outside’, in which the real world’s objects arepresumed to dwell according to the external observer. Both an internal ’subjective’ representation and anexternal ’objective’ reality exist only in the observer’s domain of descriptions, not in the domain in whichthe observed individual’s experiences exist. To the extent that the observed individual’s experiences areconcerned, both internal and external objects are quasi-objects. A pre-linguistic individual (quasi-observer), therefore, uses only first order distinctions. These are his experiences. An external observermight find some relation between these distinctions and something in the environment in which theindividual is observed to exist. This ’something’ is what I called a ’quasi-object’.

This amounts to saying that in the domain of experiences there is not so much an objective-subjectivedichotomy, as an absence of this dichotomy: that which is experienced (the purported ’objective part’),and the way in which it is experienced (the purported ’subjective part’), are one and the same«81».Only when an individual becomes a language user does this dichotomy come into existence for him. I willreturn to this difference between the domains of experiences and of language in section 7.3.2. whendiscussing the boundary of an autopoietic system in terms of a language barrier. I will also return toMerleau-Ponty’s concept of flesh when discussing the ’via negativa’ (sections 7.4.1 and 7.4.2) and theconcept of ’projective identification’ (section 11.4). For these concepts can be clarified, I think, by whatMerleau-Ponty wrote about the flesh. In particular, the difficulties of a language user to access the domainof first order distinctions will recur as issues of our highest interest.

80. Elsewhere (e.g. Goudsmit, 1989a) I have described this divergence of perspective in terms of what William James (1890) calledthe ’psychologist’s fallacy’ (cf. footnote 112, page 51).

81. This is different from saying, as does for example Ciompi (1982), that experiences and cognitions are always going together,inseparably, though they are not identical. Such a conceptualization is in fact along the same lines as the classical AI approach,according to which both cognitions and emotions are ways of information processing (cf. Abelson’s (1963) definition of emotions as’hot cognitions’ (as discussed by Stegge & Meerum Terwogt, 1993, p. 237)).

Conversely, according to the position I am taking here, experiences (feelings) are cognitions at this pre-linguistic level offunctioning. For that reason, ’that what is experienced’ is not distinguished from ’the way in which it is experienced’ (cf. the list ofpairs given in section 2.2.3). This position is congruent with Maturana’s, cf. "Cognition as a biological phenomenon takes place in aliving system as it operates in its domain of perturbations, and as such it has no content and is not about anything" (Maturana, 1987a,p. 361).

We may also recognize this absence of distinction in a quotation from Heidegger about feelings: "What is phenomenologicallydecisive in the phenomenon of feeling is that it directly uncovers and makes accessible what is felt.... Feeling is having a feeling-for,and so much so that in it the feeling ego at the same time feels its own self" (M. Heidegger, 1982/quoted by Denzin, 1984).

42

3.2. Naive objects and the coordination of disagreement

In this section I will attempt to use Maturana’s notions of ’second order coordination/distinctions/consensuality’ in order to describe the emergence of objects in a social domain between individuals. I willmaintain that the objects that occur through second order distinctions are perceived in a way that I willcall ’naive’, i.e. no distinction is made between that which is perceived and the way in which it isperceived. Furthermore, in the next section I will add to this idea that an observer of these naive objectsdoes not distinguish between logical types, such as elements and classes.

Experiences can be shared with other individuals. This is a matter of first order coordination of actions.Remaining conflicts arise there where these coordinations fail, as in physical collisions. Sometimes theseremaining problems can be dealt with by a second order coordination of actions, as I discussed in section3.2.

During structural coupling between living systems, the interactants behave in a way that becomescoordinated, due to the development of their structural isomorphism. The coordination of these actions is aspontaneous outcome, not one that is intentionally sought for. It is a mode of growing together.

These coordinations of actions constitute distinctions that are consensual to the two interactants«82»,and as such they are shared experiences. Since they are generated by a structural coupling between thetwo participants, the shared experiences take place as a shared activity. Here the inter-individualexperience is at the same time intra-individual!«83»

As said already in the previous section, no objects are distinguished by first order consensual distinctions.An external observer, however, might find a correspondence between the interactive patterns and someobject as he observes it in the environment of the two interacting individuals. The question is,can weunderstand the emergence of objects in terms of experiences and interpersonal interactions?I will try togive a conceptual frame within which Maturana’s ideas on objects in terms of coordinations of actions, donot seem as far-fetched as they may have done at first sight.

The emergence of objects is comparable to what happens when two predators have caught the same prey.Then, after some fighting, they may decide to divide the food, so that both may start eating of it, insteadof continuing their fight. This, indeed, is a case of first order coordination of actions.

Likewise, a situation may arise in which a personcannotshare his experiences with another person in animmediate and self-evident way. Now it is not a prey to be shared, but a difference of experiencesbetween two persons. The way in which the experiences of two actors do no longer fit together,constitutes the issue to be coordinated. This may result either in the first order coordination of actionsbeing restored, or in the construction of a stable intermediary between the diverging experiences of thetwo actors. As a result of this situation, objects are perceived. Here the divergence between the persons’non-identical experiences is itself coordinated, as if two flat images are integrated into a stereoscopicimage. The actors contribute to the intermediary in different ways. They no longer have, as in first orderconsensuality, a shared movement, but rather a shared outcome of non-shared movements. This is howMaturana’s idea of coordination of coordination of actions can be related to the origination of objects in aconsensual domain.

Second order consensual distinctions may occur as soon as the individuals no longer coordinate theirsimple actions (i.e. share their experiences) in a self-evident way. The consensual management of theoccurring divergences of experiences is a matter of second order coordination (distinction) of actions. Aslong as experiences are shared unproblematically, no language and no objects arise. Likewise, when in the

82. see e.g. Plooij (1979) for a field study of mother-child interactions in chimpanzees.

83. Cf. Baranes (1991) who remarks that it is the encounter of two persons in one mental space that is required for a therapeuticprocess to be possible.

43

course of a second order coordination of actions the first order coordination of actions no longer needs tobe coordinated (that is: when the interactants (again) start to share their experiences smoothly andunproblematically), then language disappears.

However, when the (first order) coordination of actions (i.e. the sharing of experiences) is no longertaking place smoothly, then they may become a quasi-object for the interactants. A certain discoordinationhas arisen.

What I call here a discoordination, in fact does not exist as a perceivable object for the interactants. It is aquasi-object for them, in that they may beobserved(by an external observer) to deal with it, withoutthemselves observing it. The shared way of dealing with the discoordination that has arisen is whatMaturana calls a second order coordination. It is through this that the ’interference pattern’«84»becomes a token for an object.

We can understand this as the occurrence of a ritual, esp. one that puts a variety of diverse experiences ofits actors into one higher level pattern. Thus, a ritual can be seen as an agreement at the level ofbehavioral patterns, where otherwise a collision of physical forces might have taken place«85». Abeautiful illustration of the origination of a ritual can be found in Fassbinder’s film ’Querelle’, where anarmed confrontation between Querelle and his brother smoothly turns into a pattern of ritualizedmovements, as if it had become a dance. It is through this ritual that an ’about’ arises, i.e. somethingaboutwhich the ritual is a de facto agreement.

In a different terminology a similar idea can be found in William James’ (1912) essays on radicalempiricism. Here James develops the notion of an object that is shared by two actors, as a kind ofcontinuation of a shared interface. Though his terminology is different (e.g.: ’contents of mind’), heclearly stipulates the identity of the locus of touch between two individuals (’you and me’), as well as thereality of a shared object: if our hands can touch and find each other real, then they may also share aworld that is in between them (e.g. a rope that they both hold)«86».

At a more complex level, the consensual construction of objects is apparent in the phenomenon ofspeaking turns between conversation partners. Here too, the contributions are not identical, the words andmovements are not shared, as in first order distinction. In speaking each contributes to the delineation of ashared subject matter, but each participant has its own way of doing so. It is this very differentiation ofperformance that is absent in first order consensuality.

The distinction between an observation process and the object observed can be present as well as absentin the performances of an observer. If it is absent, then the observer’s act of perception isnotaccompanied by any awareness of itself. The kind of object perceived thus has as a defining characteristic

84. cf. section 2.3.2, see footnote 57 (page 33).

85. (cf. Volkan, 1988, p. 103): "... rituals to maintain these minor differences keep a psychological gulf between the opposinggroups that absorbs the flow of aggression and, at least in time of peace, keeps opposing groups from killing one another."

cf. also: "One of the manifestations of playfulness is the eeryday use of ritual, which, according to Erikson, "in a viable culturalsetting, represents acreative formalizationwhich helps to avoid bothimpulsiveexcess andcompulsiveself-restriction, both socialanomie and moralistic coercion" [Volkan refers here to Erikson, 1978, p. 82] (Volkan, 1988, p. 112)

86. Cf.: "That body of yours which you actuate and feel from within must be in the same spot as the body of yours which I see ortouch from without. ’There’ for me means where I place my finger. If you do not feel my finger’s contact to be ’there’ inmy sense,when I place it on your body, where then do you feel it? Your inner actuations of your body meet my fingerthere: it is there that youresist its push, or shrink back, or sweep the finger aside with your hand. (...)

In general terms, then, whatever differing contents our minds may eventually fill a place with, the place itself is a numericallyidentical content of the two minds, a piece of common property in which, through which, and over which they join." (1912, p. 84-5)and:

"... your hand lays hold of one end of a rope and my hand lays hold of the other end. We pull against each other. Can our twohands be mutual objects in this experience, and the rope not be mutual also? What is true of the rope is true of any other percept. Yourobjects are over and over again the same as mine" (1912, pp. 78-9)

44

that it is not distinguished by the perceiving person from the way in which it is perceived! It does notexist as an external objective thing, perceived by a subject ’inside’. This mode of perception will be called’naive’. The naive perceiver does not distinguish between what is perceived and how it is perceived.As anaive perception is performed, the naive perceiver does not know that he does not make this distinction;rather, he embodies its absence. Thus, naive perception is typical of a huge class of sensitiveperformances by human as well as non-human beings. Naive perception is the outcome of second ordercoordination of actions.

Indeed, this naive mode of perception is prior to a mode in which the distinction between object andsubject is being made! This latter mode of perception can be called ’critical’, denoting that the perceiverdoes distinguish between logical types, as well as between the ’what’ and the ’how’ of his observations.Given this notion of naive perception, the question where and how object and subject should meet, or,similarly, how transitions between the domains of the physical and the symbolic are possible, can bereframed as a search for a point of common origination of subject and object. Paradigmatic for the searchfor such a common point of origin is Merleau-Ponty’s concept of the ’flesh’ (see section 3.1).

Naive perception originates in and develops out of this junction of touching and touched, which in fact isa coincidence of an act of distinction and that which is distinguished. There is not a subject perceivingand an object that is distinct of it. Rather, in naive perception there is the act of coordination of divergingexperiences, which become tokens for the object perceived. The object thus constructed, exists due to andthroughout the coordination acts of the observer.

Naive objects can be found in the performance of children ("car has no ears!"«87», "handkerchief issweet"«88»), poets, and others who, unlike regular scientists, do not need to detach themselves fromtheir subjects of attention. It also happens in the perception of art. An example of the coincidence ofobject and process in naive perception can be found in little preschool children’s appreciation of thingsthey like. For instance, it is not merely tea with milk and sugar that they want, but it is their ownperformance of pouring the milk and the sugar into the cup that is essential, not a dispensable extra. Theact of preparation and its outcome are not distinguished.

Similarly, Van den Berg (1956, p. 224) discusses Jean Cocteau’s attempts to pick up the traces ofchildhood memories, while walking the way from home to school. He then discovers that these traces aresituated in the walls of the houses, which as a child he touched with his hand while tracking along them.Thus, Van den Berg concludes, the memory traces are in the walls, at the level of the school child’shands, not at the level of the adult’s hands. Van den Berg, for the sake of his own argument, thenconcludes that memory traces do not exist in a person’s head. This may seem too far an exaggeration; thepoint, however, is that Van den Berg likes to stipulate a non-psychological, action oriented and (what Icall) naive perception of memories.

An important elaboration of (what I call) naive perception can be found in the writings of Gadamer.According to Gadamer (1960/1986), it is typical of the observation of play (a term used for the games ofchildren, but equally for the theatre performance of a drama) that the observer is participating in the play,in that he contributes to its realization. In observing the play, the observer forgets the distance betweenhimself and the play. This is particularly clear for the case of the theatre visitor, when he becomesenchanted by the play, so that its ’hermeneutic identity’«89», as Gadamer (1986b, p. 25)«90»calls it, becomes apparent. This loss of distance between the observer and the observed play is a case ofnon-distinction between content and form. It implies a non-distinction between the observed object and theway in which it is perceived.

87. Ben at 2 years 9 months. Cf. also Arieti, 1976, p. 208.

88. Eva at 2 years 4 months

89. cf. footnote 431 on page 146

90. (1977, p. 33)

45

Similarly, it is in the performance of a work of art (say music) that the performer also loses sight of thedistinction between the music as written down by the composer, and the way it is performed. (Notice thathere ’performance’ corresponds to ’form’, and ’work of art’ to ’content’.) The performer, therefore, alsobecomes a naive perceiver of what he is performing! Gadamer describes the performance as a processduring which the work itself becomes indistinguishable from the way in which it is performed (a processdubbed "ästhetische Nicht-unterscheidung", 1960, pp. 111/112«91»). This aesthetic non-distinctionholds both for the performer and for the observer (audience) who participates in the performance byattending it«92».

Gadamer writes:"If the identity of the work [of art] is [being addressed], then the genuine reception and experiencesof a work of art can exist only for one who "plays along," that is, one who performs in an activeway himself. Now how does that actually happen? Certainly not simply through retention ofsomething in memory. In that case there would still be identification, but without that particularassent by virtue of which the work means something to us.What gives the work its identity aswork? What makes this what we call a hermeneutic identity? Obviously, this further formulationmeans that its identity consists precisely in there being something to "understand," hat it asks to beunderstood in what it "says" or "intends." The work issues a challenge which expects to be met. itrequires an answer - an answer that can only be given by someone who accepted the challenge.And that answer must be his own, and given actively. The participant belongs to the play." (1986b,p. 25-6)«93»

Notice that also Merleau-Ponty (1945, p. 177) discusses works of art in terms of the impossibility todistinguish the expression from the expressed. Gadamer contrasts the position of aesthetic non-distinctionto the position of the critic of art, who has to analyze the performance itself, in contrast to the performedpiece«94».

Naive perception can also happen in watching a movie. A good example can be found in the film’Carmen’ (1983) by Carlos Saura (cf. Hall, 1995). This film, full of flamenco dances and passionateencounters, enchants the spectator as he forgets himself and becomes susceptible for the passions enactedand danced by the actors,without being able to distinguish whether his passions are stirred by the form orby the content; i.e., by the physical properties of the film, such as the physical appearance of the actors;

91. ’aesthetic non-differentiation’ (Gadamer, (TM) 1986a, p. 105); I prefer the term ’aesthetic non-distinction’

92. I will return to this concept of aesthetic non-distinction in section 9.4.

93. "Wenn das [’Angesprochen werden’] die Identität des [Kunstw]erkes ist, so hat es immer nur für den ein wirkliches Aufnehmen,eine wirkliche Erfahrung eines Kunstwerks gegeben, der "mitspielt", d.h. der eine eigene Leistung aufbringt, indem er tätig ist. Wodurchkommt das eigentlich zustande? Doch nicht durch bloßes Festhalten von etwas im Gedächtnis. Auch dann ist Identifikation gegeben,aber nicht die besondere Zustimmung, durch die das "Werk" etwas für uns bedeutet. Was ist es, wodurch ein "Werk" als Werk seineIdentität hat? Was macht seine Identität, wie wir auch sagen können, zu einer hermeneutischen? Diese andere Formulierung meintoffenkundig, daß seine Identität eben darin besteht, dass etwas daran "zu verstehen" ist, dass es als das, was es "meint" oder "sagt",verstanden werden will. Das ist eine von dem Werk" ergehende Forderung, die auf ihre Einlösung wartet. Sie verlangt eine Antwort, dienur von dem gegeben werden kann, der die Forderung annahm. Und diese Antwort muß seine eigene Antwort sein, die er selber tätigerbringt. Der Mitspieler gehört zum Spiel." (1977, p. 34)

94. "... it is a secondary procedure if we abstract from whatever meaningfully addresses us in the work of art and wholly restrictourselves to a "purely aesthetic" evaluation.

That would be like a critic at the theater who exclusively took issue with the way the production was directed, the quality ofthe individual performances, and so on. Of course, it is quite right that he should do so - but the work itself and the meaning it acquiredfrom us in the actual performance does not come to light in this way. The artistic experience is constituted precisely by the fact that wedo not distinguish between the particular way the work is realized and the identity of the work itself." (1986b, p. 29)["... daß es eine sekundäre Verhaltensweise ist, wenn man von dem abstrahieren sollte, was einen durch ein Künstlerisches Gebildebedeutsam anspricht, und man sich gänzlich darauf beschränken wollte, es "rein ästhetisch" zu würdigen.

Das wäre so, wie wenn der Kritiker einer Theateraufführung sich ausschließlich mit dem Wie der Regieleistung, mit derQualität der einzelnen Rollenbesetzungen und dergleichen auseinandersetzte. Es ist ganz gut und richtig, daß er das tut - aber das istnicht die Weise, wie das Werk selber und die Bedeutung, die es für einen in der Aufführung gewann, sichtbar werden. Gerade dieNichtunterscheidung zwischen der besonderen Art, wie ein Werk zur Reproduktion gebracht wird, und der Identität des Werkes dahintermacht die künstlerische Erfahrung aus." (1977, p. 38)]

46

or whether the spectator becomes excited by the narrative contents of the Carmen story, originallyconceived by Mérimée in 1845. On the other hand, a non-involved, critical observer, such as Gadamer’scritic of art, will be able to distinguish between the form and the content, that is, between features of theperformance, on the one hand, and on the other hand the original story as intended by its author. But as,in naive perception, the way the play is received mixes up with the object perceived, the enchantment andexcitement of the spectator is indistinguishable (that is: for the spectator!) from what he sees.Accordingly, the performance is no longer distinguishable from the play itself.

Ter Hark (1990, p. 215), discusses Wittgenstein’s philosophy of psychology. Wittgenstein presents aconceptual differentiation according to which sensations give us information about the external world,whereas emotions, unlike sensations, are directed at an object. Thus understood, the terms ’sensation’ and’emotion’ parallel the presently used terms ’first order distinction’ and ’second order distinction’, theformer being the individual experiences and the latter being the ways an individual deals with them as anaive (poetic) object. Cf. also Wittgenstein’s idea that an emotion, unlike a sensation, does not have abodily locus, but rather, is situatedin its object (Ter Hark, 1990, p. 217). Finally, notice that the notion ofa common reference between individuals seems to be missing in Wittgenstein.

3.3. Naive objects: Vico and poetic language

In this section I will argue that naive objects are, as Vico called them, poetic objects. They do notdistinguish between logical types, such as elements and classes. A naive observer does not reflect aboutthe object he is perceiving; he does not perceive it as distinct from his own modes of observing it. Anobject perceived in such a non-reflexive way is not perceived as the member of a class, a substance that isthe carrier of certain accidents (this will be the case for critical objects, as I will elaborate in chapter 4).

What Maturana constructed with respect to the occurrence of consensual constructions of reality as aninteraction phenomenon between individuals, Giambattista Vico«95»constructed with respect to theorigination of language and thought in the ancient times.

Vico presents an argument that is concerned with the origination of objects. The primal situation is one inwhich primitive people are imagined to be confronted with a thunderstorm:

"... as everyone runs to shelter from the thunder, all in a state of fear, an opportunity exists forthem to realize that it is thesame thingthat they all fear; and a look or a gesture will communicatethis. What we might call a ’moment of common reference’ exists between them." (Shotter, 1991,p. 386)

But this is not a matter of inference, from sensory information to an underlying entity, for"[i]t was by a topos, or common "place" (startling sounds are likely in the first "place" to beutterances of some being or other), that primitive men passed from the thundering sky to JupiterTonans, Jove the Thunderer; but this was a "sensory"topos in that the passage had the immediacyof feeling rather than the discursiveness of deliberate inference. To be hearing thunderwas to belistening to Jove. Only by a much later criticism could observed fact and invented explanation bedistinguished, and only then could thetopos itself be identified and formulated, and rules for its usedrawn up." (Bergin & Fisch, 1986, p. xl)

This ’much later criticism’ is one that has been formulated by Gadamer, among others, as the work of acritic of art, who does not have the aesthetic non-distinction«96». A critical«97»scientist (likethe critic of art) would be able to distinguish the fact of thunder from its explanation, but not so primitiveman. For Vico is describing human performance here at the level of ’common sense’, which he defines as

95. esp. the New Science (Vico, 1744/1986).

96. cf. footnote 94, page 45

97. I will discuss my usage of the term ’critical’ in chapter 4.

47

’judgment without reflection, shared by an entire class, an entire people, an entire nation, or the entirehuman race" (1986, p. 63/1744, §141)

The ’sensory topos’, that which according to Vico, was shared between these primitives, shouldnot, Ithink, be understood as a sensory experience common between the persons involved«98», inconjunction with its explanation, but rather, conversely, as the occurrence of individual experiences inseveral persons, that found the explanation ’Jove’ as a focal point in common. It was this latter that cameto function as the anchor point for the coordination of subsequent experiences, as well as of theexperiences of other persons«99».

It was Vico’s discovery that the language that is arising thus is primarily of apoetickind. That thesepeople spoke poetically was what Vico considered as his major discovery«100».

The kind of poetic objects that were available to these people, is discussed by Verene in terms of’imaginative universals’. They are contrasted with the so called ’intelligible universals’, which we wouldcall abstract concepts:

"Imaginative universals are the form of thought for the first two of Vico’s three ages - the age ofgods, the age in which men ordered their world in terms of gods, and the age of heroes, the age inwhich men ordered their world in terms of heroic figures. The imaginative universals of the firstage are formations of the immediately experienced forces of nature, such as Jove’s presence as thethunderous sky, or transformations of physiological bases of human existence, such as Juno, theform of everything relating to marriage, the social ordering of sexual actions. The heroes of thesecond age give form to human qualities. Thus Achilles represents "all the deeds of valiantfighters" and Ulysses embodies "all the devices of clever men."

The intelligible universal, the form of thought characteristic of Vico’s third age, the age ofmen, is that of the generic concept of Aristotelian logic. In this age men think thoughts that haveno figurative presence about them." (Verene, 1981, pp. 71-2)

Thus, the kind of object obtained by primitive (first and second age) man, according to Vico, is of ametaphoric quality.

"Vico’s discovery of the imaginative universal is the discovery of a principle of identity that is

98. as in Shotter’s reading: "these first anchor points are to do, not with ’seeing’ in common, but with ’feeling’ in common, with the’giving’ or ’lending’ of a sharedsignificanceto sharedfeelingsin an already sharedcircumstance." (Shotter, 1991, p. 388; cf. alsoShotter, 1987b, p. 5)

99. Shotter relates the occurrence of sensory topics to the recognition of feelings in others: "... a sensuous totality linking thunderwith the shared fears at the limits of one’s being, and with recognition of the existence of similar feelings in others because of sharedbodily activities." (Shotter, 1991, pp. 386-7)

Also in: "And here it is the fear shared in common which provides the first fixed reference point which people can ’find again’within themselves and know that others ’feel the same way’. (Shotter, 1991, p. 386)

Also Verene mentions the occurrence of ’same feelings’, in:"The thunderous motion of the sky that covers the world like a ceiling sweeps over the heads of thegiganti whose bodies involuntarilytremble in imitation of the celestial motion. The motion instilled in their bodies becomes the quasi-mental motion of the gesture, thehalf-voluntary movements that fix their sensations on a difference between the conditions of the earth and of the sky. The ability to fearand fear again in the same way produces a common point of focus, an object, and a common form of feeling among the giants. Theyfeel the same thing in the same way." (Verene, 1981, p. 91)

Expressions such as ’to fear in the same way’, ’a common form of feeling’, and ’the recognition of similar feelings’ have thedisadvantage of suggesting that the similarity between the various feelings of the persons involved logically preceded the origination ofthe focal object. Instead, I critically venture to submit, Vico can also be read as proposing that it is the construction of the poetic object,the focal point, that permits the sensations to be coordinated! This is similar to the idea that these sensations, though diverging andvarious, are ’tokens’ (in Maturana’s sense), in a process of second order coordination. The agreement between thegiganti or otherprimitives, I think, should be understood as a consensus on a shared focus, not as a consensus on sensations (in the sense of section2.2).

100. "We find that the principle of [the] origins both of languages and of letters lies in the fact that the early gentile people, by ademonstrated necessity of nature, were poets who spoke in poetic characters. This discovery, which is the master key of this Science,has cost us the persistent research of almost all our literary life, because with our civilized natures we [moderns] cannot at all imagineand can understand only by great toil the poetic nature of these first men." (Vico, 1986, p. 21-2/1774, §34)

And: "... poetry is shown to have been the first common language of all ancient nations..." (Vico, 1982, p. 140/1725, §253)

48

linked with the notion of metaphor as the fundamental epistemological element. The metaphor isthat by which identity is originally achieved in perception.It is the form perception mostimmediately takes. Metaphor is the first (...) of the operations of mind in the act of knowing."(Verene, 1981, pp. 79-80, italics added)«101»

Hence, these objects do not distinguish between elements and classes:"For the poetic mind poetic characters are not colorful or "rhetorical" embellishments of generaabstracted from empirical particulars; the particulars are not thought ofanalogically, as beinglikethe ideal portraits of the generic individual of the fable. The individual cannot be thought of ashaving a reality apart from the generic character; he is what he is only through his identity with it."(Verene, 1981, p. 75)

and:"Meaning can be achieved only if a sensation can become a particular that is not cancelled by thepresence of the next sensation. The mind can have something before it through its power toproduce an identity. A sensation is apprehended as the being of the other sensations in the motionof the flux. The time of the flux is cancelled by a locus within itself. A single sensation becomes apermanent reference point of the flux of sensations which now have their being in it. Through thisfixed point of sensation, or particular, the meaning of the flux can be repeatedly grasped. (...) Thefixing of sensation in the flux does not occur by analogy, by likening one sensation to another, byfinding similarities in this flux. The fixation of the particular is not a process in which onesensation in the flux is graphed aslike the rest. It is a process in which theis itself is made."(Verene, 1981, p. 82)

The achievement of a ’fixation of the particular’ is the achievement of using experiences as tokens forsome object, and of coordinating them accordingly. I will return to the idea of a ’selected fact’ whendiscussing Bion’s work in chapter 11.

In the perception of poetic objects, there is no distance between the actor and his acts, no reflection onwhat he is doing or how. The naive object, as a metaphor, arises spontaneously, in a way not designed orforeseen«102». The objects perceived are there, due to a bodily involvement of the actor, one that isso vital for perception that neither any distance from the act of perception is possible, nor a differentiation

101. see also footnote 102, as well as section 4.2

102. In section 4.2 I will use the term ’new naive object’, in order to emphasize this spontaneous emergent quality. In a similarvein, Merleau-Ponty describes philosophy’s invention of new ’openings upon Being’:

"... if philosophy can speak, it is because language is not only the depository of fixed and acquired significations, because itscumulative power itself results from a power of anticipation or of prepossession, because one speaks not only of what one knows, so asto set out a display of it - but also of what one does not know, in order to know it - and because language in forming itself expresses, atleast laterally, an ontogenesis of which it is a part. But from this it follows that the words most charged with philosophy are notnecessarily those that contain what they say, but rather those that most energetically open upon Being, because they more closelyconvey the life of the whole and make our habitual evidences vibrate until they disjoint. Hence it is a question whether philosophy asreconquest of brute or wild being can be accomplished by the resources of the eloquent language, or whether it would not be necessaryfor philosophy to use language in a way that takes from it its power of immediate or direct signification in order to equal it with what itwishes all the same to say." (Merleau-Ponty, 1968a, pp. 102/3)["... si la philosophie peut parler, c’est parce que le language n’est pas seulement le conservatoire des significations fixées et acquises,parce que son pouvoir cumulatif résulte lui-même d’un pouvoir d’anticipation ou de prépossession, parce qu’on ne parle pas seulementde ce qu’on sait, comme pour en faire étalage, - mais aussi de ce qu’on ne sait pas, pour le savoir -, et que le language se faisantexprime, au moins latéralement, une ontogénèse dont il fait partie. Mais il résulte de là que les paroles les plus chargées de philosophiene sont pas nécessairement celles qui enferment ce qu’elles disent, ce sont plutôt celles qui ouvrent le plus énergiquement sur l’Être,parce qu’elles rendent plus étroitement la vie du tout et font vibrer jusqu’à les disjoindre nos évidences habituelles. C’est donc unequestion de savoir si la philosophie comme reconquête de l’être brut ou sauvage paut s’accomplir par les moyens du langage eloquent,ou s’il ne lui faudrait pas en faire un usage qui lui ôte sa puissance de signification immédiate ou directe pour l’égaler à ce qu’elle veuttout de même dire. (1964, p. 139)]

It is precisely this usage of language in a new way, such as ’to express an ontogenesis of which it is a part’, that is required inpsychotherapeutic conversations. It requires a usage of language that, in order to find ’new openings upon being’ is beyond, or ratherbeneath, the conventional representational mode. I will return to this below in section 7.3.3. Cf. also the fragment from IJsseling (1990)on Heidegger, in footnote 65 on page 36.

49

as to universals and particulars. The properties of the naive object are necessary and intrinsic to it, and arenot perceived as accidents that are contingent and assignable. The objects thus perceived do not exist asthe members of a class, but rather do they appearboth as a class, or genus, and as elements of it.Nological types are distinguished or recognized«103».

The poetic objects thus perceived are naive objects. In regular conversations they usually arise asmetaphors«104». The term ’naive’ denotes the absence of a distinction between the object’s form onthe one hand, and on the other hand its content«105». Accordingly, I will call the observer of anaive object anaive observer, and the discourse in which the naive objects occur anaive discourse.

But why, one may ask, do these poetic or naive objects come into existence at all? Why should personsconstruct those ’imaginative universals’, why should a sensation be ’apprehended as the being of the othersensations in the motion of the flux’, so that ’a single sensation becomes a permanent reference point ofthe flux of sensations’?

The answer, I think, is to be formulated in terms of their usefulness and functionality. It is in order tosurvive in a medium (in Maturana’s sense) of experiences of other actors, experiences that may divergeslightly or strongly. Poetic or naive objects are a way to integrate these differences into a coherent entity,and, accordingly, to simplify the world«106». This is what the Jupiter Tonans denoted according toVico: a focal object, for which their experiences could be used as converging tokens. It was also whatWilliam James’ meant in his examples of the rope and the touch between two persons«107».

I will now address one major issue concerning naive objects, viz. that, once thinking in a criticaldiscourse, it is very difficult to speak about objects and perceive them without making the distinctionbetween logical types (class-element, form-content, etc.) that is so typical for critical thinking. This is oneof the problems of empirical research into psychotherapeutic processes: the objects of psychotherapeuticconversations are very often lacking the form-content differentiation, and their translation into criticalterms may be impossible. I will return upon these matters in chapters 7 and 9.

103. "The poetic mind can univocally predicate a poetic genus of a collection of elements in a way that makes no sense to the mindthat thinks in terms of intelligible class concepts." (Verene, 1981, p. 76)

and:"For the mind in possession of intelligible universals and abstract words, in possession of the fundamental distinction between thing andproperty upon which Aristotelian class logic rests, such forms of predication by the mythic mind appear as either nonsensical thought orunnecessary literary embellishments of what are in reality properties common to a class of objects." (Verene, 1981, p. 77)

One is reminded of this lack of distinction between class and element, when reading Geertz’ (1975) description of the Balineseconception of ’person’:

"To apply (...) designations or titles (or, as is more common, several at once) to a person is to define him as a determinate pointin a fixed pattern, as the temporary occupant of a particular, quite untemporary, cultural locus. To identify someone, yourself or anyoneelse, in Bali is thus to locate him within the familiar cast of characters - "king," "grandmother," "third-born," "Brahman" - of which thesocial drama is (...) inevitably composed. (...)

The immediate point is that, in both their structure and their mode of operation, the terminological systems conduce to a viewof the human person as an appropriate representative of a generic type, not a unique creature with a private fate." (p. 50)

104. Thus, a psychotherapist may find himself inventing a particular metaphor upon hearing a report from his patient, e.g.: "[A]client has been saying that she needs to be independent from her husband, but whenever he comes home, she drops everything andfollows him around. The therapist responds, "And there goes the little puppy dog pattering after him."" (Rice, 1974, p. 310). Then sucha metaphor is full of the therapist’s personal involvement with and his emotions concerning the topic of the conversation.

Cf. also: "The patient’s statement took on new meaning, but I think it would be more acurate to say that the (remembered)statement was now a new statement for me, and in this sense was being made for the first time." (Ogden, 1994a, p. 10);

and: "The act of re-presenting one’s experience in the form of a dream constitutes the creation of a new experience, a newintegration which is immediately undergoing deintegration (as reflected in the experience of the dream as a fading, ephemeral, barelyknowable psychic event). (Ogden, 1992b, p. 616-7)

105. In section 4.3 I will go into some more detail about the differentiation between (abstract) form and (concrete) content.

106. cf. Pattee’s (e.g. 1972b, 1995) ideas of simplification as a basic instrument of survival in living beings.

107. cf. footnote 86 on page 43.

50

3.4. The absence of the naive object’s form

As has been said already, the term ’naive’ is chosen in regard of the fact that the kind of observation thatis involved does not allow abstractions, i.e. disengagements from the personal and bodily involvement ofthe observer. That which is observed, is observed without a reflective detachment. This usage of the term’naive’ implies that no formal properties of the naive object can be observed by the naive observer, noranything else that one might consider its form. If a meta-observer would specify a particular abstractquality of an object, then for a naive observer this abstract quality would not exist.

For instance, until about the age of 3, my children had difficulties explaining the locus of their pain, afterhaving hurt themselves by falling on or against something (e.g. the edge of a table). When asked ’wheredoes it hurt?’, they almost invariably showed me the object inflicting the pain, indicating away from theirown bodies, toward this object (e.g. table edge). Instead, the parent was generally more interested in theaching spots on the body, butthat required from the child some more capacity of abstraction than it hadavailable at those moments; for that would have required from the child (and may even have taught themin the end!) a capacity to distinguish its own body from the hurting object, instead of perceiving thisobject through the aching body.

Naive perception, as an act, implies theabsenceof a distinction between logical types, and, hence, theobject perceived can be described as acoincidenceof logical types. The point is, that the absence ofdistinction between logical types is typical of naive perception. This absence of distinction can bedescribed by us only in terms of a coincidence (e.g. the non-distinction between expression and expressed,between dancer and dance«108», between performance and performed work of art). But it is inorder to be able to say all this, thatwe must distinguish them first, and then declare themindistinguishable! This is only possible by describing indistinguishables as coinciding parts, i.e., by saying"’this’ and ’that’ are the same, do not differ, though they have been defined previously as distinct; bynow, we declare their distinction to be nonexistent." Form and content of a naive object are quasi-objectsfor the naive observer.

It is the impossibilitynot to make a distinction between terms after having introduced their identity, thatis also of interest to Heidegger in his essay on the ’Satz der Identität’:

"Since the era of speculative Idealism, it is no longer possible for thinking to represent the unity ofidentity as mere sameness, and to disregard the mediations that prevails in unity. Wherever this isdone, identity is represented only in an abstract manner." (Heidegger, 1969, p. 25)«109»

For example, when taking a child to a fair and having it enjoy it, we should not expect the child to havean eye for the social misery of some of the shopkeepers at the fair. That would require from the child acapacity to abstract from the pleasurable world in which it is moving around«110».

A naive object cannot be said to have a form«111». If one nevertheless wishes to speak of it, thisnecessarily means that one is speaking of something that is a quasi-object for the naive observer. Only a

108. Cf. footnote 504, page 170.

109. "Seit der Epoche des spekulativen Idealismus bleibt es dem Denken untersagt, die Einheit der Identität als das blosse Einerleivorzustellen und von der in der Einheit waltenden Vermittelung abzusehen. Wo solches geschieht, wird die Identität nur abstraktvorgestellt." (Heidegger, 1990, p. 12)

110. I once bought an audio cassette of Dutch traditional children’s songs. The piano accompaniment on it appeared to beexcessively ornate and solistic. It is here that, as a critical observer, one starts to wonder whether the piano player really likes hismodest role of musically supporting the children, or whether he would perhaps have preferred a solo career. In short, as a criticalobserver one loses one’s naive involvement with the music.

111. which may be: its meaning or its essence, cf. section 4.3

51

meta-observer may perceive it and believe«112» it is ’implicitly’ present in the naive observer’sperception. The naive observer may of course be said to ’perceive’ this form, but this can only be a quasi-perception, comparable to how a non-linguistic quasi-observer could be assigned the ’experience’ ofparticular objects (as in 3.1).

The form of a critical object is very similar to what Vico called the intelligible universals«113»:these are the abstract concepts or entities that are distinguished from the particulars that are related tothem.

Gadamer gives a clear example«114»of a naive perception (which he calls ’actual experience’), incontrast to a more critical one:

"It is to move out of the actual experience of a piece of literature if one investigates the origin ofthe plot on which it is based, and equally it is to move out of the actual experience of the drama ifthe spectator reflects about the conception behind a performance or about the proficiency of theactors. This kind of reflection already contains the aesthetic differentiation of the work itself fromits representation«115». But for the meaningfulness of the experience as such it is, as wehave seen, not even important whether the tragic or comic scene which is played before one takesplace on the stage or in life - if one is only a spectator. What we have called a structure is oneinsofar as it presents itself as a meaningful whole. It does not exist in itself, nor is it experienced ina communication accidental to it, but it gains, through being communicated, its proper being."(Gadamer, 1986a (TM), pp. 105/6)«116»

I will return to the issue of regaining one’s observational naivety in chapter 9. Clearly, according toGadamer this naivety has something to do with the non-distinction between what a thing is and how itcomes to the observer.

To continue the example, the adult who accompanies the child on the fair might thus attribute to the childsome ’implicit’ perception of the social misery, maintaining for example: ’if the child would have hadsome attention for what I tried to show him, he would have been able to recognize what I meant’. Andthus the adult may continue to believe that the fair which he himself has visited, was the same one as the

112. I have discussed such a belief as a "psychologist’s fallacy" (cf. Goudsmit, 1989a). As above, it is again such a meta-observerwho commits a ’psychologist’s fallacy’ and erroneously takes for granted that the naive observer is able to take his own perspective andsee the object in the way the meta-observer sees it (I will introduce the term ’critical perception’ to this purpose, in chapter 4. Thefollowing two quotations from William James’ magnum opus (1890) describe the idea of the "psychologist’s fallacy":

"The great snare of the psychologist is theconfusion of his own standpoint with that of the mental factabout which he ismaking his report. I shall hereafter call this the ’psychologist’s fallacy’par excellence." (vol I, p. 196)

and: "Many philosophers (...) hold that the reflective consciousness of the self is essential to the cognitive function of thought.They hold that a thought, in order to know a thing at all, must expressly distinguish between the thing and its own self. This is aperfectly wanton assumption, and not the faintest shadow of reason exists for supposing it true. As well might I contend that I cannotdream without dreaming that I dream, swear without swearing that I swear, deny without denying that I deny, as maintain that I cannotknow without knowing that I know. (...) The philosophers in question simply substitute one particular object for all others, and call itthe objectpar excellence. It is a case of the ’psychologist’s fallacy’ (...).Theyknow the object to be one thing and the thought another;and they forthwith foist their own knowledge into that of the thought of which they pretend to give a true account". (vol. I, pp. 274/5)

113. cf. the quotation from Verene, section 3.3, page 47

114. in a passage that immediately follows the one to be cited in section 9.4, page 169 below

115. Palmer (1969, p. 175) translates the German ’Darstellung’ as ’presentation’, which seems to be more accurate. (Cf. footnote480 on page 166)

116. "Es ist ein Herausfallen aus der eigentlichen Erfahrung einer Dichtung, wenn man die ihr zugrunde liegende Fabel etwa aufihre Herkunft hin betrachtet, und ebenso ist es schon ein Herausfallen aus der eigentlichen Erfahrung des Schauspiels, wenn derZuschauer über die Auffassung, die einer Aufführung zugrunde liegt, oder über die Leistung der Darsteller als solche reflektiert. Einesolche Reflexion enthält ja schon die ästhetische Unterscheidung des Werkes selbst von seiner Darstellung. Für den Gehalt derErfahrung als solchen ist es aber, wie wir gesehen haben, sogar gleichgültig, ob die tragische oder komische Szene, die sich vor einemabspielt, auf der Bühne oder im Leben vor sich geht - wenn man nur Zuschauer ist. Was wir ein Gebilde nannten, ist es, sofern es sichderart als ein Sinnganzes darstellt. Es ist nicht an sich und begegnet dazu in einer ihm akzidentellen Vermittlung, sonder es gewinnt inder Vermittlung sein eigentliches Sein". (1960 (WM), p. 112)

52

one that the child had enjoyed. It may be unimaginable to the adult that the social misery that he hadinferred wouldnot be a property of the fair that the child perceived, even not an unseenproperty«117». What the adult then does not realize is that the fair that the child has perceived wasa naive one: it was not the same purportedly ’objective’ (or better:concrete) fair perceived from apurportedly ’naive’ point of view, but rather a different fair; one from which no abstractions are possiblewithout losing one’s object out of sight«118». In other words, a naive objectcannotbe reduced to away of interpreting or perceiving some objective thing!

Husserl has dealt with a comparable idea where he describes the ego that only lives in its intentionalacts«119». Here it is the absence of an awareness of this ego that Husserl takes as the criterion ofwhat I called naive perception:

"... if we ’live’ in the act in question, become absorbed, e.g., in the perceptual ’taking in’ of someevent happening before us, in some play of fancy, in reading a story, in carrying out amathematical proof etc., the ego as relational centre of our performances becomes quite elusive.The idea of the ego may be speciallyready to come to the fore, or rather to be recreated anew, butonly when it is really so recreated and built into our act, dowe refer to the object in a manner towhich something descriptively ostensible corresponds." (Husserl, 1977, Vol. II, p. 561)«120»

The naive observer lives the perception of his object, because it exists due to his performance. Thisperformance itself is not visible for the naive observer, as it appears to him as the naive object, not as hisway of perceiving and constructing it!

When, and only as long as, an observer does not notice himself (or his ’ego’) being involved in the act ofobserving, this observer must be ’naive’: the kind of object that he perceives must be a ’meaningfulwhole’«121», not polluted by conceptual reflections that might turn the observer into a criticalposition. Geertz puts it thus:

"People use experience-near concepts spontaneously, unselfconsciously, as it were, colloquially;they do not, except fleetingly and on occasion, recognize that there are any "concepts" involved atall. That is what experience-near means - that ideas and the realities they disclose are naturally andindissolubly bound up together. What else could you call a hippopotamus? Of course the gods arepowerful; why else would we fear them? The ethnographer does not, and in my opinion, largelycannot, perceive what his informants perceive. What he perceives - and that uncertainly enough - iswhat they perceive "with," or "by means of," or "through" or whatever word one may choose."(1975, p. 48)

The naive object does by definitionnot have a form, and that which a meta-observer might wish todescribe as its form (i.e. as some or other inferable abstract quality of it) belongs to a domain ofdescriptions that is different from the naive observer’s domain of descriptions. The form of the naiveobject does not exist for the (naive) observer as a distinguishable entity. As a form, it can only be denoted

117. Hence, the too easy construction of an ’unconscious’ knowledge, which is often attributed to a naive observer.

118. This phenomenon is not restricted to children, but happens at all levels of intellectual sophistication. Cf. coming to one’s’sober’ senses, after having fallen in love. When this happens, the previous perceptions are reappraised as naive, and the currentperceptions as critical. The opposite transition also happens: after having become ’naive’, an observer loses his critical distance. This hasbeen described a.o. by Devereux (1967) in respect of the incapacity of anthropological field researchers to keep a detached, scientific,position while engaging in the social affairs of some observed tribe. Notice that it is the tangible involvement that enables the naiveperception to occur.

119. I owe this reference to Bakker, 1966, p. 85, footnote 40

120. "... leben wir sozusagen im betreffenden Akte, gehen wir z.B. in einem wahrnehmenden Betrachten eines erscheinendenVorganges auf, oder im Spiel der Phantasie, in der Lektüre eines Märchens, im Vollzuge eines mathematischen Beweises u. dgl., so istvon dem Ich als Beziehungspunkt der vollzogenen Akte nichts zu merken. Die Ich-vorstellung mag "in Bereitschaft" sein, sich mitbesonderer Leichtigkeit hervordrängen, oder vielmehr sich neu vollziehen, aber nur wenn sie sich wirklich vollzieht und sich in eins mitdem betreffenden Akte setzt, beziehen "wir" "uns" so auf den Gegenstand, daß diesem sich Beziehen des Ich etwas deskriptivAufzeigbares entspricht." (Husserl, 1913, Bd. II, Teil I, p. 376/V§12b)

121. as Gadamer calls it in the fragment quoted above on page 51

53

by a meta-observer. The naive observer’s position does not allow a more reflected perception«122».

Like the coordination of first order distinctions not always goes smoothly and may give rise to a secondorder coordination, yielding the objects that I have qualified as naive, so also two persons may fail incoordinating their naive objects. The comparison of two naive objects, one opposing the other, may startan attempt to reconcile them. If the persons manage to coordinate their divergences,critical objectsarise.I will turn to these in chapter 4. On the other hand, if their coordinations fail, a power relation mayestablish. I will spend a few words on that issue in section 4.3.1, after having introduced the concepts of’critical object’, ’substance’ and ’accident’.

122. This can also be illustrated by the following example from psychotherapeutic practice. A young man of about 25 years camefor therapy and complained of nocturnal fears that something or someone would be around in his room. For that reason he always sleptwith a little light on, and he kept his housedoor unlocked at night, so that he could, in case of emergency, always get out of the housewithout extra impediments. His fears had started at the time when he was fifteen years old, at which time his parents were in the courseof a divorce. His father had remarkably aggressive and desparate fantasies at that time, sometimes threatening the children and theirmother to come and kill them. After the young man had perceived the visit of a ghost at night, who tried to strangle him, he wasbrought to the general practitioner the next morning. The doctor had tried to reassure him by pointing out the hallucinatory nature of hisexperiences, and he had prescribed some sedatives. This, however, did not bring the desired release. The patient felt misunderstood andnot taken serious. He was not prepared to consider his percepts as mental stuff. When the patient had swallowed all pills at once, in anapparent attempt to draw all attention from his parents and bring them back together, the doctor returned that he would not give anyother medicines because of this abuse of the pills. What hadnot happened was that someone asked the patient about the ghostperceptions in a respectful way, i.e. in a way that enabled the patient to preserve his naive perception. So when, ten years later, thetherapist allowed the patient to tell about the ghost, it was the first time that this patient did not feel left alone when relating his story.By calling the perceptions hallucinatory, the general practitioner had intended to reassure the patient, but actually he had let him downby refusing to receive the patient’s unbearable experience. (I will return to this in terms of dealing with projective identifications inchapter 11. In fact, the doctor’s behavior amounted to a refusal of what Bion (1962b) called a beta element). It is precisely the(doctor’s) critical position that the patient could not combine with his own naive perceptions. The latter were dealt with by the doctor asif they were some idiosyncratic (hallucinatory) interpretation of the patient, not as something real. Below (see 5.1) I will call such aninterpretation an ’internal object’.

Conversely, by taking the ghost as a serious possibility, and hence by accepting it as a naive object, the therapist showed not tobe afraid of getting involved in these matters, and quickly succeeded in sharing with the patient his naive object, to the extent ofinterrogating him about it. What is more, it turned out to be entirely irrelevant for the purpose of the treatment whether the patient’sperceptions had been hallucinatory or not. What did matter was the patient’s attainment of a position in which he could freely relateabout his observations, as well as about the accompanying emotions, without feeling left alone.

4Critical objects

In this chapter and the following chapters objects are described in which the form-content distinction isrelevant. Of critical discourses two varieties will be presented, that are inspired upon the medievalcontroversy between realists and nominalists. It will be maintained that the critical discourses allow theconstruction of ’mental’ or ’internal’ objects, both by others and by these persons themselves. This willbring us in chapter 6 to the difference between ’classical’ and ’non-classical’ science. The former isqualified by an integration between the two critical discourses (realist and nominalist), whereas in thelatter this integration has gone lost, due to a certain self-reference of the operations of the investigator.This will prepare the ground for chapter 7, which pertains to the issue of self-reference and the problemsto deal with it within the critical discourses. It will be maintained that autopoiesis, both of an observedand of an observing individual, cannot be described in terms of critical discourses. Instead, the notion ofinterface or boundary becomes salient. It is this notion that will eventually be pivotal for the final chaptersof this book.

The structure of this chapter is as follows. In section 4.1 I will introduce the concept of critical object. Insection 4.2 I will discuss the perception of an abstract form as a naive object. Section 4.3 contains themajor idea of this chapter, i.e. the development of critical objects as ways to manage interpersonal conflictsituations.

4.1. Introduction

Notice that what the adult in the example of the fair, given above«123», is missing, is a commonreference point with the child, one about which he may agree with the child, and hence demarcate thepoints of disagreement (and label them accordingly in terms of ’divergent interpretations’). Precisely thisis what a critical discourse would provide him with.

When an observer is able to conceive the form (see 4.2) of an object, i.e. when he is able to describe theobject in terms that in some way abstract from its concretely perceivable features (the content), then let uscall this observercritical and let us call the object, the form of which he thus conceives: thecriticalobject. The difference between the naive and the critical observer is that the former does not distinguishbetween form and content, whereas the latter does make such a distinction.

A critical object, then, is a form-plus-content constellation, that can replace the perception of a naiveobject. The content is perceived as aconcrete object«124»that can be observed irrespective of itsabstract properties; the form is conceived in an abstract sense as something beyond the content, somethingconceivable to be present, but not immediately perceivable. A critical object thus consists of a form and acontent. That what is concrete and perceivable for the critical observer is defined as itscontent. That whatis abstract and conceivable is defined as itsform«125».

Critical objects come into existence in a way analogous to naive objects. Conflict situations have at leasttwo parties that have more or less incongruent practices (or one person having two more or less

123. section 3.4, p. 50

124. or as a set of such concrete objects

125. Notice that ’form’ does not have any metaphysical (aristotelian) connotation. Furthermore, notice that the notions ’abstract’ and’concrete’ are redundant after these definitions of form and content.

54

55

incongruent practices). Like the naive object is made as a modus between individuals to deal withsomewhat incongruent first order distinctions (experiences), so also is the critical object established inorder to establish an objective point of agreement, and to release the conflict from its remaining threats.This is particularly useful between two conflicting parties, but it may also occur within one person whoperceives two conflicting or incompatible naive objects.

In terms of coordinations of actions: when naive objects are not entirely shared with others, then criticalobjects are a solution for this, as a way to coordinate the divergence between two persons’ non-identicalnaive objects (cf. 4.3). Conversely, to the extent that naive objects become again sharedunproblematically, the use and the usefulness of critical objects disappears. This happens for examplewhen consensus concerning a matter of taste is being re-appreciated as a truth. Two persons, who fromtheir own personal appreciations agree on the quality of a particular wine, may conclude that the wine ’infact’ is a good one«126». Then they end up describing the wine in all but objective and detachedterms. The wine itself may contribute to that as well, of course.

A naive object generally arises as a way of coordination between individuals concerning the first ordercoordination of actions that they are having. A critical object generally arises as a way of coordinationbetween individuals concerning the second order coordination of actions that they are having, that is,concerning their naive objects.

Naive objects result out of situations in which experiences are not shared unproblematically (cf. 3.2);critical objects result out of situations in which naive objects are not shared unproblematically (cf. 4.3). Inthose situations, the interacting persons disagree about a particular naive object that they are both involvedwith.

Both in the case of naive objects and of critical objects, the resulting topic of agreement functions as ashared topic of orientation. Naive objects come into existence as the coordination of experiences; criticalobjects arise as a coordination of naive objects. One is tempted to think that observers have a need for an’archimedean point’, as is formulated e.g. in cartesian philosophy«127», as a way to establish ashared topic of orientation, a point of reference for persons when coordinating their lives, as e.g. inconflict resolution. Such an archimedean point may be a relatively parsimonious tool for keeping peace. Itis only after the confrontation between various critical systems, that a cultural relativism can occur as inour century, and perhaps at other moments of intercultural encounter. In such a cultural relativism, theincompatibility of various archimedean points is solved by agreeing that such a point does not really exist,and that all scientific accounts are ’narratives’.

4.2. Gestalt formation and incorporation: the naive perception of form

We may distinguish two classes of naive objects: those that are new (to be callednew naive objects), andthose that, before becoming perceived as naive objects (to be calledderived naive objects), are conceivedas the form of an already existing naive object. New naive objects are primarily those objects that arise asmetaphors (cf. section 3.3«128»). Psychotherapy becomes useful in helping people bring forth thesenew naive objects, which are principally based only upon first order distinctions (experiences). Or, inother words, in a psychotherapy a person may make a new start with creating objects out of hisexperiences. The person may experience relief if he is enabled and capable to make such a new start of

126. Cf.: "... our thoughts would terminate in a complete empirical identity, there would be an end, so far asthoseexperiences went,to our discussions about truth. No points of difference appearing, they would have to count as the same." (James, 1912, p. 86)

127. see sections 5.2.1 and 9.2.2

128. See especially footnote 102 on page 48; also footnote 101 on page 48.

56

constructing naive objects out of experiences, instead of out of previous objects«129».

Like a naive object comes into existence by taking the first order distinctions (experiences) as a token, i.e.by (second order) distinguishing these (first order) distinctions (see section 2.3), so also an abstract formitself can become a (derived) naive object. This happens when a critical observer manages toplay withthis form, in such a way as to use its content as a tool for perceiving (instead of abstractly conceiving)this form.

In order to do that, the critical observer needs to ’forget’ the critical object’s content, and have it availableas a tool for perceiving the derived naive object as a gestalt. The critical object’s content disappears outof sight. It becomes obscured«130». It is this transition from the concrete object into an instrumentof perception, that is entailed by a change of the distinctions that are made. For the concrete objects, asthey serve as instruments for perceiving the form as a naive object, are no longer distinguished as theywere before. The sensory impressions that were the first order distinctions do no longer appear as tokensfor the concrete objects; instead, theymix upwith the second order distinctions, to make up new secondorder distinctions. Furthermore, new first order distinctions arise, due to the instrumental usage of theconcrete objects, and these new distinctions become the new tokens for the newly emerging naive object,which is a gestalt, the elements of which are the concrete objects.

The following scheme summarizes these transitions:

old situation :first order distinctions:

experiences, to be used as tokens for the concrete objects

second order distinctions:coordinations of these experiences, making them into tokensfor the concrete objects

new situation :first order distinctions:

newly occurring experiences due to the instrumental usage ofthe concrete objects

second order distinctions:a combination of the first and second order distinctions ofthe ’old situation’, using the first order distinctions ofthe ’new situation’ as tokens for the newly emerging gestalt(consisting of the concrete objects of the ’old situation’ asits elements)

In the ’new situation’ the second order distinctions are a mix of previously first and second orderdistinctions. The critical object’s form is no longer something reflected or thought about, but rathersomething seen (butnot seen as a form that has a content)«131»

129. cf. footnote 104 on p. 49, on metaphors as new naive objects. In section 7.2.1 I will suggest that this capacity of creating newnaive objects out of experiences is an effect of what is called ’health’. See also Dreyfus & Dreyfus (1986, p. 41), where they discuss theorigination of new ideas in science.

130. Cf. also Maturana’s notion of ’obscuration’, as mentioned in section 2.3.4, esp. footnote 66.

131. This is formulated by Merleau-Ponty in terms of the access we have to a (perceptual) essence: "There is aWesenof red, whichis not theWesenof green; but it is aWesenthat in priniple is accessible only through the seeing, and is accessible as soon as the seeingis given, has then no more need to be thought; seeing is this sort of thought that has no need to think in order to possess the Wesen"(1968a, p. 247)["Il y a un Wesendu rouge, qui n’est pas leWesendu vert; mais c’est unWesenqui par principe n’est accessible qu’à travers le voir, etest accessible dès que le voir est donné, n’a dès lors plus besoin d’êtrepensé: voir, c’est cette sorte de pensée qui n’a pas besoin depenser pour posséder leWesen..." (1964a, pp. 300-1)]

57

The emergence of the form as a new naive object is usually described in terms of gestalt formation. Anew gestalt becomes perceivable as a naive object, to the extent thata) the observer’s original first order distinctions (experiences) are no longer used as tokens, but instead

are combined with their second order distinctions;andb) the gestalt’s constituent elements (the concrete objects) are integrated, i.e. incorporated.

To the extent that the concrete objects remain perceivable, the gestalt itself remains only conceivable asan abstract pattern, but, on the other hand, if they are integrated into a gestalt, then they disappear asperceivable entities.

We notice therefore a certain trade-off: either the observer uses his first order distinctions as tokens andperceives the constituents (i.e., the content of the critical object), or he perceives its form as a naiveobject, and does no longer (second order) distinguish his original (first order) distinctions as tokens.

The activity that transforms a concretely perceivable object (a critical object’s content, or ’concreteobject’) into a tool for perceiving a newly emerging naive object, is one that Polanyi (e.g. 1966; see alsoShotter, 1985, 1993) calls ’knowing from’. Shotter has extensively discussed the act of ’knowing from’,which he considers a third kind of knowledge, next to ’knowing that’ and ’knowing how’.

It is by looking through the concrete objects as tools, i.e., by using them instrumentally and by ’knowingfrom’ them, that the new naive objects arise«132». Shotter (1984, p. 212) describes a dualitybetween what he calls ’tool’ and ’text’: it is by incorporating a particular thing as a ’tool’, that we canknow other things ’from’ it (i.e. use it instrumentally); if we do not incorporate it, it serves us a as a’text’, the meaning of which we can find out.

Whenever a critical object’s form becomes perceivable as a new naive object, it becomesknown fromthecritical object’s contents, which are the concrete objects. The perceiving individualincorporatesthese, i.e.he starts to use the concrete objects as tools for perception. This makes them disappear as perceivableobjects. It is as if the observer abstracts from them, disregards them, whereas in fact he uses theminstrumentally. To the extent that they still remain perceivable, it is at the cost of the perceptibility of thecritical object’s form as a newly arising naive object.

Notice that we have by now two modes of expressing the same thing:a) knowing an objectfrom another objectb) distinguishing newly occurring first order distinctions, by means of acts of distinction that are a

combination of previously performed first and second order distinctions

I call incorporation of an object the change of status of this object that happens as it is factually used asan instrument for the perception of some naive object (that will usually be called a gestalt). By now, Iclaim, we have found a way to understandincorporation in terms ofgestalt formation. I will now firstgive some examples of incorporation.

Merleau-Ponty (1945, p. 167) gives the example of a blind man using a stick as an extension of his own

This conception of seeing an essence can be taken as a paradigm not only for other perceptual modalities, but also for theperception of forms in general (either essence or meaning).

132. Cf. also: ""I noticed that he was out of humour." Is this a report about his behaviours or his state of mind? ("The sky looksthreatening": is this about he present or the future?) Both: not side-by-side, however, but about the onevia the other." (Wittgenstein,1976, p. 179 / part II, section v). [""Ich merkte, er war verstimmt." Ist das ein Bericht über das Benehmen oder den Seelenzustand?("Der Himmel sieht drohend aus": handelt das von der Gegenwart, oder von der Zukunft?) Beides; aber nicht im Nebeneinander;sondern vom einen durch das andere." (Wittgenstein, 1971, p. 285)]

Notice that the German original version says ’the one through the other’, which in fact is more close to the idea of ’knowingfrom’.

58

body. Thus, the blind man has his stick ’participate’ in his own body«133». The way in which heperceives his environment by means of the stick necessarily evades his attention. It has becomeincorporated. That is, the stick’s tactile sensory impressions in the blind man’s hands are not experiencedfocally, but rather ’subsidiarily’«134». They have become second order distinctions, allowing thestick’s encounters with the bottom to become new first order distinctions, i.e. to become the tokens for theobjects that are newly perceived. The stick, as a perceivable object, disappears during this transformation.Before, however, the stick was perceived, the tactile impressions in the blind man’s hands were first orderdistinctions that became tokens for this stick!

if the stick’s tactile then the blind man provided that:impressions are: perceives:

1st order distinctions the stick other 2nd orderdistinctions are made

2nd order distinctions the bottom other 1st orderdistinctions are made

Secondly, let us remember von Foerster’s experiments in which people came to perceive 4-dimensionaldice, named tesseracts. It was through their learning to coordinate particular manipulations of keyboardbuttons, which were connected to computer steered pictures, that these subjects came toseethe tesseract(cf. von Foerster, 1985, pp. 74 ff.«135»). Accordingly, von Foerster understands the emergence ofthe tesseract into a perceivable object, as the ’token’«136»for a newly attained stable mode ofperformance in the experimental subjects. In our terms: the tesseract arose as a new naive object byincorporating the various pictorial changes related to the various button manipulations. The latter were nolonger noticed as events, but they became the medium through which the new object became perceivable.

Third, many mathematicians are known for their skill to perceive the issues of their interest as ’reallyexisting’ entities«137», and not merely as symbols between which formal rules apply. Then newlyinferred aspects of these mathematical ’naive objects’ may first be conceived only as the form of a criticalobject, and later become perceivable as naive objects by incorporating (by knowing them from) thecontents of the available critical objects«138».

Another example is mentioned by Shotter (1989, p. 10): ’Castle Wolfenstein’, a video game involving anintricately built imaginary edifice. Though Shotter presents it as an example of the construction of whathe calls ’imaginary’ objects, I tend to believe that this Castle Wolfenstein is as real as any other castle,but only as soon as one gets involved in the game and becomes able to find one’s way through it, that is:as soon as one is able to perceive the castlefrom its constituents (concrete objects, i.c.: the features of thevideo game). Then it passes from being an abstract form to being a naive object, and it starts to be acomprehensive entity that is really existing, like a piece of music that is being performed. Thus its quality

133. In section 6.2.2 I will mention a similar metaphor used by Bohr in the context of quantum mechanics.

134. The terms ’focal’ and ’subsidiary’ are from Polanyi, 1966.

135. Cf. also von Foerster’s (1992) statement: "Only that we can grasp does exist." However, I prefer his spoken formulation givenin his lecture at the Bergamo conference "Evolution and Cognition", september 1990: "Nothing exists, unless you can grasp it". Cf. alsofootnote 422 on p. 144.

136. cf. footnote 63 on p. 35

137. Kampis (1991, p. 476) mentions this in terms of ’mathematical platonism’. (cf. also footnote 178 on page 70)

138. a similar transition is discussed in section 8.4.1

59

of video game moves to the background«139». It becomes also a consensual object for those whoare involved with the game, just as the Dow Jones Index has become a real object (and not just somethingimaginary) for those involved in the games of stock market transactions, and just as the astronomicalobjects that Galileo observed, existed only for those who were willing to use his telescope instrumentallyaccording to his instructions«140».

To the extent that a player does not succeed in integrating its elements, Castle Wolfenstein remains aform, a thought construct that one may definitely call with Shotter an imaginary entity. To the extent thatthis is the case, a comparison with a ’real’ castle is made by means of a comparison of its components(bricks versus video-images). Only when considered at the level of its components, Castle Wolfenstein isan imaginary object (that is, a form) that can only be conceived, not perceived.

How is incorporation possible? How does the use of concrete objects as instruments for perceptionoriginate? How can the perception of the contents of a critical object be made undone, so that thesecontents can be used instrumentally?

It is through beingplayful with the concrete objects, that their incorporation takes place. I will payattention to the concept of play below in section 9.4, where play is presented as a way of ’un-perceiving’«141». It is also by being playful that the newly appearing object can appear as naive:the ’naivety’ of the object resides in its observer being immersed in play.

Let me conclude this section with perhaps the most interesting case of incorporation. Sheets-Johnstonereads Merleau-Ponty’s ideas on ’flesh’ in these terms. In particular she applies his ideas on ’flesh’ to whataccording to her«142»makes humans unique: lovemaking. "I feel the other’s movement through myflesh and the livingness of my own movement through the flesh of the other."«143»It is here, Ithink, that incorporation between persons takes place in a non-instrumental way. I will contrast this to

139. Hull (1985) even goes further and calls the video game a transitional phenomenon in the sense of Winnicott (1953). He states:"In the heat of playing, there may well be a blurring of boundaries between player and machine, so that the player imagines he is "in"the action." (p. 110)

Likewise, Varela writes in regard of the technology of ’virtual realities’, which uses gloves or suits with electronic sensors formotions, so that these are translated into particular images appearing on a monitor: "For example each time my hand, which appears as a"virtual" iconic hand in my image points to a place, the image that follows corresponds to flying to the place pointed at. Visualperception and motions thus give rise to regularities which are proper to this new manner of perceptuo-motor coupling. What is mostsignificant for me here is theveracityof the world which rapidly springs forth: weinhabit a body within this new world after a shorttime of trying this new situation (i.e. 15 minutes or so), and the experience is of truly flying through walls or of delving into fractaluniverses. This is so in spite of the poor quality of the image, the low sensitivity of the sensors, and the limited amount of interlinkingbetween sensory and image surfaces through a program that runs in a personal computer." (Varela, 1991, pp. 189-190)

140. "The one universally accepted tool employed in astronomical calculation was geometry and its use was not predicated on anyclaims of realism for the mathematical models. The problem for mathematical astronomy was to plot the relative positions of variouscelestial phenomena, not to try to explain them. nor were astronomers expected to astound the world with new revelations about thepopulation of the fixed and perfect heavens. So, whatever else astronomers were to do, it was not to discover any new facts, thereweren’t supposed to be any.

But the telescope revealed new facts. And Galileo had to find some way to accommodate them. Furthermore, to make the newtelescopic findings acceptable, Galileo had to do more than merely let people "look and see" for themselves. As we have alreadydiscussed, the only time appeals to "look and see" carry evidential weight is when the observers share the same source-base. Therefore,Galileo had to identify a common source-base, in this case, it was geometry, and somehow link the telescopic observations to it." (Pitt,1992, p. 153)

141. At this place I restrict myself to an example of the phenomenon of incorporation through play. This can be found in humansucklings who are used to drink from the breast and not from a bottle. Introduction of the bottle is then facilitated when the infant isallowed to play with it by sucking it and by other kinds of exploration. Play enables the infant to accept the bottle, and this acceptanceis not a matter of learning to perceive the bottle as an object, but rather its inverse: to habituate to the bottle by no longer perceiving itat all as an external alien corpus. The breast, on the other hand, is initially not perceived at all by the neonate, but trusted andincorporated immediately.

142. Sheets-Johnstone, 1986, p. 242

143. Sheets-Johnstone, 1986, p. 237

60

what happens in the process of projective identification, see section 11.5, where incorporation can lead toan instrumental usage of the other person’s sensory and intellectual capabilities.

The perception of a critical object’s form, however, is not necessarily always a matter of gestaltformation. This form may also remain an abstract aspect. I will now turn to the two principal types ofagreement that can be obtained between critical observers, i.e. observers capable of distinguishing formand contents.

4.3. Epistemology verified: two ways of allocating ’substance’ and ’accident’ to’form’ and ’content’

The way in which naive objects are perceived may vary with the practices in which they are broughtforth. When conflict situations arise concerning a particular naive object, the two parties may come to aparticular kind of agreement that I will give a central position in my argument. For the conflict orincongruence can be settledby splitting the naive object disagreed about into a part about which nodisagreements exists and a part about which the conflict or incongruence remains.

Critical objects arise as modes of agreeing with others without a more basic lack of consensus withothers. Pask (e.g. 1975) called this ’agree to disagree’. I propose to call the agreed part’substance’«144»and the disagreed part ’accident’. By this division, a conflict or incongruencebetween two naive observers becomes a matter of qualification or appreciation (accidents), and loses itsthreat of a battle about fundamentals (substance). Conversely, if a critical discourse fails, what is left is afundamentalist opposition.

If the conflict can be managed, then the substance becomes the new issue of agreement, the accidents mayremain topics of disagreement. Irrespective of how the substance-accident split is made, what I callsubstance is that which will obtain an agreed status for the interactants: it is that which becomes the moreor less firm ground to stand on in dealing with one another. It is this agreement that is constitutive to theregular sense ofobjectivity that persons share with one another. Thus understood, this objectivity is asocially attained construct and not a feature of a pre-given reality, however convinced the interactants maybe of the objectivity and of the observer independence of reality.

There are two basic ways, which I will call the twocritical discourses, in which substances and accidentsmay become related during conversational interactions: either as form to content, or, reversely, as contentto form:

(agreed) substance (disagreed) accident

discourse A : form contentdiscourse B : content form

That is to say that the conflicting observers may agree that

144. I am using the terms ’accident’ and ’substance’ as respectively denoting ’quality’ and ’qualified’. Though the term ’substance’(’ousia’) is from Aristotle (1955, Categories, V; cf. Copleston, 1962, p. 45, esp. his note 31 on p. 263), my usage of it does not in anyway imply that I am adhering to an Aristotelian metaphysics. I am using ’substance’ in the sense of anesse per se, somethingindependently existing. I am avoiding the term ’subject’ (Aristotle’s ’hypokeimenon’, cf. Categories II), as this term, which is current inpropositional logic, has got too much the connotation of ’ego’ since Descartes (cf. Heidegger, 1961, p. II-141; see also footnote 202 onpage 78). This does not mean that a critical observer is defined as having an awareness of accidents and substances: like the terms’form’ and ’content’ they belong to the domain of a meta-observer. As such, ’accident’ and ’substance’ refer to quasi-objects for thecritical observers that I introduce. These terms enable us to specify some types of relations between and within critical objects. On aprevious occasion (Goudsmit, 1992a) I used the terms ’substrate’ and ’predicate’ respectively for what I now think can better be called’substance’ and ’accident’.

61

a) the issue of their disagreement is a matter of diverging, incongruent or controversial ’appearances’(the accidents), derived from:

b) an indisputable underlying ’essence’ (the substance), about which no disagreement exists and thatserves as a unifying principle;

or they may agree that:

a) the issue of their disagreement is a matter of diverging, incongruent or controversial ’meanings’(the accidents), assigned to:

b) ’symbols’«145»(the substance), about which no disagreement exists.

Thus:

substances/ accidents/topics of agreement: topics of disagreement:

A: as conceived (abstract) form: as perceived (concrete) content:essences appearances

B: as perceived (concrete) content: as conceived (abstract) form:symbols meanings

The idea summarized in this scheme will be one of the leading themes throughout this book. I willelaborate the two possibilities A and B as the templates for two discourses that I will name after themedieval controversy betweenrealismandnominalismrespectively.

Realism and nominalism were, until about the 14th century, the two major opposing doctrines about thestatus of abstract concepts. These so called ’universals’ (such as ’species’ and ’genera’), were taken eitheras really existing ideas (especially as existing in the mind of God), or as mere words or names (whichexisted only in human minds, but not in reality).

Instead of aiming at a contribution to these debates, I will not so much compare the two doctrines ofrealism and nominalism, as propose two kinds of discourse that serve as the contexts in which a criticalobject can be constructed. These two discourses resemble the realist and nominalist doctrines in some oftheir basic assumptions about what is more real: a concrete individual entity or some underlying principle.

The discourse that I will call realist assumes the real existence of anabstractessence which is consideredthe source of varying and diverging, at times contradictory, appearances. This abstract essence is agreed(by those in the realist conversation) to be the substance, i.e. that to which (disagreed) accidents can bepredicated. This essence, as in medieval realism, is considered to be the underlying reality. On the otherhand, the discourse that I will call nominalist assumes the existence of an independent world ofconcrete,objectively observable individual objects and facts, which are the substances about which agreementbetween observers is possible in principle, and in regard of which each person is permitted to develop hisown viewpoints, valuations and interpretations (accidents). In this discourse, like in medieval nominalism,there no assumption of an underlying reality.

Notice thatwithin each of these discourses a speaker may take for granted that substances pre-exist theiraccidents. According to the ideas presented here, however, the very split of a naive object into substanceand accident is in fact one that yields substance and accident simultaneously«146»! Conversely,

145. I will use the term ’symbol’ in accordance to its usage in mathematics: a distinct readable entity, in itself meaningless, towhich a variety of rules and definitions can be assigned (cf. also Pattee’s usage of it in his seminal 1972a paper).

146. cf. footnote 153

62

substance and accident also may disappear simultaneously as a critical object turns again into a naiveone«147».

It is interesting to observe that Aristotle discerns two types of substance, which he calls primary andsecondary substance. According to Copleston these two usages of the term are well defensible and shouldnot lead to ambiguities. Primary substance is understood as the kind of substance most easily accessible toour senses: the individual objects that we perceive in the world; secondary substance, on the other hand, iswhat is called the ’universals’ (genera, species): those substances that are considered persistent. Coplestonremarks:

"It is to be noted that the termsfirst andsecondin this respect are not valuations but mean first orsecond in regard to us (...). We come to know the individuals first«148», and the universalsonly secondarily by abstraction" (Copleston, 1962, p. 46).

But what is primary or secondary? The usages of ’substance’ in the nominalist and realist discourses, as Idefine them, are an adaptation of Aristotle’s ideas of primary and secondary substances respectively. But,apparently in contrast to Copleston as well as Quine, I contest the assumption that we come to observe theindividuals first. Instead, I think, what comes first, both developmentally and in the course ofconversations, is the perception of naive objects. When critical objects develop, an external observer is notentitled to decide what status a perceiving individual will give to the concrete objects he perceives: willthey be perceived as the appearances of some underlying essence, hence as secondary to this essence; orwill they be perceived as the individual items to which some meaning can be assigned, hence as primaryto this meaning?

Which of these two possibilities will be chosen depends on the current commitments and interests of theinteracting persons involved. If they are more interested in solid explanatory grounds, then they mayprefer a realist discourse in which it is agreed that objective essences act upon their senses in divergingways; on the other hand, if they are more interested in pragmatics, then they may prefer a nominalistdiscourse in which the concrecte individuals are the things agreed upon, as the objective symbols to whichdiverging meanings can be assigned.

The usage of terms I am proposing here, is an expression of how observers may come to agree ordisagree: that about which they agree is to be given a more objective, i.e. observer independent, status.That about which they disagree is constructed by them as more subjective, i.e. observer dependent.

Therefore, instead of an ’epistemology naturalized’, we might speak here of an ’epistemologyeconomized’ or ’bargained’. But actually, in order to honor Vico’s principle ofverum factum«149»,I would prefer the term ’epistemology verified’!

It should be noted that the two critical discourses are introduced for the description of ways of arriving atpartial consensus (i.e., about the agreed substance) between interactants. Thus these discourses are definedin terms of their function in social conflict management. This implies the claim that whatever substance-accident splits the interactants may develop, it will always be their usefulness, not their purported truth,that comes first. In this respect my own discourse is primarily about interactants’ usage of terms, andtherefore nominalist. However, I will also venture to describe some limitations of the critical discourses(see chapter 7).

I will give an elaboration of each discourse, both in the context of interobserver disagreement or conflict,

147. cf. section 4.1

148. Quine (1969, p. 89) also prefers the idea of the primacy of ’observable circumstances’, in correspondence to ’observationsentences’ that are considered more primary than concepts ("what does it mean?") or doctrines ("is it true?"): "... observation sentencesare the ones we are in a position to learn to understand first, both as children and as field linguists. For observation sentences areprecisely the ones that we can correlate with observable circumstances ..."

149. i.e., the idea that Truth is made, viz. truth is the outcome of creative constructions, the invention of a ’middle term’ betweentwo propositions that are to be connected by this middle term (cf. Verene, 1981, p. 41; Manson, 1969, p. 21).

63

and in the context of intraobserver disagreement or conflict. In the latter, it is the single observer whocombines the positions of two individual observers. The gist of the following subsections is always thatconflicting naive observers may become critical by seeking answers to these questions: what can beagreed upon? what cannot be agreed upon?/what can be agreed upon to remain a topic of disagreement?

The agreed part always serves as a foothold for the disagreed parts. Either it is a (conceived) essence, orit is a (perceived) symbol; in both cases it is agreed by the diverging parties to be an issue of agreement.

But before I give these elaborations, it is useful to discuss their opposite first, viz. the situation in whichthe interactantsfail to construct critical objects. I will discuss this now, before going into the two criticaldiscourses in sections 4.3.2 and 4.3.3.

4.3.1. The failure to attain critical objects: the manifestation of power

The present section is placed here, as it uses some concepts not introduced before. Thematically, however,it may also be useful to read it immediately after section 3.4.

Whenever two naive observers do not manage to attain agreement on a substance as well as agreement onthe existence of disagreed accidents to that substance, then a power relation may establish. This is forexample the case when a) one of the observers declares 1) his own naive object to be the real thing, and2) the other observer’s object to be a ’mere’ perception (especially a distortion) of this real thing; and b)the other observer accepts this. By this movement, both observers become critical, but in an asymmetricalway. For the less powerful observer is forced to accept that he has a less correct, less immediate access toreality. He is granted a particular private opinion, one that is presented as an accident to a particularsubstance, but he is not invited to argue with the more powerful observer about the substance itself. Thelatter, on the other hand, will be critical only in a partial way, i.e. with respect to the other, less powerfulobserver. He will continue to believe that his own perception is the ’right’ one, and keep his naive object.

For example, a psychotherapist says to his patient: "you perceive your wife’s words in a distorted way.They are in themselves quite innocent, and you look upon them as a threat for you". In fact the therapistrequires here from the patient an agreement about the ’real’ meaning of the wife’s words.

Of course such an agreement may also be prescribed by cultural norms. Thus considered, an ideology is away to impose upon the members of a group an agreement about accidents, additional to the alreadypresent agreement about the substance.

A patient relates how he remembers that when he was about six years old, it happened once thathis father was very proud of him after he had made a particular painting. Then the father boastedabout it in a home meeting with a colleague. The colleague asked for the son to bring the painting,but after seeing it he depreciated it as ’not too original’. Then the father agreed with the colleague,leaving his son with a burden of shame.

Here it is the final withdrawal by the father, that turns the burden of shame upon the son. If the father hadnot withdrawn his sense of pride, the son would have been better protected against the colleague’sunfriendly opinions. Alternatively, instead of agreeing with the colleague, the father might also havepushed the colleague into agreement with him, as in: "you really have no taste, man, this is absolutelybeautiful!" Now, in fact, the father fails to attain a critical position in his confrontation with the colleague.He might have said: "well, in that case we disagree about the beauty of this painting", and that wouldsave the son. Now he failsat the expense of his son. Below in chapter 11 I will discuss this phenomenonin terms of ’projective identification’.

I will now turn to the two basic ways of managing these kinds of conflicts in terms of critical objects.

64

4.3.2. The realist discourse

To start with an example: the various ways a particular person may behave, unpredictably and without anevident blueprint, have led other persons, often intimates, to speculate about an underlying character, oreven mental illness«150», as a ground for explanation, an underlying principle, in an attempt torelate the various incompatible phenomena. This is also what people do in difficult times: they speculateabout the will of some supreme being, according to which all those terrible things are done and in view ofwhich those things become understandable as expressions of his presence or his will«151».

The abstract substance itself is not perceived here. It is conceived as the abstract essence«152»ofwhich the appearances are believed to be anexpression. We recognize here a perspective taken a.o. byKant: the assumed ’thing-in-itself’ is taken as the abstract substance«153». Thus, the form(substance) is used here as a ground for explanation, and as such it is taken as the source from which theobserved appearances spring. This ground is a generating source, a production process. Let ’---->’ be aproduction process and ’x’ its product, then an essence-appearance relation of the realist discourse can besymbolized by ’---->x’.

This discourse is an elaboration of a particular mode of problem solving between two conflicting parties,viz. one that focuses upon the question ’wherefrom?’, one that searches for acommon (back)ground. Anexample of this can be found in the attempts of people to placate adversaries by remarks of the variety"aren’t we both children of one father?". The intended effect of this tactic is to present oneself as well asone’s adversary as mere offspring (appearances) of the same source (essence), and hence as conflictingand diverging on matters that are only accidental, not substantial. The father (God) is taken as the agreed(abstract) substance, the shared principle.

But not only physical conflicts between persons may be dealt with in this way, also their perceptions maybe taken as merely the appearances of one essence. For example, two observers may come to agree that aparticular animal, seen from a distance, appears to one as a horse, to the other as a pony. Then ’horse’and ’pony are considered as the perceived (disagreed) accidents to the (agreed) conceived substance, thegenus ’equus’ (or even the family of ’equidae’).

Likewise, the phenomena that one single person observes can be incongruent and incompatible to theextent that some underlying fixed ground is wanted. Then this single person may act with himself in away comparable to the above given two-person cases. For example, the various ways in which a piece ofwax can be molded and appear to our senses could make one wonder about what it is ’essentially’ thatappears in so many ways:

"But what is this piece of wax which cannot be understood excepting by the [understandingor]«154»mind? It is certainly the same that I see, touch, imagine, and finally it is the samewhich I have always believed it to be from the beginning. But what must particularly be observed

150. cf. section 8.2

151. I recently came across the point of view that protestant religiosity in Holland was traditionally more fervent among thosepeople living in the regions below sea level, i.e. where in the preceding ages the threat of drowning in one’s own house was unrelenting.

152. This essence is not unlike the platonic ’idea’. It resembles what Buridan (±1300-±1358) called the ’universale secundumcausalitatem’ (universal according to causality) (De Rijk, 1992, p. 40). This conception of ’universal’ has been advocated in particularby William of Champeaux (1070-1121) who was one of the champions of medieval realism (cf. Sassen, 1965). De Rijk (1981b, p. 152)remarks that William’s notion of ’universal’ was to be understood in a more aristotelian sense, i.e. as aneidos, a common essence thatresides in each individual (rather than beyond them, in a separate platonic domain of Ideas).

153. See Gram (1980) for an interesting discussion of how the impact of the kantian thing-in-itself upon the phenomena that weperceive cannot be understood but in terms of causality. As he points out, the nature of this causal relation remains problematical sincecausality is understood by Kant as a category of the intellect, andnot as a property of the thing-in-itself.

On the other hand, if we understand substance and accident to be constructed simultaneously, as I propose, this problem doesnot occur, even if the constructing actor himself may be convinced that the substance comes first (as a kind of ’prima materia’).

154. addition by Descartes’ English editors

65

is that its perception is neither an act of vision, nor of touch, nor of imagination, and has neverbeen such although it may have appeared formerly to be so, but only an intuition of the mind"(Descartes, 1641/1973 (second meditation), p. 155 [AT VI, p. 31])

A more contemporary example of this is the mode of reasoning performed by a scientific observer, say aphysician upon examination of a patient’s lungs. Is there something irregular inside, e.g. an infection, andif so, where does it reside? The physician takes a number of observations which he treats as the symptomsof the situation inside the lungs. Thus he defines appearances (noises, symptoms) from the essence (stateof the lungs, disorder). His observations are thus treated as unreliable in principle; his measurements maybe wrong or misleading at any moment due to a failure of the instruments or due to another, oftenunknown, influence. The physician, like other observers, counters this unreliability by gathering moreobservational data. But he willnot consider his observations as certain, as would the translator of anexotic text.

More generally, the classical notion ofstateof a system, is an instance of an essence, whereas thesystem’s observableoutputsare its appearances. Measurements and experimental manipulations are meantto assess the system’s state. By triggering a variety of outputs, and by considering these outputs asdetermined by the system’s state, this state is looked for as the explanation of the outputs. The inputs aretaken as mere boundary conditions.

The realist discourse, therefore, is basically a mode of speaking in terms of causal agents«155», bethey Aristotelian or other. This discourse amounts to theexplanationof appearances in terms of thoseunderlying essences that are conceived to entail them, and to the analysis of these entailments.

4.3.3. The nominalist discourse

To start with an example: the various ways in which a single fact can be interpreted may give rise tomany conflicts between persons. What does the facial expression of La Gioconda mean? what is shemeant to be doing next moment? At least the facts of the facial expression can be agreed upon; they canbe photographed, X-rayed, etc. But their interpretation, the meaning to be attached to it, and, in particular,the best ways of behaving towards a lady with this smile, may remain an issue of disagreement. Peoplemay come to agree here that they do have diverging views as to the meaning of her smile, which they donot dispute itself. Similarly, people may have diverging opinions about the treatment of persons withparticular (agreed upon) behaviors«156».

In this discourse the substance is the perceived concrete symbol, and the accident is a meaning that isassigned to it«157». Thus, unlike the previous discourse, where the form (substance) was used as a

155. cf. James, 1907, p. 51

156. cf. section 8.3

157. Cf.: "Nominalism, then, consists of the refusal to countenance any entities other than individuals. (...) The nominalist’slanguage contains no name, variable or constant, for entities other than individuals." (Goodman, 1970, p. 231)

Furthermore, in opposition to what is achieved in a realist discourse, it is interesting to find that Hume would consider anessence to be a matter of belief, and regard the supposition of it to be unwarranted:

"This very table, which we see white, and which we feel hard, is believed to exist, independent of our perception, and to besomething external to our mind, which perceives it. Our presence bestows not being on it: our absence does not annihilate it. (....)

By what argument can it be proved that the perceptions of the mind must be caused by external objects, entirely different fromthem, though resembling them (if that be possible) and could not arise either from the energy of the mind itself, or from the suggestionof some invisible and unknown spirit, or from some other cause still more unknown to us. It is acknowledged, that, in fact, many ofthese perceptions arise not from anything external, as in dreams, madness, and other diseases. And nothing can be more inexplicablethan the manner, in which body should so operate upon mind as ever to convey an image of itself to a substance, supposed of sodifferent, and even contrary a nature.

It is a question of fact, whether the perceptions of the senses be produced by external objects, resembling them: how shall thisquestion be determined? By experience surely; as all other questions of a like nature. But here experience is, and must be entirely silent.The mind has never anything present to it but the perceptions, and cannot possibly reach any experience of their connexion with objects.

66

common principle, here the form (accident) is what remains unsettled and controversial betweenobservers«158». The meaning (accident) is regarded here to reside in an observer’sway ofdealing«159»with a symbol. The symbol is taken as an in itself meaningless (but agreed upon)concrete fact of the world. Unlike the realist essence-appearance pair, the nominalist symbol-meaning paircan be symbolized by ’x---->’, where ’x’ stands for a symbol and ’---->’ for the way in which this is dealtwith.

The nominalist discourse is an elaboration of a particular mode of conflict management between twoparties, viz. one that searches for acommon problem, or common task,not a common background orunderlying principle. An example of this can be found in the attempts to placate adversaries by remarks ofthe variety "aren’t we both fighting the same enemy?", or: "aren’t we trying to solve the sameproblem?"«160». The intended effect of this tactic is to present oneself as well as one’s adversaryas sharing a common focal problem (symbol), which serves as an agreed (concrete) substance. It is withrespect to this problem that in the nominalist discourse the conflict between the diverging solutionactivities (interpretations) by disagreeing actors is reduced to the format of accidents to a shared problem,presenting the topic of their disagreement as a matter of diverging means to a same end.

This is also what always happens when, for the sake of pacifying the internal tensions of a country, anexternal enemy is sought by the leaders«161». Likewise, for the sake of cogent and compellingpersuasion, ’hard’ empirical data may be sought by persons who wish to account for their (moral,political) preferences.

In this discourse the interobserver conflict or disagreement occurs about how the concrete objects that theobservers are perceiving should be solved, interpreted, understood or dealt with. For example, two huntersmay disagree about the meaning of a particular animal’s behavior. This is the symbol agreed upon, theindividual fact. But will it flee, will it fight? Both expectations are possible, and they can be assigned asmeanings, as (disagreed) conceived accidents to the (agreed) perceived substance, the animal’s behavior.The accidents themselves are not perceived. They reside in the hunters’ diverging preferences for action.No essences are assumed.

As to intraobserver conflict or disagreement, here the nominalist discourse takes the shape of a person’swavering, when he finds himself caught between two ways of dealing with a problem, e.g. between twointerpretations of a text (that is: two interpretations of what its author may have meant). Is theresomething special intended by the author, and if so, what? The translator generates a number of

The supposition of such a connexion is, therefore, without any foundation in reasoning." (Hume, 1777a, Sect. XII, part 1, §§118-119/1975, pp. 151-153)

Likewise, James (1907, p. 47) approvingly mentions Berkeley:"It was the scholastic notion of a material substance unapproachable by us,behindthe external world, deeper and more real than it, andneeded to support it, which Berkeley maintained to be the most effective of all reducers of the external world to unreality. Abolish thatsubstance, he said, believe that God, whom you can understand and approach, sends you the sensible world directly, and you confirmthe latter and back it up by his divine authority. Berkeley’s criticism of ’matter’ was consequently absolutely pragmatistic. Matter isknown as our sensations of colour, figure, hardness and the like. They are the cash-value of the term. The difference matter makes to usby truly being is that we then get such sensations; by not being, is that we lack them. These sensations then are its sole meaning.Berkeley doesn’t deny matter, then; he simply tells us what it consists of. It is a true name for just so much in the way of sensations."

158. Buridan calls this the ’universale secundum predicationem vel significationem’ (universal according to predication orsignification) (De Rijk, 1992, p. 40).

159. This ’way of dealing’ is in the first place an activity, but, for practical reasons, it may be called the observer’s ’impression’ ofthe symbol. This ’impression’ may be a quasi-object for the observer himself. In order to confirm a such a quasi-object, the position of asecond observer is required, from which the performance of the first observer can be viewed. This second position may be taken by theobserver himself. See also section 5.3 on ’internal objects’, and section 5.3.1 on the construction of ’self’ within the nominalistdiscourse.

160. think of two parties in a war, who decide to combine forces in order to fight a third party, in spite of underlying ideologicaldifferences (e.g. the US and the USSR in WW II)

161. cf. Volkan (1988)

67

interpretations which he treats as possible modes of understanding the characters in front of him. Thus hedefines symbols (characters, sentences, descriptions) from the meaning (how the text is to be understood,what the reader is to think, believe, experience, how he is to respond and act). His observations of thesymbols are thus treated as reliable in principle. His interpretations, on the other hand, may be recognizedby others or not; they are uncertain, basically a matter of agreement between interpreters. Likewise, achess player may study the configuration of the chess pieces and waver between two strategies.

Here Wittgenstein’s (1958, e.g. §43)«162» idea comes to mind, that the meaning of a term residesin its usage. For this usage can be taken as the accident assigned to the term whose meaning it makesup«163».

Also in Hume’sEnquiry concerning the principles of morals(1777b) do we recognize the kind ofdiscourse that I am describing here. What matters to Hume for the attainment of moral judgments are ourexperiences, which he describes in terms of ’sentiments’. These sentiments are the valuations that ourmind assigns to concrete actions and situations«164». The sentiments, qua assignments of avaluation to the naked facts, lay the ground for taste and hence for (moral) action«165». Likewise,what I call ’meaning’ is equally of a pragmatic kind.

This conception of ’meaning’, in fact, has roots in the way medieval philosophers, such as Abelard (1079-1142)«166»and Buridan (±1300-±1358), dealt with universals. According to Abelard, our thinkingof universals is to be understood as a description of an intentional act«167». Likewise, Buridanunderstands a meaning as a mode of conceiving, a way in which the thinking of one and the same thing

162. Wittgenstein, 1971, p. 41 / 1976, p. 20.

163. Dummett (1963/1978, p. 190) remarks that "[t]he identification of meaning with use is a small but necessary first step: progressis to be made by asking, for each case, what use consists in and how it is to be described."

164. We may even discover what kind of actions and situations they are:"In moral decisions, all the circumstances and relations must be previously known; and the mind, from the contemplation of the whole,feels some new impression of affection or disgust, esteem or contempt, approbation or blame." (Hume, 1777b, App. I, §240/1975,p. 290)

and:"No, say you, the morality consists in the relation of actions to the rule of right; and they are denominated good or ill, according as theyagree or disagree with it. What then is this rule of right? In what does it consist? How is it determined? By reason, you say, whichexamines the moral relations of actions. So that moral relations are determined by the comparison of action to a rule. And that rule isdetermined by considering the moral relations of objects. Is not this fine reasoning?

All this is metaphysics, you cry. That is enough; there needs nothing more to give a strong presumption of falsehood. Yes,reply I, here are metaphysics surely; but they are all on your side, who advance an abstruse hypothesis, which can never be madeintelligible, nor quadrate with any particular instance or illustration. The hypothesis which we embrace is plain. It maintains thatmorality is determined by sentiment. It defines virtue to bewhatever mental action or quality gives to a spectator the pleasing sentimentof approbation;and vice the contrary. We then proceed to examine a plain matter of fact, to wit, what actions have this influence. Weconsider all the circumstances in which these actions agree, and thence endeavour to extract some general observations with regard tothese sentiments. If you call this metaphysics and find anything abstruse here, you need only conclude that your turn of mind is notsuited to the moral sciences." (Hume, 1777b, App. I, §239/1975, pp. 288-9)

165. "[Taste] gives the sentiment of beauty and deformity, vice and virtue. (...) [It] has a productive faculty, and gilding or stainingall natural objects with the colours, borrowed from internal sentiment, raises in a manner a new creation. (...) Taste, as it gives pleasureor pain, and thereby constitutes happiness or misery, becomes a motive to action, and is the first spring or impulse to desire andvolition." (Hume, 1777b, App. I, §246/1975, p. 294)

166. Abelard’s ideas about nominalism are regarded (de Rijk, 1981b) as a more moderated and more sophisticated version of theextreme and untenable position taken by his teacher, Roscellin of Compiègne (1050-±1124), who is said to have taken universals asmere ’flatus vocis’.

167. De Rijk writes: "Abelard concludes that "intelligo hominem" (...) means something like "I am carrying out man-intellection"rather than "there is (some universal) Man such that I am thinking of him". In Abelard’s view, then, the universal does not have anindependent existence in the outside world, but it owes its existence (or its "being given") to some generative way of thinking, it is aproduct of such thinking, and, as such, it is a powerful tool of understanding which allows man to speak in a general way about entitiesof the outside world.

From Abelard’s way of tackling the problem of universals, therefore, modern interpreters have rightly concluded that the masterwas familiar with the notion of "opaque or non-referential contexts" (which are established by verbs indicating mental attitudes, such asLatin intelligere)." (De Rijk, 1992, p. 39)

68

can be performed«168».

The nominalist discourse I am proposing here is pragmatically oriented, and it resembles the medievalnominalist movement, which also puts an emphasis upon the use of abstract concepts asinstrumentsforconceiving concrete things. For example, a nominalist school of the late Middle Ages was that of the’calculatores’«169», who took abstract concepts as mere tools for reasoning, first of all reasoningabout motion, without claiming that these notions corresponded to really existing entities oressences«170».

The nominalist discourse, therefore, is basically a mode of speaking in terms of qualifications and valueassignments, or, in more up to date terms: information processing. As such, it boils down to the readingof symbols in terms of their practical implications, and to the writing of symbols as a prescription(program) of these implications. Here a conception of ’state’ is implied that, contrary to the abovementioned realist one, does not entail an internal entity. It is, instead, a calculation tool, a concept usefulfor predicting future behaviors (outputs) of a system«171». Thus the nominalist view fits in verywell within a positivist (anti-metaphysical) and pragmatist tradition, however, without being identical toeither tradition.

I will now make some general and comparative remarks about the two critical discourses.

4.3.4. Expression and impression, explanation and interpretation

In the realist discourse the abstract substance (e.g.: ’the father’«172») is conceived, not perceived,whereas the accidents are perceived as concrete objects (’the brothers’). Conversely, in the nominalistdiscourse the substance is perceived (e.g. ’positions on the chess board’), the accidents are conceived(’chess strategy’). In each case either participant is allowed a particular predication of the substance.(Though the interactants do not need to go so far as to permit each other a right of possessing anindividual ’internal representation’, this type of substance-accident split prepares the ground for it).

Both realist and nominalist discourses make a clear distinction between form and content. In the formerdiscourse it is the essence-appearance pair (as substance and accident), in the latter it is the meaning-symbol pair (as accident and substance respectively) that makes up the form-content duality. The accidentsin the realist discourse are a matter ofexpression; the sons are considered expressions of one commonessence (’father’); likewise, an illness, a demon or a mental state are believed to be expressed«173»by the symptoms, like a computer displays its output.

The accidents in the nominalist discourse are a matter ofimpression(or interpretation); a problem isconsidered to yield diverging responses in different actors; the world is considered to impress«174»upon a person (investigator, patient, etc.) in a particular way, like a computer is put to work by its input.

168. See De Rijk, 1992, p. 55.

169. cf. Wallace, 1978, p. 124ff.

170. In a similar vein, as we will discuss in chapter 8, the nominalist discourse in psychiatric diagnosis uses conceptions ofdisorders as tools for deciding about treatments, not as grounds for explanation.

171. We find this conception of state in Minsky’s (1967, p. 16) definition in terms of ’equivalence class’ of input histories. For eachelement of such a class, the system’s input-output relations are equal. A different output is obtained only if the system has an inputhistory that is not in the equivalence class. Thus, Minsky’s definition of state is a functionalist one: it doesn’t matter what history ofinputs a system has undergone, as long as it responds equally to equal inputs. State is not an ontological entity, but ’just’ a conceptuseful for the prediction of outputs.

172. I am using here my example "aren’t we both children of one father?", see section 4.3.2

173. cf. section 8.2.

174. cf. section 8.3.

69

More precisely, we may take ’appear’ or ’express’ as the way an appearance is related to its essence, and,inversely, we may take the verb ’found’«175»to denote the way an essence is related to itsappearance. Thus, we may speak of an essence that appears as (is expressed by) a particular appearance,and, inversely, an appearance that is grounded (essentiated) in a particular essence. Furthermore, I amtaking ’describe’ as an operation inverse to ’impress’ (or ’interpret’). Thus, a symbol is the description ofa meaning, and a meaning is the impression (or interpretation) of a symbol (cf. Goudsmit,1990a)«176».

Schematically:

realist discourse:

manifestation ( expression ) of-------------------------->

appearance essence<-------------------------

’foundation’ of

nominalist discourse:

interpretation ( impression ) of------------------------>

meaning symbol<---------------------------

description of

Notice that the realist relations (appear, found) differ from the nominalist relations (interpret, describe), inthat the former are taken as reallyexistingrelations, whereas the latter are considered as relations thatexist to the extent that an actorperformsthem. This difference will become of our interest when the ideaof ’non-classical science’ (section 6.2), and also when self-reference will be formulated as something thattranscends the two critical discourses (section 7).

Both discourses are dualistic in that they maintain a distinction between concrete and abstract aspects,content and form. In the realist discourse the abstract aspect is considered to exist really (essence); in thenominalist discourse the abstract aspect is considered to be only a conceived qualification(meaning)«177». In either case content and form are believed to be separated.

For this reason the content of a critical object is perceived differently from the way in which a naiveobject is perceived. It is perceived as something concrete.Concrete objects, according to their definition(see section 4.1) are defined as not to coincide with some meaning or essence, unlike Vico’s imaginativeuniversals. The concrete object is perceived either as a (concrete) symbol without a meaning, though ameaning can be assigned to it, or as a (concrete) appearance without an essence, though an essence can berelated to it. Concrete objects exist as devoid of an intrinsic meaning or essence. They are like what Sartre(1943) calleden-soi. For naive objects, on the other hand, such an assignment would be beyond the mark:the naive object entails its own form unnoticed and inseparably.

The substance in the realist discourse (essence) is a point of agreement because it can be taken as a(metaphysical) causalexplanationof diverging, opposite or otherwise incompatible appearances. Theseexplanations, being part of a critical object, are made as a means to appease one’s opponent, but apartfrom that they do not serve any practical purpose. A realist discourse, therefore, is a fertile ground for thedevelopment of a corpus of explanations (’knowing what/why’). For example, in physics it may construct

175. Perhaps, Heidegger’sverb (!) ’wesen’ would be useful here as an alternative to ’found’. I have also considered the fancy term’essentiate’ for this purpose.

176. The pair ’description’ and ’interpretation’ is chosen in order to conform to Löfgren’s work on ’linguistic complementarity’ (cf.section 6.2.3).

177. By way of a metaphor the reader might like to compare the realist assumption with a convex lens, the form (substance) beingits (positive) focal point, and the nominalist assumption with a concave lens, the form (accident) being its (negative) focal point. Thepositive focal point is a topic ofconvergencebetween observers (essence), the negative focal point is a topic ofdivergencebeweenobservers (meanings).

70

abstract entities as the ’real’ things that explain our sensory experiences«178».

On the other hand, the substances in the nominalist discourse (symbols) are a topic of agreement betweenobservers, because they can be taken as a common point of departure for diverging, opposite or otherwiseincompatibleinterpretations. The meanings attached to these substances are not so much explanatory, asgoal-setting guidelines for action (such as chess strategies). A nominalist discourse, therefore, is a fertileground for the development of a corpus of predictions and controls (’knowing how/whereto’). Itemphasizes the practical, technological, usefulness of our concepts, theories and myths, not their referenceto ’real’ entities. It emphasizes the meanings of terms and sentences as their usage in practice. For thatreason, I will often use the term ’application’«179» in order to underscore the pragmatic connotationof the concept ’meaning’.

Finally, switches between the realist and nominalist discourses may occur:

agreed -> disagreed: disagreed -> agreed:

realist -> nominalist: essence -> meaning appearance -> symbol

nominalist -> realist: symbol -> appearance meaning -> essence

The former switch takes place when a critical observer changes his focus of interest, and hence hisagreements, from the abstract form of the critical object to agreement about its concrete content. This isthe case when he finds himself in need of practical decisions in regard of the critical object. The latterswitch takes place when a critical observer changes his focus of interest from the critical object’s concretecontent to its abstract form. This is the case when he finds himself in need of theoretical explanations ofthe critical object’s content«180».

4.4. Recapitulation: naive and critical discourses

The term ’discourse’ is used for a class of language games«181»that have in common a particularpresupposition, mostly implicit. Thus, two critical (realist and nominalist) discourses have been discussed.The realist discourse gives priority to the ’universals’, and hence to the quest for explanations and forunifying underlying principles. This discourse gives focal attention to the difference between ’what itreally is’ (essence) and ’how it appears to the observer’ (appearance). For instance, a general practitionerexamining a patient’s lungs by means of a stethoscope receives a lot a measurement impressions which hetries to interpret as the symptoms of a particular state of the body, e.g. a pneumonia. The pneumonia isthen considered to be the real thing (essence) that is the matter. The noises heard through the stethoscopeare considered only as the symptoms (appearances); they are often unreliable, and the physician will haveto look for as many indicators (symptoms) as possible. Furthermore, the physician is aware of theseperceptual drawbacks, and takes them as inevitable features of his personal equipment as an observer.

The nominalist discourse, on the other hand, gives priority to the ’individuals’, not to the ’universals’. Theindividuals are the points of departure for the observer. I call these individuals ’symbols’; a symbol isused as the carrier of a particular meaning or interpretation that is assigned to it. The nominalist discourse

178. Quine (1970, pp. 227/8) even speaks of a ’platonist mathematics’ that is to some extent irreducible from natural science. (cf.also footnote 137 on page 58)

179. as in: ’application of research findings’ or ’application of a psychiatric diagnosis’; also: ’application (i.e., execution) of acomputer program’

180. cf. footnote 184 on page 74

181. I use this term in the sense of Wittgenstein (1958).

71

gives focal attention to the difference between these symbols and their ’interpretations by the observer’(meanings). A meaning is only secondary to its ’symbols’. Unlike the ’appearances’ of the realistdiscourse, ’symbols’ are defined as observational data that are certain. For instance, a translator of aparticular text will have no doubts concerning the characters and sentences in front of him. He will takethem for granted, but he will be much more uncertain about what they mean, how they can be understoodand translated best according to the author’s intentions. Similarly, a researcher inspecting his raw data willtake them for granted and be much more concerned about how to analyze (i.e.: interpret) them.

The realist discourse is more oriented upon principles and explanations, the nominalist discourse moreupon goals and applications. Accordingly, in chapter 8, for instance, I will contrast activities for the sakeof explanation (e.g. etiology of illness) with those for the sake of application (e.g. treatment of illness),and I will put this in terms of the pairs essence-appearance and meaning-symbol.

Both discourses are characterized by an awareness of the distinction between the abstract concept and theconcrete perception of a particular object. In the realist discourse the abstract concepts are considered toconstitute essences of a disorder. In the nominalist discourse the abstract concepts are used instrumentally(as did the Parisian ’calculatores’). I call these two discoursescritical discourses, because each of themmaintains a distance to their objects of discussion, that allows a critical distinction between ’what?’ and’how?’. An observer in a critical discourse is defined to be aware of this distinction. Critical discoursesare the regular expression modes of scientific communities. I call ’critical objects’ the objects described ina critical discourse. Schematically:

abstract: concrete:

realist discourse: essence (what?) appearance (how?)nominalist discourse: meaning (how?) symbol (what?)

The following table relates my terms introduced here to the terms of Aristotle and of Buridan:

terms proposed here: Aristotle’s terms: Buridan’s terms:

nominalist discourse:substance (symbol ) . . . . primary substanceaccident (meaning ) . . . . . . . . . . . . . . . . universale secundum

predicationem velsignificationem

realist discourse:substance (essence) . . . . secondary substanc e . . universale secundum

causalitatem

The critical discourses can be contrasted to anaive discourse. Here, contrary to the critical discourses, nodistinction is made between the what? and the how? of perceptions. That is, a person in a naive discourseis not interested in essences behind phenomena, or meanings assignable to symptoms, nor is he aware ofhis personal involvement with a particular object or the way in which this involvement influences or evenconstitutes this object. I call ’naive objects’ the objects described in a naive discourse.

A naive discourse is one of observers immersed in aesthetic perception, wonder and emotionalinvolvement. Par excellence this holds for children, but in fact any person may enter into a naivediscourse, as soon as he gets involved in some activity to the extent of losing critical awareness of hisown modes of performance. Unlike critical discourses, therefore, one cannot access a naive discourse as amatter of personal choice. As soon as an actor becomes aware of his involvements, naive perceptionvanishes. On the other hand, a critical discourse is defined as one in which the actor is aware of these

72

involvements, and recognizes them either in terms of a meaning he assigns to a symbol (nominalistdiscourse) or in terms of the appearances he finds and attributes to a particular essence (realist discourse).Both naive and critical objects will be of our interest in our approach of psychotherapy as a phenomenonthat is hard to observe empirically. In particular, I will be interested in the difficulties of covering naiveobjects in terms of critical objects (see also chapters 8 and 9), and in the impossibility to formulate firstorder distinctions, being unreflected experiences, in terms of critical objects (see also chapters 7 and 11).

5Internal objects

In the present chapter I will discuss the construction of a particular type of critical objects, to be called’internal objects’, and I will compare the nominalist and the realist positions with respect to theirconstruction. It will turn out that the realist discourse assumes an identity between the person and hisconstruction of himself (e.g.: ’thinking substance’), whereas the nominalist discourse assumes animpossibility to attain this identity, and instead implies a complementarity between the person and hisself-construction.

The former will be illustrated in terms of Descartes’ ideas on himself as a ’thinking substance’; the latterwill be illustrated in terms of the problems both William James and Sartre discerned when a consciousperson chases after his own consciousness (termed ’judging thought’ and ’pour-soi’ respectively). It willturn out that self-reference in the realist discourse looks different from self-reference in the nominalistdiscourse. This difference became apparent with the rise of what is called ’non-classical science’. Thiswill be discussed in chapter 6. In chapter 7 I will apply these findings to our possibilities and ourcapacities to perceive a special kind of objects: organizationally closed systems. It will turn out that thereis a particular lacuna in the domain of what can be formulated in terms of critical objects. This lacuna isconstituted by the interaction boundary (interface) that an observer establishes with an organizationallyclosed system. Thus we become aware of a limit of what can be studied by means of critical discourse,which is the regular tool box of scientific research. In chapter 8 I will extend my exposition of thenominalist-realist opposition in terms of the construction of ’internal objects’ that are attributed to a third(other) person. As a major example of this, some peculiar properties of the psychiatric diagnosticclassification system DSM will be discussed there.

5.1. Internal objects and the two critical discourses

Internal objects are attributions through which a person, originally related to in the second person(addressed as ’you’ or ’Thou’), becomes related to in the third person (referred to as a ’(s)he’ or an ’it’).When an internal object is constructed, the conversation interactants together create a detachment fromthis third person. This may be the case when the behaviors of a particular third person become the topicof disagreement between people; then the conflict is usually solved by making particular qualificationsabout this person. These may range from rude qualifications such as ’mad’ and ’bedeviled’ to refinedpsychiatric diagnoses. There is always some degree of detachment from a person qualified in thesecircumstances, which may range from violent expulsion of the person, via mildly frowning about him, todetachedly and purportedly ’objectively’ diagnosing him.

’Internal’ objects, such as experiences, representations, thoughts, emotions, intentions, dispositions, states,traits, etc., are usually assigned to a person as properties of his mind«182». Thus, the socialdefinition of persons, and more generally, human ’individuality’, arise as social achievements«183»,as solutions to impending conflicts. The ’internal objects’ are socially constructed as some abstract thingin or about a third (observed) person, a thing that serves for the conflicting conversation partners in one

182. This mind is usually situatedinside the person. In fact there is no compelling reason why these ’internal’ objects should belocated inside a person. Gods, ghosts, demons, spirits, satyrs or nymphs, as they occur in most civilizations, are no less real than theother objects of our culture, ’mental’ objects included. Cf. also: "... in point of epistemological footing the physical objects and the godsdiffer only in degree and not in kind. Both sorts of entities enter our conception only as cultural posits." (Quine, 1953, p. 44).

183. Cf. also Toulmin (1979, 1986). Cf. also Jansz (1991), for an overview of ’personal’ versus ’social’ conceptions of individuality.

73

74

of two ways.Either it serves as a foundation, that is to explain and account for the queer behaviors,or asa façon de parler about these behaviors, a way of qualifying them.

In the first case, if a foundation is sought, there may be agreement about a substance that is believed to be’really there’ insidethe third person (e.g. his alleged ’soul’, ’character’, ’mind’, ’personality’, ’ego’, andthe like). In the second case, the qualifications of the observed behaviors are helpful in making(technological, moral or other) practical decisions. Here the observers focus upon the ’objectivelyobservable’ behaviors, about which agreement exists, in order to make their decisions (e.g.: "what shouldwe do with a person who behaves thus...?").

I will now elaborate the construction of internal objects in both critical discourses (sections 5.2 and 5.3).In particular, I will pay attention to this construction by the person himself. This, as will be elaborated,looks entirely different in the two critical discourses. I will use these differences when in chapter 7 I willfocus on self-reference and the impossibility to describe it in critical discourses.

As has been prepared already at the end of section 4.3.4, an observer may switch from one way ofconstructing internal objects to another. A switch from a nominalist mode to a realist mode happens whenthe observer comes to believe that the constructed mental function does exist as a mental structure that isa kind of substance«184»for the observed person’s behaviors. This type of switch is likely whenthe observer has reasons to lay more emphasis upon finding proper (causal or other) explanations for theobserved phenomena, for example when a doctor is expected to account for his deeds in terms of’knowing what’ is the matter with a patient (and in terms of ’knowing why’).

To take another example, it was in order to avoid the vagueness inherent to theoretical interpretations, thatthe early behaviorists stuck to a preference for studying pure behaviors. By describing regularities (e.g. ofthe S-R type, and, even more, of the S-O-R type), they hoped to remain true to their principles. However,they came to implicitly assume a great variety of concepts, such as neuronal connections, learninghistories, habits, response hierarchies and the like: all newly created abstractions that at times evenobtained the status of essences. For example, Thorndike, who in fact antedated classical behaviorism,presented his ’law of readiness’ in terms of ’conduction units’. Though these were lacking a precisephysiological meaning, he used them nevertheless as the abstract carriers of ’conduction’ (Hilgard, 1956,p. 18)«185». Gergen (1984) describes how, in a similar vein, the notion of ’self-concept’ wasunderstood by the neo-behaviorists as an internal mental structure, which "... is generally viewed as theproduct of environmental influence" (p. 73), but which also "determine[d] the character of behavioralresponses" (p. 72). Thus, an internal entity is constructed for the sake of explanation, and an unnoticedswitch has been made from a nominalist to a realist discourse (cf. sections 4.3.4 and 8.4).

The opposite switch from a realist mode to a nominalist mode happens when the observer comes to

184. Cf Winograd’s description of such fallacious reification:"1. A scientist observes some recurrent pattern of interactions of an organism.2. He or she devises some formal representation (for example a set of generative rules or a "schema") that characterizes the

regularities.3. The organism is assumed to "have" the representation, in order to be able to exhibit the regularities.4. (Depending on the particular sub-field) The scientist looks for experiments that will demonstrate the presence of the

representation, or designs a computer program using it to see whether the behavior can be generated by the program.The error is in the reification of the representation at step 3." (Winograd, 1981, p. 248/249)

Cf. also Zimring, 1974, p. 123: "Another [assumption is] that emotions, feelings, and thoughts exist as internal objects that cancause behavior and experience. We think that it is the anger within a person that causes him to slam a door, whether or not he is awareof the anger."

Also Giel (1982, p. 90) mentions what he calls the ’skid’ from a descriptive frame to an explanatory one (in terms of a person’s’structural background). This ’skid’ will be elaborated in section 8.4.

185. On the other hand, in the best of behaviorist traditions, Hull (1943) warns against such reification of behavior functions: "Toreify a function is to give it a name and presently to consider that the name represents a thing, and finally to believe that the thing sonamed somehowexplainsthe performance of the function." (p. 28). Then, as a ’prophylactic’ he proposes "to regard, from time to time,the behaving organism as a completely self-maintaining robot, constructed of materials as unlike ourselves as may be." (p. 27), since"[t]he temptation to introduce an entelechy, soul, spirit, or daemon into a robot is slight" (p. 28).

75

believe that the constructed ’essence’ (or ’character’, ’unconscious’, ’mental state’, etc.) is in fact amental function. This type of switch is likely when the observer has reasons to lay more emphasis uponpractical implications of his observations, implying that what matters is not so much whether or not aparticular mental entity is present, as what can be done about it. By way of example of this, we may thinkof what happens when a physician is expected to account for his deeds in terms of ’knowing how’ toproceed with a patient.

5.2. Internal objects in the realist discourse

The realist assumption is concerned with the objective existence of the internal object as a substance, aninternal state of affairs. For example, the unexpected and incomprehensible behaviors of a personconsidered insane may be explained in terms of the possession of a demon, or in terms of the existence ofa particular mental state. When such an explanation is given, the demon or the mental state are introducedas an essence, the appearances of which are the behaviors of the ill person (’symptoms’). This conceptionis usually elaborated in the tradition oftheoretical explanations, in which questions of the ’how come?’variety are given a central position. Here the internal object is constructed as amental structure. This maybe a ratherpermanentfeature, such as a person’s ’character trait’«186», but it may also be a moretransientmental state, such as a person’s state of consciousness«187»at a particular moment, aswell as his ’unconscious’ states in the sense of psychoanalysis. In brief, it may be anything as long as it isused as an abstract substance (essence), as a state of affairs that isconceivedto be internally the case. Iam therefore not so much concerned with what a mental structureis and what not, as proposing this termfor that what a critical observer is constructing, when for explanatory purposes he refers to an internalstate of affairs in another person. This type of constructions«188»contributes to the development of(metaphysical) theories about mind, and subjectivity, and that which causes them.

5.2.1. The construction of ’self’ in the realist discourse: Descartes

A realist construction of ’self’ develops when a person attempts to account for his own modes ofperformance. It happens when this person decides that a particular mentalstructuremust be (or must havebeen) the case inside himself. This may be a wish, an intention, a conviction, or something else that canbe indicated as a ’state of mental affairs’. For example, in everyday language, inspired on vulgarinterpretations of psychoanalysis, this can be expressed in statements like "I must have wished itunconsciously". The notions of ’crime passionnel’ and ’irresponsibility’ (’non compos mentis’), as theyare used in jurisdiction, connote similar references to mental structures that express themselves in thebehaviors to be explained or accounted for. The same holds for those situations in which a person agreesto be possessed by a ghost or devil. It is the internal (mental) state of affairs that is constructed here as aquasi-object, not for some observed third person, but for oneself (qua observed person). This internal stateof affairs is meant to be an explanatory ground for one’s conduct, to be used when accounts seem to berequired by others«189».

186. Giel (1982, p. 90) mentions the explanatory function of concepts like ’character trait’.

187. Such as the ’substantive parts’ as William James called those parts of the stream of consciousness that are accessible to ourintrospection (see section 5.3.1; cf. James, 1890, p. I-243). These are also the elements that we seek recourse to when accounting for ouractions.

188. Clearly, whenever such an internal structure has been constructed, it will be regarded as really existing, not as ’merely’constructed (cf. section 2.3.4).

189. This is one of the basic tenets of Shotter (1984), viz. that the reference to an inner mentality as a source of explanation ofone’s own behavior, is au fond a social act, meant to satisfy other persons.

In psychotherapeutic encounters one may find a large variety of this kind of statements, mostly meant to account for thepatient’s incapacity to do something particular. Thus a person describes his own problem in terms of:

"a certain impossibility inside myself, already existing quite a time, that impedes me, to replace that which I got from my fatherand from my mother, to replace that with something different, so that I can say: "look, this is what I do"; and this ... perhaps mightenable my acceptation of [a particular burden in the patient’s life].

76

But also in order to account to oneself for one’s own existence can such an explanatory ground beconstructed. This is what has been the primary aim of Descartes, and for a long time it has served as aparadigm for common sense and a groundwork for modern science. It seems expedient, therefore, to giveit some attention:

"... I saw from the very fact that I thought of doubting the truth of other things, it very evidentlyand certainly followed that I was; on the other hand if I had only ceased from thinking, even if allthe rest of what I had ever imagined had really existed, I should have no reason for thinking that Ihad existed. From that I knew that I was a substance the whole essence or nature of which is tothink, and that for its existence there is no need of any place, nor does it depend on any materialthing; so that this ’me,’ that is to say, the soul be which I am what I am, is entirely distinct frombody, and is even more easy to know than is the latter; and even if body were not, the soul wouldnot cease to be what it is. (Descartes, 1637/1973, p. 101 [AT VI, p. 32-33])«190»

and:"... but it cannot be that when I see, or (for I no longer take account of the distinction«191»)when I think I see, that I myself who think am nought." (Descartes, 1641/1973, p. 156 (secondmeditation) [AT VI, p. 33])«192»

It is always the observation of a particular mental performance (thinking, seeing, etc.) that permitsDescartes to conclude his essence to be a ’res cogitans’. And, once this thinking substance is proved, theexistence of the thinker himself is saved from existential doubts.

A person’s ’cogitations’ are considered the accidents (appearances) of this substance(essence)«193». But notice that something peculiar is the matter with the way substance andaccidents hinge together. For the essence isdefinedin terms of its performances: ("a substance the wholeessence or nature of which is to think"). Indeed, one may wonder what good reasons there are for

(...) It is a certain impossibility which does not, by which I, well no, I am clearly not capable to have those emotions which Ihave with respect to it, [those emotions] of which I think that they are vented through that sadness, to have such a control of them thatthey do no longer bother me."

["Een bepaalde onmogelijkheid binnen mijzelf, al heel erg lang, die het mij onmogelijk maakt om naast datgene wat ik van mijnvader en van mijn moeder heb meegekregen, daar iets voor in de plaats te stellen, waardoor ik dus kan zeggen: "nou dit doe ik dan",wat dan ..... het accepteren ervan [n.l. van een bepaalde last die de patient in zijn leven ervaart] misschien mogelijk maakt.

(...) Het is een bepaalde onmogelijkheid die het mij niet, waardoor ik, nou, nee, ik ben dus niet in staat blijkbaar om die emoties die ikdaar rondom heen heb, waarvan ik dus denk dat die zich misschien wel, als gevolg, in die triestigheid ventileren, om die dusdanig ondercontrole te krijgen dat ik daar eigenlijk dus geen last meer van heb."]

190. ["... et qu’au contraire, de cela même que je pensais à douter de la vérité des autres choses, il suivait très certainement quej’étais; au lieu que, si j’eusse seulement cessé de penser, encore que tout le reste de ce que j’avais jamais imaginé eût été vrai, je n’avaisaucune raison de croire que j’eusse été: je connus de là que j’étais une substance dont toute l’essence ou la nature n’est que de penser,et qui, pour être, n’a besoin d’aucun lieu, ni ne dépend d’aucune chose matérielle. En sorte que ce moi, c’est-à-dire l’âme par laquelle jesuis ce que je sous, est entièrement distincte du corps, et même qu’elle est plus aisée à connaître que lui, et qu’encore qu’il ne fût point,elle ne laisserait pas d’être tout ce qu’elle est." (Descartes, 1963, p. 604)]

191. Precisely this non-differentiation, mentioned by Descartes in passing, is what Merleau-Ponty considers the crucialmiscalculation of what he calls ’the philosophy of reflexion’. Cf.: "To reduce perception to the thought of perceiving, under the pretextthat immanence alone is sure, is to take out an insurance against doubt whose premiums are more onerous than the loss for which it isto indemnify us: for it is to forego comprehending the effective world and move to a type of certitude that will never restore to us the"there is" of the world." (Merleau-Ponty, 1968a, p. 36)

["Réduire la perception à la pensée de percevoir, sous prétexte que seule l’immanence est sûre, c’est prendre une assurancecontre le doute, dont les primes sont plus onéreuses que la perte dont elle doit nous dédommager: car c’est renoncer à comprendre lemonde effectif et passer à un type de certitude qui ne nous rendra jamais le "il y a" du monde." (1964, p. 58-9)]

192. cf. also his sixth meditation (1641/1973, p. 190 [AT VI, p. 78]) and the second meditation (1641/1973, p. 150 [AT VI p. 25)

193. Sartre (1943, p. 122), speaking in terms of "this cartesian substance the attribute of which is thought" ("cette substancecartésienne dont l’attribut est la pensée"), calls this essence a ’substantialist illusion’ ("l’illusion substantialiste de Descartes").

77

distinguishing substance and accident at all«194». For Descartes it was the quest for an indubitableand certain foundation of his own existence. This foundation is usually denoted with the term’archimedean point’, i.e. the foothold from which to argue and observe the world«195».

It is interesting to notice that Hintikka (1962, p. 16) regards the relation of ’cogito’ to ’sum’ for Descartesto be "not that of a premise to a conclusion. Their relation is rather comparable with that of aprocesstoits product. The indubitability of my own existence results from my thinking of it almost as the sound ofmusic results from playing it."«196»

Thus my existence is the existence of a thinking substance, and it can be inferred from its cogitative acts.But at the same time, that which is inferred (theproductof the inference), existsin these cogitations andis immediately present through them«197». My existence (’sum’), according to Hintikka, is in asense intuitively self-evident (1962, p. 15), in that I cannot sensibly think the contrary (’I do not exist’). Itis this intuitive self-evidence that does not appear to me (the thinker) as theconclusionof an inference,but rather as something that is undeniably given, due to the very fact of my thinking performance. Hence,my intuitively self-evident existence is the ground, or substance, of which my cogitations are theexpressions (appearances).Notice that according to Descartes it is the person himself (i.c. Descartes) who observes his own mentalperformances«198»and explains themfor himselfas appearances of his own existence, thus provinghis own existence (’ergo sum’). In this way Descartes’ invention, the ’res cogitans’ or ’Cogito’, is meantboth as the object and as the subject of his argumentations. The ’I’, the existence of which is sought to beproven, is alleged to be the ’I’ of the same person who wants to find this proof, and therewith to receivethe account of his own existence. The identification of these two I’s is typical for hisargumentation«199». It is from the conviction that the thoughts of the observed ’I’ are the thoughtsof the observing ’I’ as well, that the self-observing person comes to believe himself (qua observed person)to be immediatelyknowable to himself (qua observer). This ’I’ as an object of argumentation is in aparticular privileged way accessible to this ’I’ as a subject of argumentation. The Cogito is believed to betransparent to itself«200». It is this self-transparency that makes the homunculus question irrelevant

194. This is how William James in fact argued, when opting for a radical reduction of the thinker to the thought (cf. section 5.3.1).

195. See Bernstein (1983, p. 16): "It is less clear what is the Archimedean point in Descartes’ philosophy - whether it is the cogitoor God himself".

196. Hintikka refers here to Descartes’ letter to Morin of july 13, 1638, in which the lux-lumen distinction is introduced.It is interesting in the present context to find that Descartes (1638, [AT II, p. 205]) describeslux as primary tolumen, the

former being the source of the latter. We may, therefore, understandlux as comparable to the cogitative acts, the thinking process(Descartes [AT II, p. 203] speaks of ’movement or action’), andlumenas comparable to the resulting certainty about one’s ownexistence.

[It seems Hintikka is not entirely precise in his own text. For he continues: "or (to use Descartes’s own metaphor) light in thesense of illumination (lux) results from the presence of a source of light (lumen)." The terms ’lux’ and ’lumen’ are swapped here, theirdifference in fact being explained at pp. 203-205 (AT), and not at p. 209, as Hintikka has it.]

197. Cf.: "The peculiarity of [the] relation [between ’cogito’ and ’sum’] explains Descartes’s vacillation in expressing it in that hesometimes speaks of theCogito as an inference and sometimes as a realization of the intuitive self-evidence of its latter half" (Hintikka,1962, p. 17).

Here, ’cogito’ and ’sum’ are respectively meant by the former and latter half of the inference. They are the observedperformance (accident) and the existence to be proven (substance). Notice the double meaning of the term ’cogito’: it is used both forthe existence of the cogitating substance and for the very performance of cogitation.

198. Heidegger (1961, p. II-148) emphasizes Descartes concept of ’cogito’ as a ’cogito me cogitare’.

199. "... to speak accurately I am not more than a thing which thinks, that is to say a mind or a soul, or an understanding, or areason..." (1641/1973, second meditation, p. 152 [AT vol. VI, p. 27])

Cf. also Sartre’s criticism of this identification (footnote 233, p. 84).

200. "...that there is nothing which is easier for me to know than my mind" ((1641/1973, second meditation, p. 157, [AT VI p. 34]and: "... mind in its intellectual activity in some manner turns on itself, and considers some of the ideas which it possesses in

itself" (1641/1973, sixth meditation, p. 186 [AT VI p. 73])Cf. Husserl’s ideas (as mentioned by Strasser, 1985, p. 136/7) on the self-transparence of consciousness:

"... the sort of being which belongs to the mental process is such that the latter is essentially capable of being perceived in reflection."

78

to Descartes.

The Cogito, as an issue of conflict management in the realist discourse, is the form of a critical object, anabstract issue of agreement. It is agreed to be an irrefutable foundation, but one to which only its ownerhas free access. The Cogito is (implicitly) agreed to existbeyondthe domain of social interactions inwhich it is constructed«201». By implication, I (that is: my Cogito) am what I am able to knowbest«202», and even entirely. But notice that this knowledge of mine is not different from myknowledge of the appearances of this Cogito, i.e. the cogitative acts. The distinction between abstract formand concrete contents does then not so much apply to the Cogito and its appearances, as to the positiontaken by a participant in a social encounter. The Cogito as well as its appearances are concrete for itsowner, but abstract for all others.The distinction between abstract form and concrete content boils downto the distinction between the position of others and the position of myself!This is a major conclusion wecan draw from Descartes’ attempt to describe his own mental objects within a realist discourse. When inchapter 7 (section 7.1) I will return to the imperfections of the realist and nominalist discourses todescribe self-reference, I will refer to this finding.

It is from the point of view of the other that my Cogito is an abstract form, and it is from my own pointof view that it appears as concrete, i.e. perceivable to me. Descartes’ vacillation between intuitive self-evidence and inference, as Hintikka put it«203», can be understood as a wavering between the pointof view of others (inference) and the Cogito’s own point of view (intuitive self-evidence).

Critical objects exist as composites of agreed and disagreed parts. The agreed part (form) is here theexistence of a Cogito that can be inferred from the occurrence of thinking acts in a person. The disagreedpart (content) consists of these thinking acts. Their occurrence is how the Cogito appears to itself, butthey do not appear to the other persons with whom conflicts are managed«204». The self-observingperson who is convinced of the existence of his own Cogito, does so in the context of socialdisagreements or conflicts, which he attempts to manage. By agreeing that each person possesses such acogitating nucleus, people not only accept all other things as unwarranted in principle, but they also

(Husserl, 1982, p. 99/§45) ["Die Seinsart des Erlebnisses ist es, in der Weise der Reflexion prinzipiell wahrnehmbar zu sein" (1922,p. 84/§45)]

201. Clearly, this ’existence beyond the domain of social interactions in which it is constructed’ is a quasi-object for those involvedin the interaction.

202. Heidegger describes Descartes’ position thus:"Das Bewußtsein meiner selbst kommt nicht zum Bewußtsein von den Dingen hinzu, gleichsam als ein neben dem

Dingbewußtsein herfahrender Beobachter dieses Bewußtseins. Dieses Bewußtsein von den dingen und Gegenständen ist wesenhaft undin seinem Grunde zuerst Sebstbewußtsein, und nur als dieses ist Bewußtsein von Gegen-ständen möglich. Für das gekennzeichnete Vor-stellen ist dasSelbstdes Menschen wesentlich als das zum Grunde Liegende. Das Selbst is sub-iectum." (1961, p. II-155)]["The consciousness of myself is not additional to the consciousness of the things, as if it were an observer of this consciousness whoproceeds in juxtaposition to the consciousness of things. This consciousness of the things and the objects is essentially and basically firsta self-consciousness, and only as such is consciousness of ob-jects possible. Man’sself is essential for the indicated act of re-presentation, as that which underlies it. The self is sub-iectum." (my translation)]

Likewise:"... even if subsequently [the doubt’s] own existence imposes itself upon me as a limit to the doubt, as a something that is not nothing,this something is of the order of acts, within which I am henceforth confined. The illusion of illusions is to think now that to tell thetruth we have never been certain of anything but our own acts, that from the beginning perception has been an inspection of the mind,and that reflection is only the perception returning to itself, the conversion from the knowing of the thing to a knowing of oneself ofwhich the thing was made, the emergence of a "binding" that was the bond itself." (Merleau-Ponty, 1968a, p. 37)

Cf. also the above quotation from Descartes, associated with footnote 190 on p. 76.

203. cf. footnote 197 (p. 77)

204. It is interesting to notice, however, that Hintikka (1962, pp. 17-8) argues that the sentence "I exist" does have a bearing uponthe persuasion of conversation partners. The search for consensus with others, therefore, seems to be not alien to Descartes’ argumentsconcerning his own existence.

79

construct each other as isolated and inaccessible to one another«205».

Empathic understanding of a person is considered possible only by drawing analogies between my ownbehaviors and those of the other person. If bursting into tears is an expression of sorrow in my case, thenit will be in his. This is a realist understanding of empathy, one that often can be found amongpsychotherapists, who believe that, however subtle their empathic understanding may be, it ’must be’ amatter of drawing inferences from the observed behaviors of the patient«206». In section 9.3 I willpay attention to an alternative for this ’reasoning by analogy’ model of empathy.

Finally, notice that the Cogito’s self-transparency is in the first place apostulatedone. From the fact thatwe are able to inspect our thoughts and emotions, Descartes concludes that the contents of our mindcoincide with the mind that reflects on it. This implies that the Cogito’s self-transparency is complete, atleast in principle. This has been contested by William James, in his distinction of ’substantive’ and’transitive’ parts of a person’s so called ’stream of consciousness’, to which I will turn below in section5.3.1. For the present moment we may wonder that in case our Cogito’s self-access were not complete,then it might also be that in some respects it is better accessible to others than to its owner. If that is true,then there is no more reason to regard empathy as a matter of inference by analogy.

5.3. Internal objects in the nominalist discourse

The nominalist assumption is concerned not with an internal state of affairs, claimed to exist objectively,but rather with an external state of affairs: the world shared by the observers. This world is taken as asubstance, a set of symbols, of which the person in question is assumed to create his own meaning. Thismeaning is in the first place a particularway of dealingwith the ’external reality’«207», a way ofmaking a particular sense out of it, for practical purposes. In the nominalist discourse the person’s deviantor perhaps incomprehensible behaviors are considered as idiosyncratic accidents to this environment,pragmatic qualifications of it. This environment, in itself, is considered as a meaningless world of symbolsto which meanings can be given. Notice well that it is by the external observers that a third person (e.g. apatient) is assigned such a particular or peculiar meaning assignment to this environment! This meaningmay entirely be a quasi-object for this third person, also when the meaning assigned by the observedperson is called a particular ’impression’ or ’representation’ of the ’external reality’«208».

This nominalist conception has been defended most eloquently by Ryle (1949). He proposed to understandmental phenomena entirely in terms of qualifications assigned to external behaviors, instead of as entitiesinternal to the observed person.

This conception is usually elaborated in the tradition oftechnology and applied science, in which

205. Cf.: "... la relation avec autrui devient pour [la psychologie classique] incompréhensible. En effet, qu’est-ce d’abord que lepsychisme, celui d’autrui ou le mien, pour la psychologie classique? Un point sur lequel tous les psychologues de la période classiques’entendaient tacitement était le suivant: le psychisme, ou le psychique estce qui est donné à un seul. Il semblait en effet qu’on pûtadmettre sans autre examen, sans autre discussion, que ce qui est constitutif du psychisme en moi comme en autrui, c’est ce qui estincommunicable. Ce psychique en moi je suis seul à le saisir, par exemple mes sensations, ma sensation de vert, ma sensation de rouge,vous ne les connaîtrez jamais comme je les connais, vous ne les éprouverez jamais à ma place. Il résulte de cette idée que le psychismed’autrui m’appparaît comme radicalement inaccessible; du moins dans son existence même. Je ne puis pas attaindre les autres vies, lesautres pensées, puisque par hypothèse elles ne sont ouvertes qu’à l’inspection d’un seul individu, celui qui en est le titulaire." (Merleau-Ponty, 1951, p. 19)

206. e.g.: "... Allport (1937) subscribed to a complete divorcement between intuitive processes and inferential processes, although hepointed out that in any given act of understanding both processes are involved. He also asserted that it is virtually impossible todistinguish their products. We go further and assert that the process called intuition by Allport and other writers is actually an inferentialprocess in which the cues are inaccessible to self-examination." (Sarbin, Taft & Bailey, 1960, p. 181)

207. Notice that ’external reality’ is to be understood as Maturana’s term ’environment’, viz. the world as perceived by an externalobserver (cf. section 2.1.2).

208. cf. also footnote 184 (page 74)

80

questions of the ’whereto?’ variety are given a central position. Here the internal object is constructed as amental function, for example a person’s ’delusion’, his wishes, ’information processing styles’. Moregenerally, his experiences are taken as modes of dealing with the environment. These functions areconstructed by the critical observers, as ways of speaking about the person under observation. They are infact the meanings that the observers assign to the observed person’s behaviors (’symptoms’), and as suchthey are the names they give to the kind of coping that they perceive the observed person to perform withrespect to his environment. These meanings may concern ratherpermanentmodes ofperformance«209», as well as rathertransientones«210». To give an example: a person maybe regarded to be ’paranoid’. Then in the nominalist discourse this paranoia is given as a meaning to thisperson’s modes of coping with his environment; furthermore, these coping performance modes arethemselves considered as (over-suspicious, distorted etc.) meanings the person assigns to his environment.

Thus, unlike the realist assumption, the nominalist assumption entails a kind ofdouble meaningassignment«211»: the person observed is assumed to act as a meaning assigner«212», and thisis also the meaning that the observers are assigning to this person’s behaviors (without reifying these intomental structures). In section 5.3.1 I will return to this double assignment that accompanies the nominalistdiscourse.

meaning assignmentobserved person ---------------> environment

^

meaning assignment

observer

In a rather pure manner we find the nominalist assumption prevalent in the technology of model building.A model is usually meant as an action that performs a particular function«213». A model isconsidered a ’functional equivalent’, if its outputs are indistinguishable from those of the modelledsystem. The ’Turing test’ is a well known formulation of this idea. Chess playing by computers has beenelaborated along these lines. Such a model, then, is a meaning (or interpretation, action) assigned to thesymbols (i.c. a particular chess configuration). It is then no longer an issue whether or not this artificialchess is similar to chess playing by humans, but the model’s performances arenamed’chess’ because ofwhat they do, (viz. the production of chess movements that are, at least to some extent, functionally

209. such as those Wittgenstein calls ’dispositions’, a term by which he means skills (cf. Ter Hark, 1990, p. 234)

210. Such as the ’transitive parts’ (see section 5.3.1 below) as William James called those parts of the stream of consciousness thatdo not allow being pinned down by introspection. James also distinguished the ’feelings of tendency’ (cf. 1890, p. I-249) i.e. theexperiences of tending to some state of consciousness, but not yet having arrived at it. These in fact are described in a similar way ("Ifwe try to hold fast the feeling of direction, the full presence comes and the feeling of direction is lost" p. I-253). Cf. also ter Hark, 1990,pp. 241ff.

211. which is not entirely unlike Giddens’ idea of a ’double hermeneutic’, as in: "The conceptual schemes of the social sciencestherefore express adouble hermeneutic, relating both to entering and grasping the frames of meaning involved in the production ofsocial life by lay actors, and reconstituting these within the new frames of meaning involved in technical conceptual schemes." (1976,p. 79)

[’lay actors’: cf. what I call ’observed persons’; ’the frames of meaning involved in technical conceptual schemes’: cf. what Icall ’critical observer’s domain of descriptions’]

212. Clearly, these meanings, purportedly assigned by the observed person to his environment, are quasi-objects for him, unless heis made aware of them (as happenspar excellencein psychotherapeutic conversations). Whenever the meanings of the observed person,as he experiences them himself, become of focal interest to the observer, the latter enters in the former’s life world. How to putadequately into words the meanings assigned by the observed person to his environment, is one of the major stumbling blocks of theinterpretative social sciences (cf. Schutz’ (1932) so called ’postulate of adequacy’).

213. In Löfgren’s (1990) terms: "... a model of an action is not a description of the action, but another action that shares essentialproperties with the action itself." See also section 6.1.2 on Rosen’s concept of ’model’.

81

equivalent to those by human chess players). Furthermore, ’chess playing’ is here a meaning assigned byan external observer to the machine’s performances; it is not a description of some internal state of themachine.

A mental function may be anything as long as it is used by the critical observer as an abstract accident(meaning), as a mode of performance that isconceivedto be operational in a person. Again, I am not somuch concerned with what mental functionsare and what not, as proposing this term for that what acritical observer is constructing, when for practical purposes (such as: prediction and control of behaviors,etc.) he refers to a performance mode in another (observed) person. This type of constructions contributesto the development of techniques and technology in general.

5.3.1. The construction of ’self’ in the nominalist discourse: James and Sartre

A nominalist construction of ’self’ develops when a person attempts to assign meanings to his own modesof performance, like when he wishes to control and predict his own behaviors for practical purposes. Hemay then decide that a particular mentalfunctionmust be (or must have been) the case inside himself, ofa kind he could have attributed also to others«214». It is not an essence of his behaviors that is ofhis interest here, but rather a (practically relevant) meaning assigned to these behaviors. As a result, thereis always some ’self’ involved thatdoesthe meaning assignments as well as some ’self’ thatis themeaning assigned. For these two a variety of terms has been used, such as ’I’ and ’me’, ’judgingThought’ and ’empirical person’, ’pour-soi’ and ’ego’, etc. We will see that, unlike in the realistdiscourse, the members of these pairs of terms do not coincide.

Paradigmatic for this line of argument have been the ideas of William James (1890) on introspection. Inhis description of the ’stream of consciousness’ James introduces the notions of ’transitive’ and’substantive’ parts. These, according to James, together constitute the stream of consciousness«215».The transitive parts are considered the trajectories along which our consciousness arrives at its contents.Though crucial constituents of the stream of consciousness, they are themselves not accessible to aconscious inspection that is a detached observation. Whenever one might wish to grasp them as if theywere already the contents of one’s consciousness, they will actually turn into such contents, becomingsubstantive parts and lose their transitive quality. Their transitive quality, thus, is

214. That it is in the first place a construction from the point of view of anexternalobserver may be illustrated by a well knowncase from the history of astronomy, which I am in the first place due to Linschoten (1964). The British astronomer Maskelyne and hisassistant Kinnebrook differed about the exact time measurement by means of observation of star movements (see Stigler, 1986, p. 240).Though in 1796 Maskelyne did notice a systematic difference between observational data of himself and his assistant, Maskelyne clearlydid not construct a particular mental function that was to be attributed to his assistant. In fact, the difference was regarded asinexplainable, and eventually even as irregular! Exit Kinnebrook. Quoting Sanford (1888, p. 8), Stigler (p. 241) gives Maskelyne’s ownwords: "I cannot persuade myself that my late assistant continued in the use of this excellent method (....) of observing, but rathersuppose he fell into some irregular and confused method of his own, as I do not see how he could have otherwise committed such grosserrors".

Clearly, whatmight have been understood by Maskelyne as a mental function (and quite a regular one indeed), to be dealt withas an internal object in Kinnebrook, was obviously not recognized and constructed by him. Instead, he sticked to attributing to hisassistant an ’irregular and confused method’. Twenty years later, however, the very fact of these (and other) regularities gave rise to thenotion of a’personal equation’, introduced by the German astronomer Bessel. He did what Maskelyne omitted. Each observer’smeasurements could be compared with those of others and also with experimental setups (yielding respectively the observer’s ’relative’and ’absolute’ personal equations, as Sanford (1888) calls them). A range of factors (from astronomical to ’psychical’) couldexperimentally be found that determined an observer’s personal equation.

Sanford extensively describes Bessel’s actions and considerations, of which most interesting for us is Bessel’s decision toinclude alsohis ownmeasurement performance and compare it with those of other astronomers. Unlike Maskelyne, he established andaccepted for himself a particular personal equation, according to which he perceived and administrated measurement phenomena slightlydifferently from others. Thus he succeeded in constructing a mental function for himself comparable to those he constructed for others.

215. "Consciousness (..) does not appear to itself chopped up in bits. Such words as ’chain’ or ’train’ do not describe it filty as itpresents itself in the first instance. It is nothing jointed; it flows. A ’river’ or a ’stream’ are the metaphors by which it is most naturallydescribed" (1890, p. I-239)

82

elusive to conscious inspection«216». These transitive parts are considered the interconnectionsbetween the contents of our thoughts«217».

Distinguishing between the ’empirical person’ and the ’judging Thought’, James describes the latter as aprocess and the former as the object of its knowledge:

"Personality implies the incessant presence of two elements, an objective person, known by apassing subjective Thought and recognized as continuing in time. Hereafter let us use the wordsME and I for the empirical person and the judging Thought." (1890, p. I-371),

Of course the ’empirical person’ or ’empirical me’ is considered by James as an observable, like otherthings in the world, but it also exists as "a man’s inner or subjective being, his psychic faculties ordispositions" (1890, p. I-296). The ’empirical person’ is an object of consciousness, whereas the ’judgingThought’«218» is precisely not that. Anticipating Sartre, who also denied the positive existence ofconsciousness as a transcendental entity, James writes:

"The only pathway that I can discover for bringing in a more transcendental thinker would be todenythat we have any direct knowledge of the thought as such. The latter’s existence would thenbe reduced to a postulate, an assertion that theremust bea knowercorrelative to all thisknown"(1890, p. I-401)

But, as we will see, unlike Sartre, James takes the existence of this knower as equivalent to the existenceof the thought:

"If the passing thought be the directly verifiable existent which no school has hitherto doubted it tobe, then that thought is itself the thinker, and psychology need not look beyond." (1890, p. I-401)

Cartesian though this may look at first sight«219», it differs from Descartes’ position in that Jamesdoesnot require an observer-observed relation between thought and thinker«220». Indeed, the veryobserver is missing; there is no essence or ’ground’ inside the person that serves as a subject. Instead,

216. "Like a bird’s life, [the stream of our consciousness] seems to be made of an alternation of flights and perchings. The rhythmof language expresses this, where every thought is expressed in a sentence, and every sentence closed by a period. The resting-places areusually occupied by sensorial imaginations of some sort, whose peculiarity is that they can be held before the mind for an indefinitetime, and contemplated without changing; the places of flight are filled with thoughts of relations, static or dynamic, that for the mostpart obtain between the matters contemplated in the periods of comparative rest.

Let us call the resting-places the ’substantive parts,’ and the places of flight the ’transitive parts,’ of the stream of thought. Itthen appears that the main end of our thinking is at all times the attainment of some other substantive part than the one from which wehave just been dislodged. And we may say that the main use of the transitive parts is to lead us from one substantive conclusion toanother.

Now it is very difficult, introspectively, to see the transitive parts for what they really are. If they are but flights to aconclusion, stopping them to look at them before the conclusion is reached is really annihilating them. Whilst if we wait till theconclusionbe reached, it so exceeds them in vigor and stability that it quite eclipses and swallows them up in its glare. Let anyone tryto cut a thought across in the middle and get a look at its section, and he will see how difficult the introspective observation of thetransitive tracts is. The rush of the thought is so headlong that it almost always brings us up at the conclusion before we can arrest it. Orif our purpose is nimble enough and we do arrest it, it ceases forthwith to be itself. (...) The attempt at introspective analysis in thesecases is in fact like seizing the gas quickly enough to see how the darkness looks." (James, 1890, pp. I-243/4)

See also footnote 210 (page 80).

217. "The transition between the thought of one object and the thought of another is no more a break in the thought than a joint in abamboo is a break in the wood. It is a part of the consciousness as much as the joint is a part of the bamboo." (James, 1890, p. I-240)

218. A comparable terminology is used by by G.H. Mead, in his definition of the ’I’ and the ’me’ (1962, p. 175)

219. cf. section 5.2.1 on the Cogito as coincidence of thinker (observer) and observed thoughts

220. cf.: "To say that one can identify one’s own psychological states places an arduous burden on theoretical speculation. Such aconclusion would entail a concept of mind in which a psychological process could essentially ascertain its own states. Rather than asingle stream of consciousness, one would be forced into a mental dualism in which one level of process acted as a sensing andrecording device and a second process furnished the stuff to be sensed and recorded. Such a dualism is sufficiently awkward that one isled to conclude that the assumption of internal perception is a reconstructed form of the subject-object dichotomy represented in thetraditional metaphor of external perception. In this case the object is displaced inward, and one is left with the image of an inward eyein the process of perceiving itself." (Gergen, 1984, p. 66)

83

James (p. I-304) even playfully suggests the term ’sciousness’, emphasizing that it is not so much a self-awareness that is involved in our stream of consciousness, as a more primary relatedness to the objects ofour thoughts. Thus, the thought, qua process, differs from its object, and whenever the thinker himself istaken as the object of his own thoughts, it is not the thought itself, qua process, that is thisobject«221». Indeed, one would be inclined to understand this as a necessity, though James does notformulate it expressly.

This entails that any form of self-reflection creates a distance between the thinker and himself, one thatwas not present in ’sciousness’. Thus, the ’judging Thought’, which is in fact a passing ’sciousness’ thatis not concerned with itself, can be understood as a transitive part of the stream of consciousness, viz. theone that is presently actual. On the other hand, the ’empirical person’ is to be understood as a substantivepart.

Though the position of Sartre, like that of James, cannot be qualified as nominalist with respect to thestatus of universals - as is the traditional meaning of the term -, he may be called a nominalist in ourpresent, derived, sense with respect to his ideas on consciousness. According to Sartre, consciousness isbeyond our capacities to objectify. In opposition to Descartes, Sartre maintains that the ’I think’, which isthe subject of the consciousness, is never able to found its own presence«222». In this respect Sartreis also critical of Husserl’s concept of a ’transcendental ego’ that in fact is meant to offer such afoundation of consciousness«223». Instead, Sartre conceives consciousness as a ’pour-soi’, which isa hole in the world of the things, which exist ’en-soi’. It is this pour-soi that acts as a source of meaningassignments to the world. The assigned meanings are laid over the en-soi world like a net.

On the other hand, in opposition to the pour-soi, Sartre considers the ’ego’ as a constructed entity that iscontemporaneous«224»to the en-soi entities of the world, and of a comparable status. As such it iscomparable to James’ notion of ’me’. But more emphatically than James, Sartre maintains that the ’ego’ isa construct«225», made by consciousness when reflecting upon itself and searching for its ownunity. For example (and:par excellence), when one discovers oneself observed by another person, Sartremaintains«226», one is reduced to an ego that exists in the ’regard’ of the observer«227». Assuch, a person’s ego, as a locus of states, actions, and qualities«228», is also an issue of the’science called psychology’«229», but does not coincide with the person’s consciousness and hislived experiences. It never will, and therefore, what we call the ’ego’ and its states, actions and qualities,is never the essence of our consciousness, set aside the essence of ourselves. There is no such essence,according to Sartre, and hence there is no mental structure that serves as the source of our expressions.

If we nevertheless want to speak about such an essence, it must be as a ’nothingness’. It is in this respectthat we must call Sartre a nominalist, for whatever we affirmatively say about consciousness, it will be a

221. Cf.: "... this thought, that is, the I, is in itself a process that is principally unacknowledgeable for the subject. As soon as it isknown, it becomes part of the ’me’, the empirical self" (Jansz, 1991, p. 85)

222. Sartre, 1943, p. 122

223. The transcendental ego, being the subject of consciousness, was considered as consisting of pure acts. It was meant by Husserlas an attempt to radically rethink Descartes’ Cogito, but also, in a sense, to continue his programme.

224. Sartre, 1966, p. 86.

225. Cf.: "Allerdings sagt Nietzsche: "Der Begriff des "Ich" als Subjekt ist eine Erfindung der "Logik"."" (Heidegger, 1961,p. II-185)

["However, Nietzsche says: "The understanding of "Ego" as subject is an invention of "logic"."" (my translation)]

226. 1943, p. 307

227. Marthe, at the age of eight months and two weeks, clearly did not exemplify these thoughts. Whenever she discovered to bethe object of laughter and observation, even by a camera, she joined in with an abounding jollity.

228. Sartre, 1966, p. 54

229. ibid.

84

name, a model, a construct or whatever, but not the description of an essence. Consciousness is the sourceof meaning assignments, but cannot itself be qualified in positive terms.

In the nominalist discourse a split is made not between what is accessible to the ’owner’ of the Cogitoand what is accessible to the others, but rather between ’I’ and ’Me’, the untouchable (transitive) processversus the observable constructed (substantive) subject matter. Here the accessibility of the behaviors ofthe observed person is almost equal to all who are interested in the assignment of a meaning to them, sothat a ’me’ or ’ego’ is constructed as a thing about which disagreements are possiblein public. Thus, aperson may find himself in disagreement with others about ’who he is’. This public disagreement abouthis ’ego’ or ’me’ would be impossible in a realist discourse, where the other persons would not be granteda bit of authority in matters of one’s own Cogito«230». In the nominalist discourse thesedisagreements are possible, and a person might even construct his own personal uniqueness as adivergence from how he is defined by others«231». On the other hand, the inaccessibility of thisperson’s ’I’«232» is also equal for all:it is neither a symbol nor a meaning, and it cannot beconstructed as such. It is in fact beyond the domain of the nominalist discourse. When in chapter 7(section 7.1) I will return to the imperfections of the realist and nominalist discourses to describe self-reference, I will also refer to this finding.

As I mentioned above in section 5.3, there is in the nominalist discourse a doubleness of meaningassignment. The observer may not only find himself labeling his own behaviors in terms of internalobjects (i.c. mental functions), but he may also come to regard himself as the one who does the meaningassignments to his environment. This is the double assignment inherent to the nominalist discourse, nowapplied by an observer to himself.

Unlike in the realist discourse, as exemplified by Descartes’ Cogito, in the nominalist discourse these twoassignment roles do not coincide, but instead remain disjoint«233»: while the observer is observinghimself in dealing with the world, he assigns to his performance a particular meaning, viz. that hisperformance, in turn, is a meaning assignment to the world. Compared to the situation of observer (e.g.clinician) and observed (e.g. a patient diagnostically described as paranoid), the self-observer is himself ina double role (e.g. both clinician and patient). Qua self-observer he labels his own performances as aparticular way of dealing with his environment (cf.: paranoid), and qua self-observed person he is seen toassign his own meanings to this environment«234»(e.g.: ’these people all want to poison me’). The’double hermeneutics’ of observer and observed person go together here,but they do not merge!The latterrole of observed person who assigns a meaning to his environment can only come into existence due tothe former role of (self-)observer! This is to say that a person’s assignment of a meaning to hisenvironment exists as an actonly to the extent that this person is able to observe himself assigning thismeaning. If he does not observe himself in doing so, any meaning assigned by him to his environment canbe sensibly conceived only in terms of a construct made by another person«235»and must be onlya quasi-object for himself.

On the other hand, if he does observe himself in assigning a meaning to his environment, he in fact takes

230. In pyschotherapeutic sessions this is the phenomenon of the therapist who claims some certainty about the patient’s mentalstructure (e.g. a particular desire), and who requires from the patient to admit and recognize thereal existenceof this mental object.Within a nominalist discourse, on the other hand, this type of power display would not be necessary; the therapist would only speakabout the patient’s performances in functional terms, wihtout claiming such real existence, as in: ’it seems as if you are not willing tolook for a different job’.

231. Furthermore, the meanings thus constructed might become reified in terms of the ’deeper motives’ of the person, or in terms ofa ’true self’ etc. (cf. the reductions described in sections 8.4 and 6.1.1)

232. or ’judging Thought’ (James), ’pour-soi’ (Sartre), or even ’measurement process’ (Bohr)

233. Precisely the unrecognized presence of this doubleness in the cartesian Cogito is considered by Sartre as a contamination of it(1966, p. 73).

234. This is achieved by what Sartre calls our prereflexive cogito (Sartre, 1943, pp. 19-20). Cf. also footnote 42 on p. 29.

235. Clearly, this would be a meaning assignment of the variety as discussed below in section 8.3.

85

an external point of view in regard of himself. What he (as observer) then does is to consider particularmodes of functioning of his own (as observed person), and (as observer) to regard these as ways in whichhe (qua observed person) deals with his environment. Thus, the meaning that he assigns (qua observer) tohis modes of functioning is that these are meaning assignments«236»by himself (qua observedperson) to his environment. For example, he may find himself thinking and dreaming about a particularperson, and conclude that he may be in love with this person (and accordingly he may take responsibilityfor this state of affairs).

The self-observing person who constructs his own assignments of meanings to his own behaviors, does soin the context of social conflicts and attempts to manage them. By taking his behaviors as publicly agreedupon, he may use the meanings he assigns to them as legitimations of his own valuations of the world.Indeed, his individual goals and purposes may be the names and qualifications he assigns to his ownbehaviors. As such, he may disagree with others with respect to the assigned meanings, such as his ownopinion and course of actions in respect of the (agreed) facts, but not with respect to these facts. Theyserve as the agreed substance between critical observers.

To the nominalist it is (at least in the last resort) not so much of interest what consciousness is, or whatthe ’I’ is, as how to act, in respect of both oneself and others. Thus, according to Sartre it is of particularimportance not to use one’s ego as an explanatory alibi. It is when one uses his ego in order to accountfor his deeds«237», that is, when one avoids his absolute responsibility for his deeds, that onecommits the fallacy of ’mauvaise foi’. This fallacy in fact is a particular case of a switch from anominalist to a realist discourse (cf. section 4.3.4). Sartre’s concept of ’mauvaise foi’ is a case of ameaning taken for an essence (as in section 8.4). Instead of taking the ego as an explanatory ground,Sartre wishes to underscore man’s absolute freedom and responsibility.

Finally, the two roles of observer and observed (as in Sartre’s opposition between consciousness (which isa pour-soi) and the ego (which is an en-soi), or likewise as in James’ duality of ’I’ and ’me’) are notidentified to one another, as did cartesian realism in postulating the self-transparency of the Cogito.Rather, in the nominalist discourse, as for example presented by James or by Sartre, acomplementaryrelation is maintained between the knowing or observing ’I’ (or ’pour-soi’) and the known/observed ’me’(or ’ego’). The term ’complementarity’«238», as we will see, is a technical one and denotes theimpossibility to describe a process of observations (cf. the ’I’) in terms of the objects that can beperceived through it (cf. the ’me’).

It is the difference between the realist and nominalist discourses with respect to the description of the ’I’that is typical for the divergence between these discourses which occurred due to several developments inmathematics and physics since the end of the 19th century. I will turn to this in the next chapter.

236. pre-reflexive, cf. footnote 42 on p. 29. The term ’meaning assignment’, of course, does not need to make part of the observer’svocabulary.

237. as in: "it is all due to my character, which I can’t help". A version of it that looks more sophisticated is: "it is not compatiblewith my professional identity to refer you to dr X"

238. Though the term ’complementarity’ can be found in James (e.g. 1890, p. I-206), he uses it only in respect of dissociativemental states of (hypnotized and other) persons. This is not what later came to be meant by ’complementarity’, especially since theworks of Niels Bohr, to which I will turn in section 6.2.2. In any case, James’ treatment of the relation between transitive andsubstantive parts is clearly a case of complementarity, since the transitive parts cannot be perceived but by freezing them intosubstantive parts. Their ’transitivity’ escapes.

6Classical and non-classical science

"What is important, I believe, is that behavioral science shouldstop trying to imitate only what a particular reconstruction claimsphysics to be"(Kaplan, 1964, p. 11)

There is a difference between ’classical’ and ’non-classical’ physics. In the words of Pattee:"In classical physics it is assumed that only the finalresultsof measurement, that is, the numericalvalues, are the relevant information one needs to explain and predict the behavior of the system.(...) The measuring device itself is assumed to have vanishingly small physical interaction with thesystem and, in contrast to quantum theory, the act of measurement is not recognized in the formalmodel. (...) Quantum theory does not allow [the] precise conceptual distinction between laws andmeasurements or between the objective structure being measured and the subjective agency ofmeasurement. (1979, p. 220)

A major difference between what I will call classical and non-classical science is that non-classicalscience is qualified by an explicit admission of self-reference, both in calculations and in empiricalfindings, whereas in classical science self-reference was ignored or even eschewed.

In this chapter I will maintain that classical science can be characterized as a combination of the realistand nominalist discourses, a combination that became obsolete when self-reference became a more seriousissue of attention. Classical and non-classical science are understood as rooted in the realist-nominalistcontroversy.

I will attempt to explain the integration of the realist and the nominalist discourses in terms of theinvention of ’classical science’, i.e. science in which the observer himself stays out of the domain ofstudy. I will maintain that in classical science the realist and nominalist discourses could be combined,whereas in non-classical science they became no longer compatible due to the inclusion of self-reference.

Thus, we will find the following trade-off:

- if self-reference is not allowed to occur in the objects of study, then the observer is permitted tokeep a detached, ’scientific’ position, from which objective, detached, observations are possible;realist and nominalist positions can be combined into a ’classical solution’, which makes use ofmodels (meanings) as tools for performing abstract calculations about the dynamics of statechanges (essences) behind the empirical phenomena (appearances/symbols))

- however, if self-reference is allowed to occur in the objects of study, then the observer cannotmaintain his detached position; realist and nominalist positions can no longer be combined; the’classical solution’ breaks down, in favor of two separate, disjoint, modes of describing thephenomena: one in terms of ’ideal’, unmeasured states, the other in terms of measured phenomena

A conclusion for the field of psychotherapy (section 6.3) will be that self-reference has survived onlywithin a realist tradition, as a mode of performance usually called ’introspection’, whereas in fact a non-classical elaboration has been lacking so far.

86

87

6.1. Classical science: the integration of the two discourses

In classical science, i.e. the scientific tradition based upon the ideas and ideals of (among others) Galileo,Descartes, Locke, Newton and Kant, a reciprocal relation can be recognized between observable (concrete)and conceivable (abstract) terms: the abstract conceivable can be constructed out of the concreteobservable, and the latter can be deduced from the former. These two movements were symmetrical toone another, and in fact became more relevant than the old opposition between realism and nominalism.

The introduction of calculus«239» in the natural sciences allowed the usage of formal termswithoutthe requirement that each step of the calculation corresponded to some state of the system«240».The status of formal, abstract, terms thus became less relevant. These terms were usually considered torefer to real, underlying entities, but at other times not to refer to anything, and to have the status of mereauxiliary tools (for purposes of calculation or argumentation). The domain of empirical applications forthe new natural sciences itself became more and more of focal interest, and the controversy on the statusof universals faded.

More generally, through the construction of calculus present and past phenomena could be explained, andfuture phenomena could be predicted (and controlled)«241». By means of calculus it was possible toformulate general laws of nature and compute their implications. Thus, the tension between two questionswas overcome: the question about the source or foundation of phenomena and the question about theattainment of a particular goal. Above in section 4.3.4 I qualified these two questions in terms of’metaphysical explanation’ versus ’technological application’.

6.1.1. Content and form in classical science

In classical science, explanations and applications were recognized as based upon the same mechanisms,which, formulated in terms of natural laws, could be calculated both forward and backward in time. Thisrequired a partitioning of the world into states (also named phases, or conditions) and the dynamics that

239. I am emphasizing here calculus, such as for instance Descartes’ analytical geometry, and not so much ’scientific method’ (e.g.as proposed by Descartes), because only calculus allows the computable connection between explanations in terms of backgrounds oressences on the one hand, and on the other hand practical implications in terms of meanings assigned to empirical findings. The usageof a scientific method is only a necessary condition to this computable connection. It is not a sufficient one, as is apparent from thediscipline of biology, where the classical scientific method is applied without rigorous results comparable to those in physics andchemistry (cf. Rosen, 1985).

240. Cf.: "... that it is not necessary (even in the strictest reasonings) significant names which stand for ideas should every time theyare used, excite in the understanding the ideas they are made to stand for. (...) ... inalgebra, in which though a particular quantity bemarked by each letter, yet to proceed right it is not requisite that in every step each letter suggest to your thoughts, that particularquantity it was appointed to stand for." (Berkeley, 1970 (1734), p. 93)

Descartes (1954/1637) would not have agreed with this, since he took his calculations in geometry to concern issues that havean already established (platonic or ideal) status ("If, then, we wish to solve any problem, we first suppose the solution alreadyeffected..." (1954, p. 6/1637, p. 300)). Nevertheless he was also confronted with numbers that he called ’imaginary’: "Neither the truenor the false roots are always real; sometimes they are imaginary(*); that is, while we can always conceive of as many roots for eachequation as I have already assigned, yet there is not always a definite quantity corresponding to each root so conceived of" (1954,p. 175/1637, p. 380)

(*) translators’ footnote:"... This is a rather interesting classification, signifying that we may have positive and negative rootsthat are imaginary. The use of the word "imaginary" in this sense begins here."

Cf. also: "If the circle neither cuts nor touches the parabola at any point, it is an indication that the equation has neither a truenor a false root, but that all the roots are imaginary." (Descartes, 1954, p. 200/1637, p. 393). Thus, notwithstanding his own convictions,Descartesseemsto have accepted the notion of imaginary roots as abstract calculation tools!

241. Cf: "It was part of the incredible insight underlying Newton’s approach (...) to realize that where a particle is going to be mustbe somehow entailed from where it is now." (Rosen, 1991, p. 91)

88

ruled the transitions between these states«242».

In a sense, classical science was of a realist signature. The abstract entities it dealt with were believed tobe really existing. But, taken this for granted, irrespective of ontological assumptions, these abstractentities were also used as instruments for power, or prediction and control, as was for instance the case innavigation maps (cf. Latour, 1988). Hence, at calculation time«243», also nominalist aspects werepresent.

Abstract concepts were now related to concrete, empirically accessible observables in a twofold way: theywere bothusedas the calculation tools (or, appreciations, integrations, summaries, or models) of empiricaldata, thus fitting in a nominalist tradition, and they were alsoconsideredas the latent entities behind themanifest appearances, thus fitting in a realist tradition. In this way, the realist-nominalist distinction, asintroduced in section 4.3:

form : content :

realist discourse : essence/ appearance/agreed substance disagreed accident

nominalist discourse : meaning/ symbol/disagreed accident agreed substance

integrated to:

form : content :classical science : essence = meaning appearance = symbol

(abstract ideas) (concrete data)

The leading distinction became that between form and content, between the abstract and the concrete. Theultimate elaboration of this came from Kant, in his treatment of the a priori - a posteriori distinction. Butalso in Bacon we find already the ’modern’ tendency to arrange facts into abstract rules. The distinctionbetween substance and accident became less relevant. It became more of interest to those concerned withepistemology than as a tool for the description and categorization of the world. This is different from themedieval tradition of aristotelian science. Apart from this, the new leading distinction between form andcontent entailed a disregard for rhetorics and in particular a rejection of argumentation based uponauthority or tradition. Issues of debate were established in classical science, at least in principle, in termsof empirical findings and theoretical models; not in terms of a taxonomy of substances and accidents.Agreement and disagreement between observers were not recognized to play a constitutive role (cf.section 4.3) and instead were treated as a distraction from the objective study of Nature.

242. "As Wigner (...) has emphasized, Newton’s greatest accomplishment was not the discovery of the laws, but the discovery of thecrucial separation of thelaws from the initial conditions that must be determined by measurements." (Pattee, 1992, p. 182; italics added)

More generally, these initial conditions constitute the system’s initial state, and the natural laws serve to calculate the variouspossible changes of state over time.

Cf. also: "Newton’s Laws thus serve to transmute the initial dualism between system and evironment into a new dualism, thatbetweenphase (states) and forces, or betweenstates and dynamical laws. The states or phases constitute a description of system.Environment, on the other hand, gets an entirely different kind of description; it is described in terms of the specific recursion rule itimposes on states or phases. It is this dualism between states and dynamical laws that, more than anything else, has determined thecharacter of contemporary science. (Rosen, 1991, p. 95)

243. cf. Berkeley’s position, as in footnote 240 on page 87

89

The integration of the two discourses in classical science implied that the passage from essence toappearance (and vice versa) and the passage from meaning to symbol (and vice versa) were consideredparallel movements. ’Appear’ and ’describe’, as well as ’found’ and ’interpret’«244»becamepairwise coupled. Thus, the relation ’appear’ could become an inversion of the relation’interpret’«245».

abstract form (essence/meaning)| ↑

analyze | | synthesize(appear/ | | (found/

describe) | | interpret)| |↓ |

concrete content (appearance/symbol)

We will find an illustration of this integration in section 8.4 in terms of the ’contaminations of thenominalist discourse’ in psychiatric diagnosis. The combination of meaning and essence has in fact beenthe respectable and sensible core mode of thinking in classical science.

In classical science it was assumed that the laws of nature were made of the same stuff as their formalexpressions«246». There was no need to distinguish explicitly between the two. Whenever a theoryproved to be incorrect, it could be dismissed; but when it proved to be correct,the description wasconsidered as the description of a law of nature. The classical investigator had the conviction that if aparticular mathematical relationship was confirmed by empirical findings, then his formula was apparentlythe correct expression of ’how Nature works’. This still holds also for ’modern classical’ research untilnow, up to the whole Artificial Intelligence culture that looks for algorithms and procedures«247»according to which human intelligence works.

Theories were considered descriptions of mechanisms, and mechanisms were composed of structuralconstituents that perform each a particular function. Thus, ’how Nature works’ could be decomposed intostructural parts such as each part to have a particular function. Structure could be expressed in terms ofstates, and function in terms of dynamics, i.e.: changes of state. Hence, we find another way of sayingthat between states (structure) there are particular dynamics (functions) at work.

Once it was believed that a particular mechanism was found according to which particular phenomenawere generated, this mechanism served both as the ground for explanation and as the tool for prediction ofnew phenomena. This mechanism, then, was considered a particular Law of Nature (e.g. the second law ofthermodynamics). Such a mechanism was used as the tool for attaining the system’s next states, bycalculating its state transitions.

The mathematics that accompanied classical science were considered a language in which the variousLaws of Nature had been written and the formulations thus obtained were regarded as pertaining to thestates of the system under observation. These states were considered to be really existing. This was a

244. For a discussion of these terms see section 4.3.4, esp. the scheme on page 69. I am proposing ’synthesis’ and ’analysis’ here asgeneric terms that encompass realist and nominalist connotations. Whereas realist relations are considered to be ’really existing’, andthus of a causal (natural) variety, the nominalist relations are considered as performed by an actor, and thus of an artificial (symbolprocessing) kind.

245. together forming the pair ’expression-impression’ (cf. section 4.3.4)

246. cf. section 6.1.3, also footnote 256 on page 94

247. or their "functional equivalents"

90

major assumption that underlay the usage of mathematics in classical science«248».

In particular, the axioms of euclidean geometry were regarded as corresponding to the physical world, andfor that reason they were taken astrue. Furthermore, this truth could be extended to all theorems thatwere derivable from the axioms, and this to the degree that the truth value of all theorems was consideredto be determinable.

If a theorem is true, then this truth should also be demonstrable. Conversely, if a theorem is deduciblefrom its axioms, then its truth can be said to be a property that is "hereditary"«249»from theaxioms. In terms of the two discourses that I introduced, this would look thus:

form : content :

realist discourse: essence: appearance:’real’ truth theorem(axioms correspondto empirical reality)

nominalist discourse: meaning: symbol:proved truth theorem(theorem’s truthas proof outcome)

We thus find for classical mathematics an integration of the two discourses parallel to that of the classicalempirical sciences:

classical empirical science: classical mathematics :essence = meaning real truth = proved truth

Notice that the assignment of the truth value ’T’ to a theorem is a particular case of meaning assignmentto a symbol. The meaning ’T’ thus assigned to a proved theorem coincides with the truth of the axioms,which are taken to be in correspondence to reality. The theorem is regarded both a (realist) appearance ofthese axioms and the (nominalist) symbol to which the proved truth is assigned.

Classical physics would take a system’s state«250»both as an underlying entity (essence) and as aconceptual tool for calculations (meaning). The empirically observable phenomena are to be understood asthe products (appearances) of this essence, but also as the symbols to which the observer assigns ameaning. Thus, in classical (Newtonian) mechanics a compression tank was assumed not only to have aparticular state which is a real entity and which is expressed in the system’s observable behaviors (as

248. Cf. also:"Pure mathematics, as synthetic knowledgea priori, is only possible because it bears on none other than mere objectsof the senses, the empirical intuition (of space and time), and can be so grounded because the pure intuition is nothing but the mereform of sensibility which precedes the real appearance of objects" (Kant, 1783, §11/1971, p. 39-40)

and:"Pure mathematics, and in particular pure geometry, can only have objective reality under the condition that it bears merely on objectsof the senses, in respect of which this principle holds good: that our sensible representation is in no way a representation of things inthemselves, but only of the way they appear to us. From this it follows that the propositions of geometry are not determinations of amere creature of our poetic fantasy which could not be reliably referred to real objects, but that they hold necessarily of space and hencealso of everything that may be encountered in space, because space is nothing other than the form of all outer appearances under whichalone objects of the senses can be given to us." (Kant, 1783, §13, note I/1971, p. 43)

249. I borrow this term from Nagel & Newman, 1958, p. 51.

250. See the first lines of my quotation from Jammer (1966) quoted below in section 6.2.2, on page 99, for a definition of theconcept of ’state’ in classical mechanics.

91

appearances); but also was such a state considered a qualification of these observables (as symbols), andas such the state was a meaning assigned to them, a tool, indeed: amodel, for calculating and controllingthe system’s features«251».

To take an example, let us regard a formulation on the observation of experimentally generated pain inanimals: "Cats also gave "pain suggestive" responses to stimulation in the superior and inferior colliculi"(Milner, 1971, p. 168).

Here the laboratory animal’s pain seems to serve both as an abstract essence that is not at all immediatelyobvious to the investigator. The investigator only seems to permit himself to observe the concretebehaviors, in search of pain centers in the brain that could become manifest in particular loci of electricalstimulation. Meanwhile, it seems, the animal’s behavioral responses are studied as possible manifestationsof pain. Here, the animal’s pain is used as an abstract construct, an essence, possibly situated at the top ofone’s electrode, and the source of behavioral manifestations. Furthermore, the term ’pain suggestive’ isused in order to propose ’pain’ as a possible meaning that could be assigned to the behaviors observed."Is this to be qualified as pain?", the investigators must have been thinking, while watching the animal’sresponses. Here the meaning ’pain’ is the investigator’s way of dealing and relating with the essence’pain’. Thus, in this example, ’pain’ is used in two ways: as an essence of the behavioral phenomena, andas a meaning assigned to them.

6.1.2. The modeling relation in classical science

According to Rosen, the idea of a ’natural law’ contains the idea of a correspondence between a naturalsystem and a formal model of it. In the latter, formal inferences can be performed, that in a waycorrespond to causal relations within the natural system. Thus, a ’modeling relation’ is possible, consistingof a natural system and a formal system, between which the functions ’encoding’ and ’decoding’ hold.

A model is a formal system (F) that consists of a set of terms between which formal relations of inference(3) hold. Furthermore, a model is related to a natural system, in two ways. First, there is a function ofencoding(2); second, there is a function ofdecoding(4). The inferential relations within the model aremeant to represent causal relations (1) within the natural system of which the model is made.

A model is said tocommute(or: predictions are said to come out) if relation (1) leads to the samephenomena as (2) + (3) + (4). That is, ’encodings’ (measurements) can be made, upon which some formalinferences can be performed within the model, after which ’decodings’ (interpretations) can be made thatare indistinguishable from the outcome of some causal processes within the natural system (cf. Rosen,1991, p. 60). In this way can inferential processes within F(ormal system) be a model of causal processeswithin N(atural system):

251. Margenau discusses ’state’ as something that fits in with both discourses. The realist formulation is: "... validstatesare real, forthey are verifacts. The statep,r of a particle is realized when it functions properly in an empirical circuit of verification (...). They donot exist primarily because they are measurable, but because 1) they are satisfactory causal variables and 2) because they are linked tosensation, i.e. observation, by statable rules of correspondence." (Margenau, 1949, p. 298)

and Margenau’s nominalist formulation:"The postulate [of causality] requires that a state be a collection of observables suitably selected to insure predictability, or, morecarefully put, a state must comprise a "just sufficient" set of variables, or observables, so that it provides through them enoughinformation for rendering the solution of certain equations, called laws of nature, completely definite. This is, for example, whymomentum and position define the mechanical state of a particle, the laws of nature being Newton’s laws of motion" (ibid.)

92

The relations within F are considered to match those within N. They are parallel«252».

A prerequisite assumption in classical science was that the observer’s measurement activities didnotbasically disturb the object of attention. If some impacts of observation were unavoidable, then they wereconsidered to be due to mechanisms that wereunrelatedto the mechanisms studied; in other words:measurement was a matter different from the objects measured.

The same was true for control and manipulation: if a particular effect was looked for, e.g. shooting abullet from a cannon, then the acts of exactly positioning the physical cannon and installing its angulardirections, though of immediate influence on the bullet’s course, were considered physically independentevents, for which a separate model, disjoint from a model of the bullet’s course, could be given. Thus, toimplement the model’s computations was a matter different from applying the model. I will return to thenotions of application and applied science in section 9.1.1.

This means, in fact, that the functions ’encode’ and ’decode’ were considered to be separate and disjointparts, not interfering with the processes (3) of F, nor with those (1) of N! They could be subsumed underthe model as separate parts, representing the mechanisms of the measurement (or control) processes. Theycould be added to the system modeled, as additional parts, and a model of them could be added to themodel of the system. For instance, in section 6.1.3 on Galileo we will notice that the processes related tofriction, taking place in the experimental setups for measuring gravitation, had to be considered ascomplicating contaminants, as properties of the imperfection of the measurement configuration; not as partof the natural system to be modeled!

The causal workings of the devices for measurement and manipulation were considered to be disjoint fromthe causal workings of the object studied. Thus, given a model of a natural system, it was consideredalways possible, at least in principle, to perceive a new natural system N’ that contains the devices formeasurement and control, and to construct a new model F’ that contains descriptions of the encode (2)and decode (4) functions, so that, as a result, the remaining encode (2’) and decode (4’) functions, thosebetween the new model F’ and the natural system N’, were considered obvious and non-problematic inclassical science. No explicit understanding was required for performing them. In the terminology of ourchapter 3, they operated as tokens, as first order distinctions, through which the natural system could beobserved! In terms of James, they are the stream of consciousness’ transitive parts, the parts that remain

252. Cf.: "What can be defined in [classical] physical theory can be physically observed. And conversely, what can be physicallyobserved can be described and defined in physical theory" (Löfgren, 1992, pp. 136-7).

This remark can be appreciated best if it is added that Löfgren takes ’observation’ to be a case of ’interpretation’ (this will bedealt with more extensively in section 6.2.3), and ’definition’ to be a case of ’description’. His statement then amounts to: ’what can bedescribed in physical theory can be physically interpreted’, and vice versa. Hence, the two critical discourses, the realist and thenominalist, are considered equally powerful, if not equivalent, for the case of classical physics.

93

beyond the scope of attention while thought was performed.

The model, therefore, was a description of a set of states and their interrelations, and as such it was builtwithin a nominalist discourse; it was ameaningassigned to the empirically obtained data. But quameaning, it referred to some underlying real state behind those data; and as such the model served tomaintain realist thinking!

In fact, what happened was that, due to the triviality (in principle) of encoding and decoding relations,properties of the model were not distinguished unnecessarily from properties of the natural system.Thenatural system (essence) and a model (meaning) were considered not only to match, as in Rosen’s’modeling relation’, but in fact to coincide! The model is obscured by that which it expresses: the naturalsystem. We recognize here a phenomenon we discussed before«253» in terms of incorporation. Forit is the usage of a model as a tool for perceiving an essence behind observational data, that makesinvisible to its user that the model originally was a meaning that he assigned to his data. The formalapparatus, in classical science, instead of being a mere representation of the natural system, has obtained astatus comparable to that of the blind man’s stick, through which the researcher touches for the naturalsystem that is of his interest! The model has become his way of dealing and relating with the naturalsystem.

This, I think, is why the opposition between realist and nominalist positions became of less importance, ascalculus developed as a tool for modeling nature. Whereas before they had been vigorously incompatible,they now could become of mutual support. This was, of course, at the price of sticking to someassumptions, especially those on the possibility to include the encode/decode relations in the model, asseparate, disjoint parts.

As long as the world was believed to be made up of mechanisms (even in the case of such difficultsubject matters as mind, embryogenesis, or acculturation, etc.) then there was no need to look forproperties of nature that couldnot be decomposed into disjoint units, and expressed in terms ofmechanisms, nor to look for the limits of formal descriptions. For instance, the need was not felt to lookfor natural systems that contain a model of themselves, and to study the problems of encoding suchsystems into models, and of decoding such models«254». This need, in fact, became felt only withthe rise of self-reference, both as a formal and as an experimental problem. When the encode and decoderelations became themselves of substantial interest to researchers, the tables turned and the issues ofclassical science were no longer universal, but ’special cases’ of the issues of non-classical science.

253. section 4.2

254. This has been one of Rosen’s (e.g. 1978) ideas on the organization of living systems.

94

The shift from the realist-nominalist controversy towards a greater emphasis upon the abstract-concretedimension can be recognized par excellence in the work of Galileo (1564-1642). His majoraccomplishment is the construction of what has become a prototype of modern empiricalscience«255», one that uses "a methodology that provided a central place for mathematics in theexperimental manipulation of ordinary sensory experience" (Butts, 1978, p. 78). The realist-nominalistdichotomy does not do justice to Galileo’s position. In fact his work can be taken as a point of departurefrom this dichotomy. I will spend some words on this now, for the sake of illustration.

6.1.3. Galileo

Galileo is generally considered to be a mathematical realist, taking mathematical truths (of geometry andarithmetics) as the language in which the book of nature has been written«256». Furthermoreaccording to his realism the mathematical forms are considered to be the essences of the observablephenomena, and the latter are considered to be theimperfectappearances of those forms. Theirimperfection is due to all kinds of material impediments«257». Hence, the mathematics areconsidered as the unattainable limits of experimental purity«258».

"[Galileo] wants to know what would happen if moving bodies of different weights were droppedin a medium devoid of resistance. To find out, he tries a series of media of less and less resistanceand finds that difference in weight affects speed less and less until in the most tenuous mediumavailable, the difference is so small as to be almost unobservable. We may believe, he concludes,"by a highly probable guess that in the void all speeds would be entirely equal."

This "asymptotic" method assumes that as the "impediments" are removed, nature is found toact in a simpler and simpler way, closer and closer to the mathematically-expressed law that isbeing proposed. The law is true of Nature, even thoughin vacuofall does not occur in Nature. It istrue as a limit, departures from which (due to causal "impediments") can later be calculated andallowed for." (McMullin, 1978, p. 233)

We recognize here, in the idea of the asymptotic method, a particular encoding relation (see section 6.1.2)between a model and the physical reality it represents.

Of interest is also Galileo’s usage of the so called reasoningex suppositione, i.e. arguments that assumethe real existence of the things spoken about«259»that he is talking about. Wallace (1978, p. 127)considers this a major indication of Galileo’s realist attitude. Butts (1978, p. 83/footnote 4) compares

255. Cf.: "[Galileo’s free fall] law is simple and well-known: all physical bodies in a state of free fall (regardless of their differencesin weight, the property the Aristotelians had fastened attention upon) accelerate at the same rate in the direction of the surface of theearth. The way in which Galileo arrived at this law is a classic case of ’deducting the material hindrances’. His own experiments withrolling balls down inclined planes had shown him that according to normal commonsense observation (Aristotelian observation?) the lawof equal acceleration does not hold. The Aristotelians had concluded that the acceleration of a falling body depends upon its weight andalso upon the medium through which it is moving. Galileo’s insight consists in denying both Aristotelian points." (Butts, 1978, p. 79)

256. See for example Pitts (1992, p. 1) who starts his book on Galileo with a famous quotation in which Galileo calls the universe abook, written in the language of mathematics.

257. "An "impediment" is not (...) something which prevents, or lessens the force of, the application of mathematics to nature.Rather, it denotes a difficulty inrealizing in the complexity of the material order the simple relations of the mathematical system. But tothe extent theyare realized, mathematics can tell us what to expect." (McMullin, 1978, p. 230)

258. See also Garrison, 1986.

259. Reasoningex suppositionewas a realist tool of thought. The thing argued about is assumed to be really existing. On the otherhand, what the nominalists used was called reasoningex hypothesi; the latter means that the thing argued about is introduced only forthe sake of argument, as a useful fiction (cf. McMullin, 1978, p. 250). This is how it was also useful for purposes of calculation.Wallace (1976, p. 81) calls reasoningsex hypothesi"dialectical attempts to save the appearances - a typical instance would be thePtolemaic theories of eccentrics and epicycles."

95

Galileo’s realism with that of Descartes. Both Galileo and Descartes distinguished between the abstractsubstance of a thing and the way it appeals to our senses«260».

However, a realist understanding of Galileo is criticized by McMullin (1978, p. 234ff.) who emphasizesthat Galileo’s usage of the reasoningex suppositioneis more ambiguous. McMullin’s point is that forGalileo theex suppositionereasoning became more and more hypothetico-deductive in character. ThoughGalileo stuck to his claim that his mathematical descriptions of motion (a.o.) were in accordance withreality, he also found a truth criterion in their mathematical character and theirusefulness for predictionofnew phenomena. This, indeed, is what we would now call a model. Thus, according to McMullin (p. 223)"in Galileo’s later works, "a truly novel relationship between experience and scientific reason appeared," anew notion of explanation which "finds its true meaning in the anticipation of facts yet to be discovered,"rather than in the description of causes"«261».

Pitt (1992)«262»even calls Galileo an instrumentalist (p. 62ff.), who gives priority to bothmathematics and the observable results that could be obtained using these mathematics. Thisinstrumentalism has roots in medieval nominalist schools, such as those of the ’calculatores’ (cf. Wallace,1978, p. 124ff.), who took abstract concepts as mere tools for reasoning, first of all reasoning aboutmotion, without claiming that these notions corresponded to really existing entities or essences. ButGalileo’s conceptions are not identical to this nominalism. In fact, the realist-nominalist categorizationdoes not longer apply to Galileo:

"Thus, at the heart of Galileo’s most important scientific contribution we also find his most daringmethodological innovation. Galileo is claiming that the proof of the postulate needed in thedemonstration of the law of free fall will be the production of empirical truths. Now, if Galileowere truly convinced that the structure of the world was mathematical, it would seem appropriate,either here or later when those truths were produced, to argue for that claim using the empiricalproof of the necessary postulate as conclusive evidence. But he doesn’t. And, because he fails tomake that claim, it seems unreasonable to force any label on him" (Pitt, 1992, p. 50)

Butts gives a different emphasis:"Galileo’s confidence in mathematics and in the efficacy of experimentation must be taken as aphilosophically unjustified confidence. In part the confidence is justifiedex post factoby thepositive results of physics built upon his expectations, and such success was one of the outcomesthat Galileo trusted. But his mathematical realism had to wait for the philosophical justification hehimself had failed to provide." (Butts, 1978, p. 81)

In other words: Galileo may have believed in the realist status of his enterprise, and he may have arguedfor it, but for his factual modes of performance the question of realism versus nominalism did not matter,however he might have liked to have it otherwise. What mattered was that his formal descriptions were sosuccessful models, as to contribute to the onset of the classical tradition of science, where realist andnominalist oppositions were fading out.

The classical integration of the two discourses, however, did not hold for those domains of research whereself-reference slipped into the operations of the researcher. Here classical science attained a limit, beyondwhich essences and meanings did no longer coincide. This domain can been called non-classical science.

260. Cf. the quotation from Descartes in section 4.3.2 on page 64.

261. McMullin quotes here from M. Clavelin,The natural philosophy of Galileo.Cambridge (Mass.): MIT Press, 1974.

262. He returns partially from an earlier opinion: on an earlier occasion (Pitt, 1978, p. 193) he explicitly called Galileo ’a realist, notan instrumentalist’.

96

6.2. Non-classical science and self-reference in the researcher’s operations: therealist-nominalist controversy revisited

As paradigms of non-classical science are to be considered the works of Gödel, Bohr, and perhaps alsoWilliam James. More recent non-classical research programmes«263»can be found in Pask (e.g.1975), Rosen (e.g. 1991) and Kampis (e.g. 1991). In non-classical science the encoding and decodingfunctions cannot be added to the model as separate units, as they cannot be performed as operations thatare disjoint, in principle, from the natural system to which they are applied. Hence, it became moretroublesome to distinguish between the model and the reality modeled, as the encoding or decoding itselfbecame non-trivial and of substantial interest. The models in non-classical science were models ofencoding or decoding functions as well as of the things encoded/decoded. As a result, such models did nolonger claim to be complete; they were partial models, especially when the model pertained to how it hadbeen encoded, or to how it should be interpreted«264». Hence, the distinction between the executionof a set of formal operations (the application of a model) and its decoding (the implementation of themodel’s outcomes) were no longer as clear as they had been in classical science.

Hence, the idea that all performances of a system could be caught in strict descriptions, and eventually beprogrammed on computers in terms of calculations, lost a major claim, viz. that of its universality (the socalled Church-Turing thesis). This claim has been a leading idea in Artificial Intelligence. Rosen (1991,pp. 203-4) discusses its limitations in terms of the impossibility to describe semantic relations in terms ofmechanist (and hence: effectively computable) models (cf. also Rosen, 1962). Rosen and Kampis,however, argue for some limitations to the validity of this thesis. Rosen uses Gödel’s famous idea (cf.Nagel & Newman, 1958) of the unprovability of a particular self-referential, but true, statement. In asimilar vein, according to Rosen, it is possible that a particular materially realized (physical) systemcontains aspects of self-reference, and for that reason cannot be represented and decomposed in terms of astep by step procedure (’effective computation’). Living beings are such systems. A device that operatesby means of a program, however complicated, cannot be a proper model for a living being, which is self-referentially organized. Kampis (1991, p. 472ff) gives an extensive treatment of these problems,maintaining that "thinking and processual logic (...) can be creative and can cross all algorithmic barriersand enable humans to indeed transcend the Gödelian limits that bound computers." (p. 480)

One of the most pertinent properties of classical science had been specified as a prohibition (Whitehead &Russell, 1910-1913) of self-referential forms (e.g. paradoxes) to occur in logical expressions«265».If this rule is violated and self-reference is allowed to enter into a scientific procedure, then it becomesimpossible to keep a rigorous bookkeeping between properties of the object studied, and properties of theresearch process.

Nonclassical science took self-reference explicitly. It was no longer possible to combine realist andnominalist assumptions, but instead a renewed split arose. The pivotal characteristic of non-classicalscience is that some aspects of the object of study have become indistinguishable from the researcher’s/observer’s own operations. Furthermore, the impact of the researcher/observer upon his object of studybecomes substantial to the extent that the outcome of the research efforts are ineradicably changed bythese very efforts. I will illustrate this in terms of Gödel’s famous undecidable proposition, and in termsof some of Bohr’s ideas on complementarity.

263. Maturana & Varela (e.g. 1980) do not express their ideas in terms of models of natural systems; Maturana’s (1987a) theory ofthe observer would certainly count as non-classical, as would Varela’s work with Thomson and Rosch (1991). Varela’s work incognitive science, on the other hand (e.g. 1989), tends more to classical approaches.

264. Cf. Lögren, 1990; see also section 6.2.3.It is the impossibility to construct a complete formal (mechanist) model of the circularity of the living organization, that brings Rosen topropose a comparison with semantics:

"In fact, it is not too far wrong to say that an organism (...) is itself like a little natural language, possessing semantic modes ofentailment not present in any formal piece of it that we pull out and study syntactically." (1991, p. 248)

265. Russell is said to have been delighted when the prohibition of self-referentiality was finally proved to have been incorrect(Spencer Brown, 1972).

97

6.2.1. Gödel’s formal undecidability: the proof process mixes up with the contents of the proof

Hofstadter (1979, pp. 420ff.) discusses Cantor’s so called ’diagonal argument’, which has been used toprove the impossibility to assign a unique positive integer number to each element of the set of realnumbers. In classical mathematics, the mathematician’s procedure itself would remain beyond the scope oftheorizing, but in Cantor’s diagonal argument this was no longer the case. Now the act of encodingbecame substantial.

One of the outcomes of the crisis of mathematics that developed during the turn of the 19th century, wasthat the blend of the two discourses was made undone. Among other things, this was an impact of thediscovery of the so called non-euclidean geometries. Once it was accepted that the axioms of amathematical system did not necessarily correspond to nature, their truth was no longer a matter ofcorrespondence, but rather one of arbitrary choice. Formal axiomatic systems were nowprimarilyconsidered as a matter of definition of terms, not as descriptions of Nature, and the definitions of axiomsand other mathematical terms became regarded primarily as pertaining to their use in calculations, not as areference to something in empirical Nature. This culminated in the quest for a complete formalization ofmathematics, so that all mathematical truths were to be made mathematically provable. As a result, it wasattempted that mathematics was to lose all its apparent reference to any reality, and to gain an absoluterigor. This was David Hilbert’s programme.

Instead of the correspondence question, the formalization of mathematics required that the axioms of aparticular mathematical system, if not true in themselves due to a correspondence to the physical world,were at least consistent with one another«266». This, in fact, led to a search for criteria to decidethe consistence of systems of axioms, and it is with respect to this search that we find Gödel’s famousproof of the necessary incompleteness of axiomatic systems in the field of arithmetics. His proof amountsto the rejection of the idea that a system of axioms can be proved to be complete. This is to say that forany axiomatic system of arithmetics there will be at least one true proposition that, though formulatedaccording to the proper syntax, cannot be proved true or false within this axiomatic system. It is called’undecidable’. A statement that maintains its own undecidability, then, may be true yet (and: therefore)undecidable. The truth of such a statement is beyond the domain of formalization of the axiomatic system.

The crucial property of such an uncanny proposition is that it pertains to its own provability. This is whatthe famous sentence constructed by Gödel says«267». Here the predicate of provability itself ismapped, in an intricate way, to the contents of the sentence to be proved. As a result, a paradox isconstructed which effectively impedes the assignment of a truth value to the sentence (and this isprecisely what the sentence is maintaining). Henceforth it was considered impossible to attain the rigor ofcomplete formalization in mathematics. The question remained: if the truth of the sentence cannot beformalized, then what is it?

As we are faced with the existence of true propositions that cannot be assigned a truth value within anaxiomatic system, we are brought to the core of the matter: a split has emerged between what is true withrespect to an axiomatic system on the one hand, and on the other hand what can be proved within it. Inthe words of Wang (1988):

"I believe that [Gödel’s] idea is that his [incompleteness] theorem presses upon us the need for’assumptions essentially transcending arithmetic’ even for solving problems in arithmetic, just aswe use physical laws that transcend sense perception. G[ödel] then says that, of course, under thesecircumstances mathematics may lose a good deal of its ’absolute certainty’" (p. 315)and:"The contrast between formal systems and the axiomatic method (as G appears to understand them)may be seen as a fundamental aspect of what might be called G’s dialectic of the formal and the

266. cf. Nagel & Newman (1958, p. 14)

267. Lucid discussions of his famous 1931 paper can be found in Nagel & Newman (1958) and Hofstadter (1979).

98

intuitive.(...)His incompleteness theorems are a spectacular result of his application of the dialectic between theintuitive concept of arithmetic truth and the more precise concept of formal provability as revealedby concrete instances of formal systems. (...) The dialectic is especially transparent in the processof G’s discovery. He first saw that arithmetic truth is not definable in arithmetic. Then he noticedthat provability (in a formal system) is definable, and constructed a proposition that is true andexpressible in the system, but not provable in it." (p. 201)

In classical mathematics this difference did not exist. But now that a split between ’real’ truth and provedtruth has emerged, the former does not, as before, pertain to empirical reality. Instead, it is oftenconsidered to pertain to some ideal, platonic, reality«268».

For our present purposes it is relevant to notice that a major element of Gödel’s proof was the translationof meta-mathematical statements (such as about the consistency of a particular axiomatic system) intonumber theoretical propositions. In other words, it was the formulation of statements which were aboutnumbertheory, in terms of propositions aboutnumbers. In the case of Gödel’s proof, it was a statementabout the absence of a particular proof process itself that was reformulated in terms of numbers. Thus, astatement about a proof process became treated as if it were ’just’ a matter of numbers, and the contentsof the proof were mixed up with a numerical translation (the so called ’Gödel-numbering’) of the proofprocess itself.

The outcome of the proof process is that the sentence constructed is not decidable. This is precisely whatthe sentence itself says. Its truth, therefore, exists as the impossibility to prove it«269»within theconfines of the axiomatic system in which it has been formulated! It is this circumstance that elevates thetruth of Gödel’s sentence above the truth qua provability that exists within an axiomatic system. The truthof the Gödel sentence concerns the proof process itself, and hence the computational acts of themathematician. Thus the sentence describes an act of assignment of a meaning (’undecidable’), such thatamong the symbols to which this meaning is assigned, the very meaning to be assigned has been encoded.Indeed, a divergence has occurred between a ’real’ (platonic) truth (viz. the incompleteness of axiomaticsystems and the undecidability of self-referential Gödel-sentences) and a proved (formalized) truth (suchas the truth of other, non-selfreferential theorems within an axiomatic system). Notice also that thedifference between the proof process and its subject matter has become less clear.

As has been said above, a ’real’ truth is something that is maintained within a realist discourse; a provedtruth occurs in a nominalist discourse, where it can be assigned to a theorem as a meaning to a symbol.The self-reference of Gödel’s sentence, therefore, is one that occurs within a nominalist discourse. Thequalification ’undecidable’ is assigned to this sentence as a final meaning, after attempts to prove it trueor false have failed. More precisely, the qualification ’undecidable’ pertains to these failures. Hence, itstruth is that of the necessity of failure of proving this sentence within the axiomatic system within whichit has been formulated. This truth, therefore, pertains to the limits of provability. Though Gödelconsidered this as a ’real’ truth, we may notice at this place that it does pertain to (the limits of) themathematician’s operations. It is the self-reference of these operations that turns out to be self-defeating.

268. Cf.: "Classes and concepts may, however, also be conceived as real objects, namely classes as "pluralities of things" or asstructures consisting of a plurality of things and concepts as the properties and relations of things existing independently of ourdefinitions and constructions.

It seems to me that the assumption of such objects is quite as legitimate as the assumption of physical bodies and there is quiteas much reason to believe in their existence. They are in the same sense necessary to obtain a satisfactory system of mathematics asphysical bodies are necessary for a satisfactory theory of our sense perceptions and in both cases it is impossible to interpret thepropositions one wants to assert about these entities as propositions about the "data", i.e., in the latter case the actually occurring senseperceptions." (Gödel, 1990 (1944), p. 128)

Wang (1988, p. 188) refers to Gödel’s position as a ’conceptual realism’.

269. shadows of a ’via negativa’, cf. chapter 7

99

More about the limitations of self-referential operations will be said in chapter 7«270». First I willturn now to a comparable situation of self-referential operations which can be found in the domain ofmeasurement of quantum phenomena.

6.2.2. Bohr’s understanding of complementarity in quantum mechanics: the measurement processmixes up with the contents of the measurement

Bohr’s commitment was to find a new conceptual framework that could account for the anomalies heencountered within the framework of classical physics. His ’principle of complementarity’ was introducedto this purpose. His (e.g. 1934) topic of concern was that during quantum measurements an interactionoccurred, inevitably and necessarily, between the measuring device itself and its object. The measuredproperties of this object then could no longer be conceived as existing irrespective (and independent) ofthe measuring device.

Though it is possible to define a quantum particle’s development in time (e.g. according to the so calledSchrödinger wave equation«271») and describe its course in terms of causally determined changes,this course of the particle remains inaccessible to empirical observation. We can either (obtrusively)observe the system’s spatiotemporal state, or we can describe (unobtrusively) the system’s abstract causaldevelopment. But we cannot integrate the two into one coherent representation.

Jammer summarizes Bohr’s position on complementarity:"The "state of a system," as understood, for example, in ordinary mechanics, is the set of all the(position) coordinates and the momenta of the constituent parts and implies the possibility of usingthese data for the prediction (or retrodiction) of the structural properties of the system. Prediction(or retrodiction), however, is possible only if the system is closed, that is, unaffected by externaldisturbances. For an open system, strictly speaking, no "state" can be defined. Now, the evolutionof the structural properties of the system is governed by the law of causality; their progress in timeis the causal behavior of the system. But according to the quantum postulate, in atomic physics,any observation of the system implies a disturbance. In other words, a system, if observed, isalways an open system. A space-time description, however, presupposes observation. Hence,concluded Bohr, the claim of causality excludes spatiotemporal description and vice versa. Theusual causal space-time description, that is, the simultaneous use of complementary descriptions, ismade possible in classical physics, argued Bohr, merely because of the extremely small value ofthe quantum of action "as compared to the actions involved in ordinary sense perceptions.""(Jammer, 1966, p. 352)

Thus, what we could measure, according to Bohr, was no longer indicative of the (conceived) state of thesystem before measurement! Clearly, such a situation does not fit with the assumption, held by classicalscience, that a system’s state qua essence coincides with its state qua meaning«272». We meet here,therefore, a renewed divergence of the two discourses. We recognize this divergence in Bohr’s oppositionof ’definition’ and ’observation’:

"On one hand, the definition of the state of a physical system, as ordinarily understood, claims theelimination of all external disturbances. But in that case, according to the quantum postulate, anyobservation will be impossible, and, above all, the concepts of space and time lose their immediatesense. On the other hand, if in order to make observation possible we permit certain interactionswith suitable agencies of measurement, not belonging to the system, an unambiguous definition of

270. The reader may be reminded of the distinction I made in section 4.3.4 between realist relations ’appear’ and ’found’ versusnominalist relations ’interpret’ and ’describe’. The latter were associated with the actions of an observer, the former not. The self-reference of Gödel’s sentence is one of nominalist relations.

271. Though Bohr (1928, p. 589) does mention Schrödinger’s work, the reference to his wave equation has become conventionalsince von Neumann (1932, p. 5).

272. See section 6.1.2, esp. footnote 252.

100

the state of the system is naturally no longer possible, and there can be no question of causality inthe ordinary sense of the word. The very nature of the quantum theory thus forces us to regard thespace-time co-ordination and the claim of causality, the union of which characterises the classicaltheories, as complementary, but exclusive features of the description, symbolising the idealisationof observation and definition respectively." (Bohr, 1928, p. 580/Collected Works VI, p. 148)

The realist discourse is the one in which causal laws are expressible that are valid for the undisturbedobject, the nominalist discourse is the one in which spatiotemporal measurements are obtained, those thatcan be used for (statistical) prediction and control«273», but which do not allow a precise accountof the causal development of individual quantum objects«274».

What impedes the possibility of an all-encompassing formulation in line with the classical solution, is thetransition from the causally determined flow of events to the a-causal situation during and after themeasurement interaction. This transition itself cannot be measured, whereas, furthermore, eachmeasurement disturbs the object measured in an unpredictable way. It is precisely at the encounterbetween the measured system and the measurement device that a discontinuity is created in the quantumsystem’s state (a so-called ’collapse of the wave function’)«275». This discontinuity is whatseparates here the realist from the nominalist discourse.

Thus, we can compare mathematics and physics with respect to the split between realist and nominalistdiscourses:

mathematics physics

realist discourse:essence ’real’ truth’ unmeasured state,

causally developing

nominalist discourse:meaning proved truth spatiotemporally describable

state, tool for statisticalprediction

The crux of the matter is that it is impossible to measure the things that happen when a measurementdevice interacts with a quantum object and disturbs it. Dependent on the measurement arrangement used,and dependent of features of the measurement interaction itself (that cannot be measured) the wavefunction collapses in one way or other. Different measurement arrangements lead to measurement results

273. i.e., in terms of chance distributions of measured events

274. A modern elaboration of this has been given by Pattee. This author concludes from Bohr’s results that complex phenomenarequire an approach in which two complementary languages are required, one concerned with (dynamic, physical) laws and one with(static, symbolic) constraints or (computational) rules. Laws and constraints, according to Pattee, can never be reduced to one another,and together they constitute two complementary modes of description of empirical reality:

"In order to justify the need for a complementary mode of description we must look for significant situations where one modeof description fails, and fails in such a fundamental way that it cannot be patched up by adjustments of parameters or functional forms.Such situations are well known in dynamical descriptions. They are called singularities by mathematicians and instabilities byphysicists." (Pattee, 1978, pp. 198-9)

Furthermore, according to Pattee, a measurement is to be seen as a transition between these two modes of description: "Since(...) both laws and computation are universal, there appears to be an ambiguity in where lawful description should end andcomputational description should begin. Measurement is, by definition, the process that separates the two" (Pattee, 1992, p. 183)

Notice that measurement, thus understood, occurs at the transition from a dynamic to a symbolic domain, not at the reversetransition. According to Pattee, Bohr wanted to explain the measurement process in uniquely causal terms, and hence was faced with aninfinite regress, whereas in fact he needed a mode of description that was complementary to the language of causal (dynamic) processes(cf. Pattee, 1992, p. 183).

It is interesting to notice that to Pattee’s two complementary modes of description (’laws’ and ’constraints’) correspond to therealist and nominalist discourses respectively. The former mode enables explanations; the latter enables descriptions in terms ofinformation processing.

275. During this transition the system’s state parameters collapse from complex values to real values, cf. Penrose (1989, p. 250)

101

that cannot be combined into one representation. The properties measured are not compatible (such as:wave and particle) or cannot be measured simultaneously (such as: position and momentum). As a way ofdealing with this uncommon situation, Bohr proposed to take into account how an experimentalphenomenon had been obtained«276», and, furthermore, "that we must, in general, be prepared toaccept the fact that a complete elucidation of one and the same object may require diverse points of viewwhich defy a unique description."«277». This conception of complementarity is called by vonWeizsäcker (1955, p. 522) ’parallel complementarity’.

The compilation of these diverse points of view is a task within the realist discourse, which is fit forreconciliating the diverging varieties of appearance with respect of one single underlying entity. In thisdiscourse Bohr’s term ’definition of the state of the system’«278» is understood as pertaining to theessence behind the empirical phenomena, and as such it aims to provide a description of the ’pure’,undisturbed state of the system; a description, that is, of a side of the system that cannot be approachedempirically.

It is interesting to make here a comparison with Galileo’s idea of mathematics as describing a reality thatis freed from the material ’impediments’ that make up the imperfections of physical phenomena. Thisreality, according to Galileo, can only be approximated in experimental arrangements«279», but notreally attained in its purity. There is, however, one important difference between Bohr and Galileo. As wesaw, the issue of realism versus nominalism (the reality of the pure mathematical forms) was for Galileo’sscientific practice not too vital. His experimental approximation of the pure forms was considered a matterof degree, and the degree he needed was ruled by the prospects of prediction. In the quantum situation, onthe other hand, empirical approximations of the pure form were no longer thinkable, since the pure formwas severely disturbed by the measurements. Here the realist versus nominalist controversy regained itsvirulence. For what is it the mathematical notions (such as those of Schrödinger’s wave equation) referto? Were they really existing entities, and if so, how should we understand the difference betweencomplex (and imaginary) values and real values?

A strong case in favor of a realist interpretation of Bohr’s position has been given by Folse (1992, p. 52).Bohr’s way of thinking, according to Folse, remains compatible with a kantian framework, in which theexistence of the ’object behind the phenomena’ is also presupposed«280». This in itselfunperceivable object is regarded the subject matter of what Bohr calls ’definition’, whereas ’observation’pertains to a mode of description in terms of the empirically accessible spatiotemporal properties. Butthere is an ambiguity in Bohr’s usage of the term ’definition’. For it may mean either that its subjectmatter is a kind of thing in itself, the essence of the phenomena observed, or that it is a concept, not areal thing. This is the old opposition between real and nominal definitions«281».

276. Cf.: "While, within the scope of classical physics, the interaction between object and apparatus can be neglected or, ifnecessary, compensated for, in quantum physics this interaction forms an inseparable part of the phenomenon. Accordingly, theunambiguous account of proper quantum phenomena must, in principle, include a description of all the relevant features of theexperimental arrangement." (Bohr, 1963, p. 4, as quoted by Folse, 1992, p. 61).

277. Bohr, 1934, p. 96/Collected Works VI, p. 212. Because the English text of 1934 is, qua translation, considered not entirelysuccessful (Kalckar, 1985, p. 195), I include here and at other places, when available, a German translation additional to the citationsfrom the English version:

"Wir müssen im allgemeinen darauf gefaßt sein, dass eine allseitige Beleuchtung eines und desselben Gegenstandesverschiedene Gesichtspunkte verlangen kann, die eine eindeutige Beschreibung verhindern" (Bohr, Collected Works VI, p. 205)

278. Bohr, 1928; see the quotation from Bohr on page 100 above.

279. cf. section 6.1.3

280. In particular, Folse (1992, p. 59) points at the following passage: "We are aware even of phenomena which with certainty maybe assumed to arise from the action of a single atom, or even of a part of an atom. However, at the same time as every doubt regardingthe reality of atoms has been removed and as we have gained a detailed knowledge even of the inner structure of atoms, we have beenreminded in an instructive manner of the natural limitation of our forms of perception." (Bohr, 1934, p. 103)

281. ’definitio realis’ and ’definitio nominalis’ in scholastics

102

On the other hand, if the subject matter of a definition is regarded as an abstract concept, then we findourselves in a nominalist discourse, which deals with the interpretation of symbols. In fact, Bohr oftenexpresses himself in terms of the empirical description of experiences as well as in terms of the orderingof these experiences«282». Indeed, there are in Bohr’s writings many passages in which he takes afirm nominalist position«283».

It is the language of classical physics that he considers to be necessary to order the experimentallyobtained quantum phenomena, but he also considers the language of classical physics to be basicallyinsufficient to this purpose«284». For example, the quantum objects are qualified by states in acomplex space, i.e. complex numbers are used for the description of these states. The classical frameworkdoes not permit a clear representation of these states. More important, however, is the fact that theclassical framework does not permit a description of the interaction processes between the measuredobject and the measurement device. Bohr was particularly interested in these interaction processes, whichhe would have liked to formulate in the conceptual framework of classical science (i.e.: using the classicalcombination of causality and spatiotemporal descriptions).

Hence, when Bohr found himself challenged by an impossibility to adequately measure the interactionbetween object and measurement device and account for it in the classical (causal plus spatiotemporal)terms, he took this as evidence for the necessity of finding a proper conceptual framework for interpretingthe new empirically obtained experiences«285». It was thus that Bohr was concerned with "aninvestigation of the conditions for the proper use of our conceptual means of expression"«286».That is, given his quest for measuring the interaction processes during measurement, his emphasis shiftedfrom a causal description of these processes to the conditions for understanding them in causal terms. Theproblem ’why can the causal processes that take place during measurement not be measured?’ changedinto: ’why are our conceptual tools inadequate for understanding quantum measurements in causal terms?’.The fact that the empirically obtained spatiotemporal descriptions did not fit in with the conceptual(causal) frameworks, which the observer has already available, triggered a rethinking of theseframeworks«287». I will return to this change of emphasis, from the object studied to theconceptual framework in section 7.2, when discussing the limits of observing self-reference. For thepresent moment it suffices to notice that, as a result of his efforts, Bohr turned his attention towards hisinstruments, i.c. his tools for conceptualization. This, indeed, can be seen as a transition opposite to thatof incorporation. We might call it ’excorporation’.

282. Bohr, 1934, p. 92/Collected Works VI, p. 208 (German: "Einordnung der Erfahrungen"; Collected Works VI, p. 203)

283. E.g.: "... a subsequent measurement to a certain degree deprives the information given by a previous measurement of itssignificance for predicting the future course of the phenomena. Obviously, these facts not only set a limit to theextentof theinformation obtainable by measurements, but they also set a limit to themeaningwhich we may attribute to such information. We meethere in a new light the old truth that in our description of nature the purpose is not to disclose the real essence of the phenomena butonly to track down, so far as it is possible, relations between the manifold aspects of our experience." (Bohr, 1934, p. 18/Collectedworks VI, p. 296)

284. "... it is decisive to recognize that,however far the phenomena transcend the scope of classical physical explanation, theaccount of all evidence must be expressed in classical terms. The argument is simply that by the word "experiment" we refer to asituation where we can tell others what we have learned and that, therefore, the account of the experimental arrangement and of theresults of the observations must be expressed in unambiguous language with suitable application of the terminology of classicalphysics." (Bohr, 1949, as quoted by Holton, 1970, p. 1018)

and:"... only with the help of classical ideas is it possible to ascribe an unambiguous meaning to the results of observation." (Bohr, 1934,p. 17/Collected Works VI, p. 295)

285. "The main point to realize is that all knowledge presents itself within a conceptual framework adapted to account for previousexperience and that any such frame may prove too narrow to comprehend new experiences." (Bohr, 1958, p. 67; as quoted by Honner,1982, p. 25)

286. Bohr (1958, p. 2); quotation cited by Honner (1982, p. 24)

287. See Honner (1982, p. 23) on Bohr’s interest in the "general limits (or conditions) for the possibility of forming unambiguousconcepts."

103

Bohr’s thinking led to a fully nominalist discourse! If the term ’definition’ pertains to a concept, then it isconcerned with the way in which this concept can be used for interpreting empirical experiences andobservations«288». A definition of a concept, then, becomes a prescription of its usage whenempirical data are to be interpreted.

The conception of complementarity that is implied here, is different from the ’parallel complementarity’.Here the scientist’s own conceptual framework is of focal interest. Von Weizsäcker (1955, p. 524) callsthis ’circular complementarity’, and he considers it to be Bohr’sown conception of complementarity. Themajor difference with parallel complementarity is that two modes of description are not, as in parallelcomplementarity, considered mutually complementing, but that one mode of description is regarded asinstrumental for making descriptions in the other mode. Thus, the best example: through thinking incausal terms we attempt to interpret the spatiotemporal descriptions of our empirical observations.

Bohr recognizes a tension between the definition of a concept (such as causality) and its actual usage fordealing with empirical information. The usage of a concept impedes a full detachment from it, and, on theother hand, the analysis of a concept (and the act of defining it) impedes fully using it for observations inthe mean time«289». The conceptual tools that we use as instruments for ordering our observationscannot be analyzed while being used. When analyzed, there are other conceptual tools required, whichescape being analyzed at that very same time.

There is some evidence«290»that, as to these ideas on the limitations of our conceptualframeworks, Bohr has been inspired by the ideas of William James on the stream of consciousness andthe impossibility of introspectively observing its transitive parts without turning them into substantiveparts«291». Indeed, we find ourselves back on familiar ground: the tension between the usage of a(conceptual) instrument and its analysis, which is the issue of Bohr’s circular complementarity, is similar(and probably inspired by) William James’ relation between substantive and transitive parts, and it returnsat several places in his work in terms of the relation between subject and object«292».

The following comparison can be made between some of James’ and Bohr’s core notions:

288. A concept, thus understood, is a dynamic process, to which the term ’interpretation’ would also be appropriate. It is thisdynamic conception of ’concept’ that has been elaborated extensively by Pask in his monumental conversation theory (e.g. Pask, 1975).I leave this vast topic beyond the scope of the present text, though I do recognize that many issues presented here could be unfolded interms of conversation theoretical notions.

289. "... the nature of our consciousness brings about a complementary relationship, in all domains of knowledge, between theanalysis of a concept and its immediate application" (Bohr, 1934, p. 20/Collected Works VI, p. 298)

290. Von Weizsäcker (1955, p. 525) recalls as a personal experience with Bohr that "the only occidental philosophers whom I heardoften and emphatically being cited by Bohr, were Socrates and William James". Folse (1992) suggests a more indirect connection viathe Danish philosopher Ho/ ffding. See also Holton (1970, p. 1031ff.) and Jammer (1966, pp. 178/9) for discussions of the relationbetween Bohr and James.

291. "The unavoidable influence on atomic phenomena caused by observing them here corresponds to the well-known change of thetinge of the psychological experiences which accompanies any direction of the attention to one of their various elements." (Bohr, 1934,p. 100/Collected Works, VI, p. 216) ("Die unvermeidbare Beeinflussung der atomaren Erscheinungen durch deren Beobachtungentspricht hier der wohlbekannten Änderung der Färbung des psychischen Geschehens, welche jede Lenkung der Aufmerksamkeit aufihre verschiedenen Elemente begleitet." (Collected Works VI, p. 206))

and:"... the unavoidable influencing by introspection of all psychical experience, that is characterized by the feeling of volition, shows astriking similarity to the conditions responsible for the failure of causality in the analysis of atomic phenomena." (1934, p. 23-4/Collected Works VI, p. 301-2)

292. E.g. Bohr (1928, p. 590).

104

James: Bohr:

substantive parts concepts (subject matter of definitions)transitive parts use of concepts (for ordering experiences)

By using an instrument, Bohr maintains, one cannot perceive that instrument. As an example Bohr (1934,p. 99) speaks of a stick that, when grasped firmly, can be used as an instrument for touching objects inthe environment. The tactile sensations in the hand then escape from attention, and instead the distal edgeof the stick takes the quality of a tactile organ. It is there, at this distal site, where the person observes theobject he is touching with his stick; no longer is the palm of the hand the boundary of the person as asensing unity, but the edge of the stick. Conversely, if the stick is held loosely, it cannot be used as aninstrument of touch, and it appears to the observer as a stick, i.e., as an independent object, sensed in thehand. In fact I discussed this phenomenon already in section 4.2 in terms of ’incorporation’«293»,and in section 2.3.4 , esp. footnote 65, page 36, one the usage of words as instruments to bring forthobjects. Likewise, other tools, such as concepts, can be taken as an object of study only when ’holding itloosely’, i.e. at the price of not using the concept for arranging empirical data. The complementaritybetween observation and definition which Bohr mentions several times in his Como paper (1928), thenarises as one between the usage of the instrument versus its analysis. This is entirely a nominalist affair,concerned with the study of the limits of arranging empirical data.

To finish this section, let me quote from Honner (1982, p. 22), who gives a bright remark on Bohr’sposition:

"In my view, then, the following uncompromising schema for Bohr’s underlying train of thoughtmust be adopted. First, the situation in quantum physics and the situation with respect to thediscussion of consciousnessboth entail similar instances of the one argument. Secondly, this oneargument rests on the analysis of the conditions for the possibility of unambiguous communicationabout processes at and beyond the boundaries of human experience."

In chapters 10 and 11 we will find opportunity to discuss comparable ’similar instances’, when discussinghow a psychotherapist may develop a particular self-experience, when involved in particular encounterswith patients.

6.2.3. Löfgren’s ’linguistic complementarity’

It is the mixture, within the nominalist discourse, of process (proof process or measurement process) withthe outcome of the process (proof or measurement) that is the locus of confusion. Löfgren has focusedupon this.

Löfgren, from a point of view of mathematical automata theory, elaborates the thesis that within alanguage (e.g. a computer language) it is impossible to give a complete description of the way in whichthe sentences of the language are to be interpreted«294». He uses the term ’linguisticcomplementarity’ for this impossibility. The concept of linguistic complementarity comprises that it isimpossible for any language, provided it is sufficiently rich, to contain such a complete description of itsown interpretation processes, i.e. a description of the processes and functions that convert the sentences ofthis language into fully operative interpretations (e.g.: running computer programs or calculation

293. It is interesting to notice that Merleau-Ponty (1945, p. 167, see section 4.2 above) used a similar metaphor. Both for Bohr andfor Merleau-Ponty this metaphor illustrates that we cannot simultaneously pay attention both to an object and to the method by which,or the instrument through which, it is perceived.

294. ’Interpretation’ and ’description’ as defined in section 4.3.4.

105

processes)«295».

interpretation^ || |

interpretation | | descriptionprocesses | | processes

| vdescription

If a description would exist that specifies theway in which it should be interpreted, then theinterpretationof this description would coincide with theinterpretation processrequired for it.

Equivalent to this is to say that, within a formal system, the meaning (interpretation) of a descriptioncannot coincide with the interpretation processes. In a functioning system (computer), the interpretationprocesses arenot interpretations of the description (e.g. a program in Pascal), but they are theinterpretations of a description in a meta-language (e.g. a description of the operating system, given inmachine language).

Now this idea is applied by Löfgren to Bohr’s circular complementarity. In particular, Löfgren maintainsthat there is a self-reference implied in the idea of measuring the interactions by the measuring device andthe object to be measured. In order to know the object as it was before measurement, it would benecessary to know in detail the precise ways in which this object interacts with the measurementinstrument and how it has been influenced. Precisely this is impossible, as it would require a measurementof the very measurement process, which would entail an infinite regress«296». In order to avoidthat, a self-referential measurement would be required, one that registers its own process as it takes place.But this would get stuck on (circular) complementarity: we cannot observe the instrument while using itfor our observations. Let us note, in passing, that such self-referential ’measurement’ is only possible forimmediate experiences, at a more elementary level than object perception, and at the price of losing theobject qua perceivable entity. I will discuss that in chapter 7.

The impossibility of self-measurement, Löfgren maintains, is not due to a limitation of classical physics orits causal language. It is a feature of languages in general that self-reference cannot be expressed fully (cf.Löfgren, 1990). His major step, then, is toreduceBohr’s ideas on (circular) complementarity to hisconcept of linguistic complementarity.

Löfgren points at a similarity between the works of Gödel and those of Bohr. Both authors wereconcerned with self-reference: Gödel with self-reference of arithmetic propositions, Bohr with self-reference of measurements. In quantum physics, unlike in classical physics, there is a limitation to whatcan be observed, comparable to the limitations of what can be proved in mathematics.

295. As an example of this we may think of a computer program written in a particular language, say Pascal. It is then impossibleto have the program also contain a set of (Pascal) instructions for the computer as to how to interpret the Pascal statements intocomputer actions. A program for reading and executing Pascal statements (either a so called ’interpreter’ or a ’compiler’) is presupposedto be already installed on the computer before it can transform the Pascal lines into actions. This auxiliary program that performs thetransformation (Löfgren calls this theinterpretation process) should itself also be written in a language already available on thecomputer (such as machine language). The computer should ’know’ the language before it can process a program written in it; theprogram itself cannot provide its own ad hoc interpreter or compiler.

Any (fancy) program that would accomplish that, would be self-referential in that it says how to interpret its own lines,including the lines that prescribe the interpretation process!

See also Löfgren (1992, p. 116).

296. cf. Löfgren (1989, p. 302ff.); Pattee (1992, p. 183)

106

self-reference of operations in:

mathematics physics

undecidable proof process unmeasurable measurement process

Both authors also hit upon the impossibility to arrive at a particular goal: the incompleteness offormalization of arithmetics and the incompleteness of description of quantum measurements. Löfgren(1988, p. 144) claims that the linguistic complementarity is the crucial link between these domains ofmathematics and physics, since it pertains to all languages (and hence is valid both for physics andmathematics).

Löfgren’s work, I think, is to be considered the most parsimonious elaboration of this notion ofincompleteness that is present both in the work of Gödel and of Bohr. I will use Löfgren’s insights,especially his concept of ’linguistic complementarity’, in the next chapter, when describing the limitationsof realist and nominalist discourses (section 7.1, especially 7.1.2). The linguistic complementarity is to beunderstood as the ultimate formulation of self-reference, within the confines of a nominalist discourse.

To end this section, I will make some terminological comments that can easily be skipped for the courseof my argument.

6.2.3.1. Remarks on terminology

Löfgren (e.g. 1992, pp. 145-148) translates in terms of this linguistic complementarity the quantummechanical measurement problems, especially the impossibility of self-measurement«297».

It is remarkable that Löfgren, in order to justify this step, understands measurement not as somethingapplied to physical reality, but todescriptionsof physical reality! Measurement and observation, then, arereduced by Löfgren to (linguistic) operations ofinterpretationof these descriptions«298». Anadequate theory of quantum physics, according to Löfgren, should contain a description of how themeasurements should be performed. But if measurements are reducible to interpretations, then such atheory is reducible to a description of how it should be interpreted«299». The impossible case ofself-measurement in quantum mechanics is thus presented by Löfgren as reducible to the impossible caseof formulating a sentence that describes its own interpretation processes.

It is thus that Löfgren de facto imposes a nominalist discourse upon Bohr’s ideas on complementarity,where a more natural way to recognize a nominalist discourse in Bohr is overlooked. Löfgren expressesBohr’s tension between ’observation’ and ’definition’ in his own terms, i.e. as a tension between’interpretation’ and ’description’ respectively. This is somewhat curious, since, as we saw, Bohr’s notionof ’definition’, being concerned with the arrangement of experiences, is a more likely candidate to betranslated as Löfgren’s ’interpretation’; likewise, ’observation’, which is for Bohr a matter of

297. "It is in recognizing the quantum mechanical measurement problem as self-referential, with the measurement process, asinterpretation process, itself belonging to the quantum physical domain, that we encounter a phenomenon of language for which thelinguistic complementarity takes the form of Bohr’s primary view." (Löfgren, 1992, p. 145) [’primary view’: i.e., that of Bohr’s Comopaper, 1928]

298. "... that observation by measurement is how to interpret the state symbols of the formalism" (Löfgren, 1992, p. 148). Cf. alsoLöfgren, 1992, p. 136; 1991, p. 64; 1989, p. 298.

299. "The measurement process is in quantum mechanics part of the quantum mechanical domain. Thus, with measurement regardedas interpretation, a description of quantum mechanics must contain a description also of how it is to be interpreted." (Löfgren, 1989,p. 310)

107

administrating the data, might fit in Löfgren’s notion of ’description’«300».

Bohr’s concepts: Löfgren’s concepts: my understanding:

definition description interpretationobservation interpretation description

The usage of ’interpretation’ and ’description’ that I am proposing is more in line with the terminology ofPattee. His (e.g. 1992) concept of ’measurement’ is comparable to what Löfgren calls ’descriptionprocess’. Furthermore, for Pattee measurement is an operation in which a time dependent (dynamic)process is transposed to a time independent (static) description. For Löfgren, to the contrary, measurementis an operation through which a static description comes to time dependent processes (the ’interpretation’,’meaning’ or ’model’)«301».

Instead of Löfgren’s reduction of self-measurement to self-interpretation, one might venture to think, itcould make better sense to translate the idea of impossible self-measurement as an impossible self-description, i.e. as the impossibility to describe in a description the very description processes required forwriting down that description. Likewise, the impossibility to write down the interpretation process that isrequired for interpreting the very same description, parallels the impossibility to analyze a conceptual tool(a definition, cf. section 6.2.2) while using it; it also resembles the impossibility to think the (transitive)processes that lead to a (substantive) thought. The terminology Löfgren uses in order to reduce Bohr’sconcept of (circular) complementarity to his own concept of linguistic complementarity, therefore, seemsto be not maximally plausible, even though this reduction itself, I think, is very valuable for recognizingthe nominalist discourse in Bohr’s thoughts.

Finally, we may extend the comparison between the concepts of James and Bohr (see the diagram on page104) with those of Löfgren:

James:substantive parts(empirical Me)

transitive parts(judging Thought, I)

empirical findings

Bohr:concepts(subject matter ofdefinitions)

use of concepts (forordering experiences)

empirical phenomena

Löfgren:interpretations(e.g. automaticcalculation)

interpretationprocesses

descriptions (e.g.computer program)

300. However, it must be acknowledged that Bohr’s usage of the term ’definition’ is not always consistent. For example, in thefollowing fragment this term is used as ’description’ or ’measurement’:

"The reciprocal uncertainty which always affects the values of these quantities is, as will be clear from the preceding analysis,essentially an outcome of the limited accuracy with which changes in energy and momentum can be defined, when the wave-fields usedfor the determination of the space-time co-ordinates of the particle are sufficiently small." (Bohr, 1928, p. 583/Collected Works VI,p. 151).

301. cf. footnote 213 (p. 80)

108

6.3. Psychotherapy and self-reference

The nominalist discourse easily affords the formulation of a principle of complementarity betweenobserving process and observed object, whenever it is attempted to self-referentially describe a system. Onthe other hand, in the case of self-reference the realist discourse would straightforwardly claim the identityof the observing process and the object of observation (as in the cartesian Cogito).

self-reference:realist discourse: self-transparence (of the Cogito to itself)nominalist discourse: complementarity (between the observed and the

act of observation)

It is precisely here that we recognize a renewed split between the nominalist and realist discourses. Sincethe realist discourse deals with self-reference in a way entirely different from the nominalist discourse,their classical integration is no longer possible.

This integration consisted of the primacy of the analysis-synthesis pair of operations«302», so thatabstract conceptions were related to concrete perceptions, disregarding the allocation of substance andaccident to form and content. Classical science, as an integration«303»of realist and nominalistdiscourses, could not include self-reference as a topic of study without losing its ground, i.e. thisintegration.

When psychology as a newly developing discipline dealt with self-reference, traditional realist approacheswere opposed by nominalist ones (like William James opposing the realism of the ’Hegelians’ and’Fichteans’«304», but also like Sartre opposing the realism implicit in Husserl’s transcendentalego«305»), etc. The self-transparent Cogito was thus opposed by a complementarity between ’I’(pour-soi) and ’Me’ (ego), between the processes of observation and their outcomes, the things observed.

Likewise, psychotherapy, to the extent that self-reference is involved, has not succeeded in finding a wayto continue the classical combination of realist and nominalist discourses. Mostly it became realist,emphasizing essences (self-transparent Cogito, self-consciousness, self-observation, self-insight, etc.).Psychoanalysis, as a paradigm, has become an enterprise into the depths of the patient’s mind, assumingthe idea of full self-accessibility in principle. To the extent self-reference has become a topic of interest, ithas been so at the price of a certain vagueness. Also an appeal to the reader’s sympathy was and still istypical for those studies on therapeutic techniques in which the therapist’s ownsubjective feelingsare

302. cf. section 6.1.1

303. It is also the loss of this integration that characterizes modern developments in the field of science studies, where emphasis isgiven more to the observational (meaning assigning) history of science than to foundational issues (which are more about formal a priorimatters, and hence have more of an essentialist look). The latter are regarded as outdated themes stemming from a past era, viz. theEnlightenment (e.g. Rorty, 1979). Instead of rethinking the relation between realist and nominalist discourses, now that theEnlightenment solution has broken down upon self-reference and other issues, these science studies stick to a (nominalist) emphasis onhistorical observation of scientific developments, without any claim of ’getting into’ them and contributing to them substantially. As ifby doing so one would hopelessly be trapped by some of the old realist assumptions of Enlightenment!

304. e.g.: "... Transcendentalism is only Substantialism grown shame-faced, and the Ego only a ’cheap and nasty’ edition of thesoul. All our reasons for preferring the ’Thought’ to the ’Soul’ apply with redoubled force when the Soul is shrunk to this estate. TheSoul truly explained nothing; the ’syntheses,’ which she performed, were simply taken ready-made and clapped into her as expressionsof her nature taken after the fact; but at least she had some semblance of nobility and outlook. She was called active; might select; wasresponsible, and permanent in her way. The Ego is simplynothing: as ineffectual and windy an abortion as Philosophy can show. Itwould indeed be one of Reason’s tragedies if the good Kant, with all his honesty and strenuous pains, should have deemed thisconception an important outbirth of his thought.

But we have seen that Kant deemed it of next to no importance at all. It was reserved for his Fichtean and Hegelian successorsto call it the first Principle of Philosophy, to spell its name in capitals and pronounce it with adoration, to act, in short, as if they weregoing up in a balloon, whenever the notion of it crossed their mind." (James, 1890, p. I-365)

305. cf. section 5.3.1, esp. footnote 223

109

considered of pivotal importance. This vagueness and this appeal are apparent both in psychoanalyticliterature and in writings on client-centered and experiential therapy. They are in sharp contrast to theobjectivist phraseologies of the behaviorists.

On the other hand, vigorous nominalist trends, like behaviorism, hardly dealt with self-reference, as thiswas considered too often a matter of (reprehensible) introspection. Behavior therapy, as a majornominalist trend in psychotherapy, has even obtained its successes, at least for a long time, due to anexplicit disregard of the therapeutic (interactional) processes themselves. Instead, behavior therapy focusesnearly exclusively upon the proper formulation of techniques and procedures«306». I will returnupon this issue in chapters 8 and 9.

Thus, whereas in the exact sciences self-reference became a topic most fruitfully«307»dealt withwithin a nominalist environment, in psychotherapy the reverse seems to have happened: self-reference hassurvived as introspection only within a realist environment; within a nominalist environment it has beenreplaced by a behavioral technology, one that requires an external controller (however benevolent) of theperson under observation. In experimental contexts the latter has been the research subject, inpsychotherapeutic contexts this has been the patient.

At stake in modern psychotherapy is the reciprocal relation between the (realist) postulation of latent(mental) essences behind manifest (behavioral) appearances on the one hand, which require a considerableamount of introspection and other ’cogitations’ to be unearthed, and on the other hand the (nominalist)assignment of (therapist’s) meanings to (patient’s behavioral) symbols (which hardly allow for self-referential formulations). Or, in other words, at stake is the relation between explanation, the convergentsearch for unifying underlying principles on the one hand, and on the other hand the divergence of thevarious pragmatic implications that can be assigned to a (behavioral) symbol.

psychotherapy:realist discourse: nominalist discourse:

-excavation of essences -assignment of meanings-behavior as appearances -behavior as symbols-theoretically oriented -practically oriented-diagnosis as explanation -diagnosis as tool-self-reference as introspection -self-reference ignored

Here realist and nominalist discourses mix up without integrating. Explanations and practical solutions toproblems are independent of one another, if not incompatible. At best they alternate, but do not integrateas they did in the classical way. A non-classical approach has been missing thus far.

When a discipline, such as psychotherapy, or, to take a point of departure, psychiatric diagnosis, isnotsuccessful in connecting the two discourses, then it might be prudent to look for a new kind ofinterconnection between them and to question the tenability of the classical solution as interconnection.Or, in other words, when a discipline is not successful in relating the explanatory principles that underlieempirical research phenomena to the practical implications that are to be assigned to the research data,then not so much the quality of these data is to be improved by means of (more and more) sophisticatedstatistical and psychometric tools (which is what is usually done in psychology), but rather are therelations between essences and meanings, between explanations and applications, to be reconsidered(which is usually overlooked). But what could be an alternative way of relating essences and meanings?

I will deal with this question in two steps. First, I will make a rather formal point in chapter 7, where Iwill describe the limitations imposed upon us by the usage of the two discourses in which critical objects

306. This is not at all to say that these are performed rigorously and without respect for the unique circumstances of the patients.

307. see the motto of this chapter

110

can be formulated. There, self-reference will be situated beyond the two critical discourses. Second, inchapters 8 and 9 I will explore how a reappreciation of naive discourse can help us to relate essences andmeanings. Chapter 8 will focus on the field of psychiatric diagnosis and how the two critical discoursesdo not permit a classical solution, and chapter 9 will focus on psychotherapy in general and the self-referential structure of exploration processes.

7Self-reference, the one-sided boundary and thenegative wayIn the present chapter I will attempt to present an argument that I consider as crucial for this book. It isthat the self-reference of an entity (e.g. the organizational closure of a living system) can be approximatedby an observer only if he engages himself in acts of observation that are self-referential too. This is to saythat, though one may attempt to observe self-reference, one will not succeed. Self-reference cannot beobserved, but only joined, or, if one likes, participated in. I will attempt to situate self-reference beyondthe confines of the two critical discourses (section 7.1). I will maintain that in the observation of anautonomous entity, once it is attempted to get a glimpse of the autonomy itself, the process of observationjoins«308»with the production processes of the autonomous entity. Furthermore, I will maintainthat an observer thus involved in a self-referential observation process, ceases to be an observer at all.From this it will be concluded (section 7.2) that it is precisely at the interface of the observer and theobserved autonomous system that there is a limit to critical discourses. I will call this limit a ’one-sidedboundary’ (section 7.3). It is a limit both to our thinking, and of the object studied. Finally, I willmaintain (section 7.4) that it is in the recognition of this limit that a negative way to our understanding ofthis limit is possible. Thisvia negativais a process of studying one’s failures to understand something(i.c. self-reference), as a way of coming to grips with it.

For the moment, let us consider the course of my argument thus far. Departing from autopoiesis andrelated concepts, I dwelt upon Maturana’s (predominantly 1987a) conception of language as coordinationsof actions/distinctions, and I extended it in terms of quasi-objects, naive objects and critical objects, naivediscourses and critical discourses, internal objects, and the construction of ’self’. Then I discussed inchapter 6 the way in which, in classical science, realist and nominalist discourses became mutuallycomplementing instead of incompatible, and I presented non-classical science as a departure from thisliaison, due the inclusion of self-reference. By now I am to maintainthat there is no way either for therealist nor for the nominalist discourse to empirically describe self-reference, because neither of themallows the split to be undone that is typical for critical objects, the split between form and content, andtherefore between (agreed) substance and (disagreed) accident. So, in other words, I am implying that thelanguage that arises in the critical discourses is insufficient to make an adequate description of self-reference; self-reference cannot be specified in terms of an essence-appearance pair, nor in terms of ameaning-symbol pair. Hence, a self-referential entity, such as an autopoietic system, is delimited by aone-sided boundary, that is at the same time delimiting our critical understanding.

7.1. ’Self-accidence’ and ’self-subsistence’

Both realist and nominalist discourses have in common a breach between the content and the form of thecritical objects. In fact, the content and the form of a critical object cannot coincide, because they havebeen constructed by their users precisely for the sake of distinguishing an agreed from a disagreed part ofa disputed object. Were they to coincide, the constructed critical object would fail to enable thisdistinction between agreed and disagreed parts, and a social conflict would lack this particular type ofcontrol. If they werenot constructed as apart from one another, then it is because there was no need forthe interactants to construct them at all, neither in a realist nor in a nominalist discourse!

In both critical discourses objects are constructed as substance-accident pairs, so that the interactingpersons can agree about the substance, but remain in disagreement about its accidents (see section 4.3).

308. or more precisely: structurally couples

111

112

But a critical object, thus constructed, can never be self-referential! For in order to be self-referential, theaccident would need to be its own substance, or, equivalently, the substance would need to be its ownaccident. This can only be the case for critical objects in an imperfect way.

Indeed, the realist and the nominalist critical objects«309»can be taken as the constraints betweenwhich a third possibility is to be situated. But why should one look for such a third possibility? Because itmay offer a new integration between the two discourses in a way that self-reference is no longer omittedand excluded, as in the classical calculus solution. In fact, self-referenceis the encounter of form andcontent, it is their moment of identification; that is: self-reference is possible as critical discourse is left.This may be a valuable extension to a discipline in which the classical solution does not work. Forinstance, a new connection between explanation and application may show up.

Clearly, psychotherapy is such a discipline. The psychotherapist’s performance is often not theory ormodel driven, but based upon immediate intuitions, hunches that only afterwards can be accounted for interms not available at the moment of performance«310». Clearly, the psychotherapist does not actaccording to a theoretical model, which he uses in order to calculate his next step and in order to predicthis patient’s next mental state. Instead, the models used by the psychotherapist are of an entirely differentkind: they serve him to reattain some clarity of mind, after the interactions with his patient have confusedhim. As I will discuss more extensively in chapter 9, the psychotherapist does not operate only within theconfines of critical discourses. Indeed, a major puzzle concerning his work may be related to the limits ofcritical discourses! If we can delineate these limits in some way, then this may be helpful to reappraisethe therapist’s job.

In order to include self-reference and not exclude it, as did classical science, we may give it a place andsituate itbetweenthe two critical discourses, as a limit case of both. But this means that we, as long aswe are critical observers, remain incapable of expressing self-reference. While putting into words a self-referential situation, we are bound to express ourselves in the accident-substance dichotomy; we, at leasttemporarily, designate one part of the thing as the accident, another as the substance, whereas actuallysuch a description in terms of a substance-accident relation is inadequate. This means: we are constrainedto our regular usage of critical objects, whereas in fact these do not fit for self-reference! The realist andnominalist discourses each permit only an imperfect approximation of self-reference, and this imperfectiondiffers for the two discourses.

The imperfection of the realist approximation of self-referenceis that the accidents (appearances) are nolonger accessible to others. Such a critical object is no longer the outcome of an agreement between twoparties; there is not an agreed substance and a disagreed accident: both substance and accident havebecome private to their owner. Hence, if this is how self-reference is constructed within the realistdiscourse, then it has lost its quality of a critical object. There is the theoretical construction of aparticular essence, without any way for this essence to appear to an observer who isbeyondthis essence.For instance, andpar excellence, we saw in section 5.2.1 that the Cogito is defined by Descartes in such away as to not allow inspection from the part of a third person (observer). The appearances of the self-transparent Cogito are, strictly speaking, as abstract to the external observer as the substance of whichthey are appearances. It is not too difficult, then, to discard a self-referential substance as a metaphysicalconstruction. Indeed, this has been done a.o. by the logical positivists with respect to the cartesian Cogitoand all other mental constructs, and also in a less rigorous way by the behaviorists. In other words, thereis no way in the realist discourseto specify the appearances of a self-referential essencein a wayindependent of this essence. The appearances (accidents) in fact can be reduced to the essence (substance),and the whole concept of a self-referential essence is caught within an idealist, non-empiricist deadlock.

Conversely,the imperfection of the nominalist approximation of self-referenceis that it is constrained to

309. consisting of the pairs essence-appearance and symbol-meaning respectively

310. cf. Van Strien (1986 or 1997)

113

complementarity, most parsimoniously formulated by Löfgren as ’linguistic complementarity’«311».Thus it is impossible to formulate the transitive parts of one’s stream of consciousness, to catch (anddescribe) the ’I’ that does the thinking performances (such as the self-reflections). We have attained thisconclusion above in section 5.3.1. Likewise, as we saw in section 6.2.3, it is impossible to give adescription (e.g. a computer program) of the interpretation processes required for turning that very samedescription into an action (such as an automatic calculation)«312». A specified performance(interpretation) never coincides with the process through which it arises (the interpretation process, inLöfgren’s terms). In other words, there is no way in the nominalist discourseto specify a description(symbol) that self-referentially describes how it is to be interpreted or defined. And precisely this is whatmakes up the imperfection of the nominalist approximation of self-reference.

In both discourses self-reference remains characterized by a deficient specification of the content of thecritical object. In the realist discourse this is a deficient specification of the accidents (appearances) of theself-referential substance (e.g. the Cogito). In the nominalist discourse it is a deficient specification of thesubstance (set of symbols) that is supposed to describe the self-reference. This is a deficiency, however,that cannot be repaired within either of the critical discourses. It is intrinsic to the substance-accidentdistinction itself, and as long as this distinction holds, self-reference remains split up in terms of thisdistinction. At best, one can use this deficiency within a ’negative way’, as a token for the self-referentialthing that cannot be expressed properly!

It is in the surpassing of these discourses that also for self-referential issues explanations can be integratedwith applications, essences with meanings. This, as we saw«313», had been achieved in classicalscience for topics other than self-referential ones; it was not until self-reference became studied carefullythat the two discourses were set apart again.

A less imperfect approach to self-reference takes into account the circumstance that led to the breakdownof the classical integration of the two discourses; it is the collapse (or indistinguishability) of form andcontent,so that now the form becomes identical to its own content. This collapse entails the collapse ofaccident and substance:

form becomes its own content

realist discourse : substance becomes its own accident(essence becomes its own appearance)

nominalist discourse : accident becomes its own substance(meaning becomes its own symbol)

If in the realist discourse the form becomes its own content, then it is the substance that becomes anaccident to itself. I will call this collapse the’self-accidence’ of the substance. That is, a peculiarrelationship between essence and appearance arises: the essence becomes its own appearance.

Likewise, if in the nominalist discourse the form becomes its own content, then it is the accident thatbecomes a substance to itself. I will call this collapse the’self-subsistence’ of the accident. That is, apeculiar relationship between meaning and symbol arises: the meaning becomes here its ownsymbol«314».

311. See section 6.2.3.

312. cf. my comparison between the concepts of James, Bohr and Löfgren at the end of section 6.2.3.1.

313. see chapter 6

314. This can also be described in terms of the coincidence of an operand and the operator that operates upon it. This treatment ofan operator as its own operand has been discussed e.g. by Kauffman & Varela (1980, p. 195). (See also footnote 40 on p. 29)

114

’Self-accidence’ and ’self-subsistence’ are uncommon but at least uncontaminated terms that I use here topoint at self-reference as a collapse of form and content. They are terms that equivalently denote self-reference as a substance-accident collapse (coincidence)«315». However, ’self-accidence’ and ’self-subsistence’ differ in their origination. They are both approximations of self-reference, but they stem fromdifferent discourses«316».

The realist and nominalist discourses were defined in section 4.3 in terms of the allocations of substanceand accident to form and content. With the disappearance of the form-content distinction, the differencebetween the two discourses also disappears. Though a description of self-reference may be endeavoredwithin a realist as well as within a nominalist discourse, it is eventuallythe same form-content collapsethat is attempted to be formulated in either discourse. Neither discourse, however, is able to express itadequately; self-reference is a limit case that goes beyond the boundaries of either discourse. It is not acritical object, though it may be approximated in terms of critical objects! It is not a naive object either,but in a naive discourse it can be ignored spontaneously, and questions about the status of self-referenceare begged more naturally, without the actor losing his involved and non-detached position.

The notions of ’self-accidence’ and ’self-subsistence’ denote in fact the limits of critical discourse.Though defined in different terms and stemming from different discourses, both surpass the scope ofcritical objects. The concept ’self-accidence’ surpasses the idea of an essence-appearance split. Forexample, it may pertain to the unperceptibility of a living being’s (first order) distinction. ’Self-accidence’differs from the self-transparency of the cartesian Cogito in that the latter keeps its duality of substancecum accidents. Likewise, the concept ’self-subsistence’ surpasses the idea of a meaning-symbol split. Forexample, it may pertain to the unperceptibility of the observer’s own (first order) distinctions. The notionof a self-transparent Cogito implies a deliberate reduction of the accident (thought acts) to the substance(thinking essence); the notion of complementarity, on the other hand, is opposed to this reduction andinstead implies the irreducibility of substance and accident. Both notions, however, present this reductionas an issue, and hence entail a duality between the reducendum and that to which it is to be reduced (inthe case of the realist discourse), or not to be reduced, respectively (in the case of the nominalistdiscourse). On the other hand, ’self-accidence’ and ’self-subsistence’ direct at the absence of this duality.

realist discourse: nominalist discourse:

self-ref within : self-transparent ’Cogito’ (linguistic) complementarityself-ref beyond : ’self-accidence’ ’self-subsistence’

We may express this duality also in terms of an identity between a relatum«317»and the process ofreference to it. Then the terms ’process of reference’ and ’relatum’ can be defined in terms of ’form’ and’content’ respectively:

process of reference: relatum:realist discourse essence (substance) appearance (accident)nominalist discourse meaning (accident) symbol (substance)

315. Strictly speaking, we could also define self-reference as ’content becomes its own form’. Then for the realist discourse thiswould be: ’appearance becomes its own essence’, and the term ’self-subsistence’ would apply here. For the nominalist discourse itwould be: ’symbol becomes its own meaning’, which were to be named ’self-accidence’. Such a duplication of terminology, howevercorrect in itself, would not improve the clarity of my argument, as the reader may be willing to agree.

316. cf.: both1/x and -1/x asymptotically approximate0 asx → ∞Likewise, as form approximates its own content, the effability of the self-reference goes to zero.

317. I borrow this term from Löfgren (1990) who uses it for the ’thing referred to’.

115

Let ’---->’ be a process of reference and ’x’ the relatum, and let ’---->x’ and ’x---->’ denote the pairs ofessence-appearance and symbol-meaning which I presented in sections 4.3.2 and 4.3.3 respectively, thenfor the case of self-reference the relatum ’x’ is replaced by the process ’---->’, though in two differentways:

In the realist discourse the process of referenceproducesitself qua relatum, and the relatum (appearance)can be specified only in terms reducible to the process of reference (essence). In the nominalist discoursethe process of reference is a meaning (interpretation), whichdepartsfrom itself qua relatum (substance).The substance cannot be specified in such a way that, upon interpretation, it yields a self-interpretingprocess. In both discourses the distinction between process of reference and relatum can be made undoneonly in the limit case of a full identity between the two. Each of the two symbolic representations of self-reference would then converge to a single point!

What results after the critical discourses have been surpassed, cannot be expressed in critical terms. It is acollapse of substance and accident, in a way the two cannot be distinguished and in fact are notdistinguished. It is the occurrence of a new situation, one in which this distinction does no longer exist.This deserves some further elaboration and illustration, which will be the subject matter of the rest of thepresent chapter 7.

7.1.1. ’Self-accidence’: the self-appearing essence

In self-accidence an essence coincides with the ways in which it appears. For that reason this essence maybe called ’self-appearing’. If we substitute ’process’ for ’essence’ and ’product’ for ’appearance’, thenautopoietic systems can be recognized as instances of self-accidence, for in these systems the network ofrelations between processes and components is both the product and the generative process«318».

Take the occurrence of adaptive«319»behaviors in an animal. No concrete empirically measurableevent (’appearance’) would be adequate as a valid measure or indicator of the processes that havegenerated it. Adaptive behavior takes place at the interaction site of the individual studied (e.g. animal),and continuously so, in a way that changes contingently upon the history of perturbations that theindividual undergoes. This is also true for those perturbations that stem from a measurement that isapplied to the individual.

That which can be factually observed and measured is not a coincidence of essence and appearance, butsomething different. As long as the observer manages to continue his mode of critical observations, whathe will obtain (through his measurements) is a frozen image of the self-accidence. The difference betweenkinetics and kinematics can serve as a metaphor«320»of this. Kinetics is like Zeno’s arrow,kinematics is like a snapshot of it; the former is the real movement, the latter is the sequence of staticimages, as in a movie. Likewise, the observer’s measurements yield at best a snapshot, or a concatenation

318. cf. the definition of autopoiesis, as presented in section 2.1.

319. I am speaking here about adaptation in an individual organism, taking place from moment to moment; not about adaptation asthe development of changes in a species due to environmental pressures.

320. which I gratefully owe to Gordon Pask, personal communication (1991)

116

of snapshots. However dynamically the measurements may have taken place, they arenot, and cannot be,measurements of the way in which the system adapts to the measurement processes that actually takeplace. The ongoing process of adaptation is not a thing (e.g. a system’s state) that can be measured andobserved. The living organism copes with each process that it encounters, also those that are set up tomeasure it, at least as long as these do not destroy the system«321».

This means that, when it is attempted to critically observe adaptive behaviors, the measuring actions ofthe observer structurally couple (join) with the adaptive processes that were intended as the topic ofobservation. These adaptive processes will not be perceived, irrespective of what in the end may becomethe actual object of perception«322».

The adaptive processes take place in the interaction with the observer. More generally put, when it isattempted to observe a system in which (generating) processes coincide with (generated) outcomes, thenthe observation process itself joins with the thing to be observed, and the observer interested in his topicof investigation would find himself compelled to watch a process that is indistinguishable from his ownperformances with which it is coupled.

On the other hand, if a measurement technique is used that does permit the observer to hold a detachedposition, then the outcomes of the measurements will lack this complication, but it is to be questionedwhat it is that has been measured. Indeed, the observations may be accurate and technically sophisticated,but do they convey the gist of what is really happening? The answer to this question is definitelyaffirmative, if one prefers to stick to classical science and to keep the observer distant from his object, atleast in principle, and dismissing the interactions between observer and observed as not of substantialinterest. The observations may be arranged in a formalism and the formalism may express somemechanism purportedly at work in the system studied. Kampis puts it thus: "Why is this proteininactivated? Because this molecule does this, that does that, and you get what you get." (Kampis, 1991,p. 272) A mechanist model cannot explain creative behavior in a system: in a mechanism each state is bydefinition explainable in terms of its preceding states and the transition functions operating on them.

The answer is tentatively negative, on the other hand, if one does not rule out a priori the possibility thatthe object of study shows itself in ways beyond the researcher’s pre-specified conceptual frames! And thisis what the observer’s critical discourse is: a pre-specified conceptual frame, consisting of substance-accident pairs. Furthermore, in adaptive processes, if anywhere, it is of interest to look for modes ofperformance in the system studied that are new or unexpected. The domain of critical objects, bydefinition, excludes the study of the various modes in which the system observed interacts with theobserver.

By now, I claim, we have found one answer to the question that remained at the end of chapter 2 (section2.4): why cannot we perceive autopoiesis empirically? It is that we cannot perceive it critically, becausethe critical observer isby definitiona distant and detached observer, who cannot take into considerationany immediate contact with his object of study. If we can go beyond critical discourses, we may perhapsperceive it.

The organizational closure of an autonomous system, then, lies beyond the scope of what can be describedand observed in critical terms. This inaccessibility to critical observation may appear to us as a barrierbeyond which all empirical methods fail. It is as if we can think only in highly abstract terms about thatwhich is beyond the barrier, without being able to perceive it. We are reminded of Kant’s divisionbetween the thing-in-itself and its appearance. Is the organizational closure such a thing in itself? I am notgoing to maintain that; to the contrary (cf. section 7.3.3). Perhaps we can find empirical methods (orbetter: hods«323») that are not critical themselves, but that can be used critically! In sections 8.5.5,

321. Maturana (1983, p. 259) calls this the domain of possible destructive interactions (cf. section 2.1.2).

322. such as a naive object, as in the clinical example of entering into a closed system, elaborated in section 10.1.2.2

323. ’method’ without ’meta’ (beyond): the remainder is ’hod’ (way); the worst translation of this would be: ’unreflected way’

117

9.5 and 11.7 I will give examples of what I will call the critical use of naive discourse (and also ofunreflected experiences).

To finish this section, I will make some remarks on a comparison between autopoietic systems andquantum systems. Maturana maintains that it is impossible for an observer to have access to a system’sdomain of states (cf. Maturana, 1982, p. 289), and he makes a general remark about the impossibility toknow a system’s state (1987a, p. 343, p. 350). Varela (1983, p. 157) maintains that structural changes aredifficult to observe at the level of neurons. One can find in Varela (cf. 1979, p. 12, p. 68) the idea of a’logical bookkeeping’ between a ’domain of descriptions’ in which the observer is free to describe anobserved system in terms of functions and purposes, and a domain in which the system’s operations andits internal dynamics take place. What one is missing in the literature on autopoiesis is a specification ofthe type of language in which these internal dynamics can be described in positive terms, or an accountwhy and how such a specification is not possible.

Whenever we want to describe something in critical terms, we need substances and accidents insomeway. In the case of autopoiesis, no proper substance (the network of processes) and no proper accident(products) can be assigned to the system without ignoring its coincidence of processes and products, i.e.without opening the system’s closed network of operations, and ’explain’ it into a stratified arrangement, afrozen image. The perturbations which are triggered by the measurement interactions are, like otherperturbations, included in the living system’s ongoing adaptive dynamics«324». But we may findother empirical ways.

7.1.2. ’Self-subsistence’: the self-describing meaning

In self-subsistence a meaning coincides with the symbol by which it is described. For that reason thismeaning may be called ’self-describing’«325». Here a substitution in terms of ’process’ and’product’ does not fit, as in self-accidence. Whereas ’self-accidence’ is concerned with the closure of anobserved system as an entity that is beyond observation, in ’self-subsistence’ the self-reference is thecoincidence of form and contentof observational acts, instead of a coincidence of form and contentof athing in which an observer is interested (and with which his observational acts become intertwined).

Self-subsistence is a state of affairs at the limit of the nominalist discourse. It does not pertain to anobject of perception, one in which essence and appearance coincide, such as in an autonomous livingbeing, but rather to an act of perception, which is an act of autonomous functioning. It is in such an actthat a meaning can be said to coincide with the symbol describing it. An example can be found in Sheets-Johnstone’s discussion of ’carnal knowledge’:

"...carnal knowledge is knowledge of a pre-epistemological subject in the sense that thehyle carriesits own meaning. [Conversely, t]he temporal structure of the experience of sheer physicality bearsout this self-contained meaning." (Sheets-Johnstone, 1986, p. 241)

As we saw, such a coincidence is generally impossible according to linguistic complementarity. Whatevermay be seen by an observer, it willnot be the act of perception through which it is seen. But that is truewithin critical discourses«326». Self-subsistence, being the limit case of self-reference within the

324. Cf.: "Notwithstanding the essential importance of the atomistic features, it is typical of biological research, however, that wecan never control the external conditions to which any separate atom is subjected to the extent possible in the fundamental experimentsof atomic physics. In fact, we cannot even tell which particular atoms reallly belong to a living organism, since any vital function isaccompanied by an exchange of material through which atoms are constantly taken up into and expelled from the organisation whichconstitutes the living being." (Bohr, 1932, p. 10)

325. In section 4.3.4 (see esp. the scheme on page 69) I proposed the term ’describe’ to relate a symbol to its meaning. Accordingly,a meaning that is its own symbol can be called ’self-describing’.

326. Cf. Reichenbach who maintains the imperceptibility of impressions, which cannot be observed themselves as objects:"I do not say that I doubt the existence of my impressions; but I have neversensedthem. When I consider this question in an

unprejudiced manner, I find that Iinfer the existence of my impressions" (1938, p. 164).

118

nominalist discourse, is the convergence of substance and accident, to such an extent that no distinctionbetween the two can be made any longer. Self-reference is here like a blind spot: it cannot be made theobject of perception, though it does play a role in perception.

For example, it is impossible for us to see in real time the sensory impressions (first order distinctions) oflight upon our retina. There is no perception of these sensory impressions themselves (and, furthermore,the absence of this perception is not noticed during perception«327»). It is through these sensoryimpressions that we are able to see, but that circumstance does not bring them under the scope of theirown perceptual range. This is true for experiences in general (which I presented in section 3.1 as cases offirst order distinctions).

First order distinctions, then, can be considered self-referential in an elementary, if not trivial, sense: in afirst order distinction that which is distinguished (cf. symbol) coincides with the way in which it isdistinguished (cf. meaning). We already encountered this coincidence as a basic property of first orderdistinctions«328»and hence also of experiences«329». In our search for meanings that areself-describing we get some intimations of a prelinguistic domain of functioning, as soon as we attempt toovercome the linguistic complementarity. Like William James’ quest for the transitive parts of our streamof consciousness, we cannot perceive our first order distinctions. But we can live and perform them, andit is precisely by doing so that they may join (structurally couple) with the first order distinctions livedand performed by other individuals.

Self-subsistence then seems to be an unattainable state. Here the observer himself disappears qua observerby losing his distance to the object of his attention, which is his own observational activity. Unlike in thecases described in terms of (linguistic) complementarity, self-subsistence requires the involvement of theobserver himself in a way that undermines his observership.

Again, I claim, we have found an answer to our question of section 2.4: organizational closure cannot becritically observed, because the observer cannot have as critical objects the first order distinctions throughwhich he himself structurally couples to the closed system’s production processes (and hence to itsclosure). The imperceptibility of such a closure«330»(as a property of the thing observed)converges to the observer’s incapacity (as a property of the observer) to use his own first orderdistinctions as objects of his attention. This convergence is of our interest now.

Cf. also: "I do not see colour-sensations, but coloured things, I do not hear tone-sensations, but the singer’s song etc. etc."(Husserl, 1977, Vol. II, p. 559) ["Ich sehe nicht Farbenempfindungen, sondern gefärbte Dinge, ich höre nicht Tonempfindungen, sonderndas Lied der Sängerin usw." (Husserl, 1913, Bd. II, Teil I, p. 374/V§11b)]

327. Von Foerster (1973) has dealt with this issue in terms of the retina’s blind spot: when an experimental subject directs his eyesto a sheet of paper in such a way that a black dot is projected exactly to the retina’s blind spot, then not only does this person not noticethe black dot, but also does he not notice a hole in his perceptual field! It is this phenomenon that, according to Von Foerster,demonstrates that the organism constructs its own reality through its own autonomous organization. The latter is the only criterion ofwhat is perceivable and what not.

To an external observer it is impossible to describe that what the experimental subject perceives,without mentioning that thelatter does not see the black dot! The observer remains confined to an external position (cf. Goudsmit, 1989c), and he is to constrainhimself to expressions of the variety ’the subject does not distinguish between ’black dot’ and ’no black dot’; or: ’the subject does notsee that he cannot see the black dot’.

328. see section 2.2.3

329. cf. footnote 81, page 41

330. I do not call it ’inconceivable’; the idea of such closure is not impossible.

119

7.2. The interface as the limit

When starting from the realist discourse, the search for the interface between observed system andobserver yields as unperceivable the first order distinctions of the observed system. These are structurallycoupled to the first order distinctions of the observer, hence a consensual domain of coupled movements.This happens beyond the critical perception of the observer: first order distinctions cannot be perceived,whether performed either by the system or by the observer himself. Thus, as the development of a domainof first order distinctions takes shape between observer and observed system, the critical observer who isinvolved in it cannot notice what happens. However, he may come to experience things which, afterhaving been recognized, can be arranged as a coupling to the experiences of the system observed. In amost lucid way this has been illustrated by Reik (1948) in his descriptions of the intuitions and hunches,which, during his work as a psychoanalyst, occurred to him as events of which the proper interpretationwas not always evident. But the observer may also couple to those mechanisms that contribute to theproduction of naive objects of the system’s participants, that is, the beliefs, convictions or (family) mythsthey together hold.

All this results, as said, in the disappearance of the critical object, and the observer may accordingly feellost and deprived of his, thus far, comfortable and stable, archimedean point of view«331». Enteringa closed system, then, entails giving up this point of view and allowing a free floating, evenly suspendedmode of attending, as if one were a psychoanalyst attending to his patient’s free associations, or as if onewere a mother in ’reverie’«332», while breast feeding her infant. During this suspension theobserver may have experiences that are identical to those of the system he studies, because theseexperiences exist as distinctions shared at the common interface, that is, as common movements.

It is an entirely different thing, however, for the observer to catch these shared distinctions, not asfeatures of the observed system’s adaptive processes, but as features of his own performance. The searchfor his own first order distinctions yields an awakening out of his blindness. This search, however, is nolonger a realist affair. It is a nominalist search by the observer for his own meaning assignments. Thesemeanings will be assigned to the observer’s own performance, in terms of meaning assignments to theobserved system’s behaviors (as in: ’I find myself being irritated’, which can be unfolded as: ’I findmyself negatively appreciating his behaviors’; where ’behaviors’, in its turn, can be unfolded as ameaningful performance, e.g. some self-defensive actions’). Here we encounter the doubleness of meaningassignment as prepared above«333».

Thus, the search for one’s own first order distinctions is of an entirely different quality than the search forthose of another individual. The latter search is rooted in the realist discourse, and carries to self-accidence as its limit. Here first order distinctions of the observed individual couple with those of theobserver himself, and to chase after the former entails chasing after the latter. The search for self-accidence, then, entails the search for self-subsistence. But if one attempts to push self-subsistence intoclear terms, one ends up in (linguistic) complementarity. This, in fact, has been a major trend of NielsBohr’s inquiries, as he became more and more concerned with the limitations of his own conceptualframework«334».

In sum, the notion of self-accidence entailed the unperceptibility of a living system’s first orderdistinctions, and the notion of self-subsistence entailed the observer’s incapacity to observe his own firstorder distinctions. The perception of self-accidence in a living being would require from an observer theobservation of a first order distinction in the observed, coupled to first order distinctions in the observerhimself, and thus amount to a case of self-subsistence.

331. I will return to this in section 9.2.2.

332. see footnote 555, page 204

333. see section 5.3. and especially section 5.3.1 for a discussion of this doubleness of meaning assignment within the nominalistdiscourse

334. see section 6.2.2, page 102

120

Self-accidence and self-subsistence do not, however, take symmetrical positions in respect of time: it isthrough the observer’s getting confused while searching for the observed system’s first order distinctionsthat self-accidence may be approached. But self-subsistence, rather than being a symmetrical counterpart,serves as a new point of departure for critical self-observation. The critical performance of the observerfades at the limit of the realist discourse, so to say, as it does at the limit of the nominalist discourse.Meanwhile, the two discourses have made contact beyond the observer’s critical perception, at theinterface between observed and observer.

In this way, first order distinctions are limit cases for what can be perceived by a critical observer, i.e. anobserver who splits his world in form-content pairs. This split, which is useful for an observer to agreewith others about something (substance) while disagreeing about something related (accident), cannot bemade with respect of what happens at the boundary between himself and his object of study.

The coincidence of form and content is beyond, as well as constrained by, the domain of what can beperceived critically, and this coincidence happens either as a thing’s unperceptibility, which I called self-accidence, or as an observer’s incapacity to perceive, which I called self-subsistence.Being two ways toexpress the idea of self-reference, both concepts ’self-accidence’ and ’self-subsistence’ pertain to onething, the interface between observer and observed.They are both indications of it, each from its ownside.

This interface is both the observed system’s boundary, at which it realizes its first order distinctions whichare compensations for the perturbations that it undergoes, and also the observer’s own boundary, at whichhe performs his first order distinctions which are the sensorial stuff (tokens) out of which he constructshis percepts. In other words, this interface is the shared boundary between two individuals, constituted bytheir consensual distinctions. It is this interface that can be understood only beyond the object-subjectdichotomy, as atertium where impression and expression are still undifferentiated«335». In chapters9 and 11 I will return to the idea of such a tertium, as the locus of a category of interpersonal encounter.

Self-reference is a limit case of critical objects. It escapes the realist discourse as an invisible boundarybetween an autonomous individual and his medium, which cannot become the object of criticalobservation. But it also escapes the nominalist discourse as a limit which the observer’s perceptionattempts to approximate when studying the interface between himself and his medium. The interfacebetween observer and observed, therefore, also constitutes the locus where the two critical discoursesmeet. It is here that psychology should look for a new relation between explanation and application, and itis here also that the realist and nominalist discourses can be brought back to one another. A first attemptat elaboration of this will be given in chapter 8.

With regard to psychotherapy it is precisely the interface between two persons that is the site where thepatient’s autonomy can be fostered«336», and for the therapist it is the way he may stay (literally)

335. cf. section 2.1.2

336. Cf. Anzieu, who underscores the importance of bodily contact for the development of what psychoanalysts call ’the ego’:"Le Moi-peau, inexistant à la naissance, émerge à partir de cette base, avec l’expérience d’une identification nouvelle à l’object de base,qui est l’identification adhésive (Meltzer) et dont la représentation est celled’une peau commune au corps du sujet naissant et au corpsde son object-support. L’interface qui relie, en les disposant de part et d’autre de la surface commune, le corps de l’enfant et celui d’unepersonne faisant fonction d’object-support, fournit l’infrastructure nécessaireau Moi pour se constituer en interface entre le mondeextérieur et la réalité psychique." (Anzieu, 1984, pp. 1396/7; italics added)["The Ego-skin, inexistent at birth, emerges from this basis, with the experience of a new identification with the primary object, which isthe adhesive identification (Meltzer) and of which the representation is oneof a skin common to the body of the subject being born andthe body of its supporting object. The interface that connects the infant’s body and the body of the person who functions as a supportingobject, by having them on both sides of the shared interface, provides the necessary infrastructurefor the Ego to arise at the interfacebetween the external world and the mental reality." (my translation)]

See also Bick (1988).

121

in contact with the patient«337». In a realist discourse one takes the interface as a feature of thepatient’s autonomy, as a source of information about his "inner world"; in a nominalist discourse onetakes the interface as a feature of one’s own autonomy, as the way in which the therapist’s performance isa meaning assigned to the situation in which he operates. Both discourses are often practiced. The formeras a way to explain what is ’really’ the matter with the patient; the latter as a way to decide how to dealwith the situation. But the interface itself is elusive to both discourses. It is here that no consensus can beattained between observers, simply because it does not permit observation at all, neither naive nor critical.Instead, what can be attained here is something between the interactants, a first order consensual domain,a locus where experiences can be shared, and from which new naive objects may be constructed. Whenthis is the case, there is identification between persons. The experiences of the one are identical to theexperiences of the other: these experiences are not situated inside the persons, but at their sharedinterface. It is the attainment of this that constitutes the passage of a one-sided boundary (cf. section7.3.2).

Furthermore, what we have found, is that the attempt to empirically observe organizationally closedsystems, such as living beings, ends up either in artifacts (open systems, mechanisms) or in the observer’sattempts to perceive the unperceivable first order distinctions. It is this result that I consider a justificationof what is often called (with a pejorative undertone) the ’subjectivity’ of clinical approaches.

This is also to say that what happens at the interface between two individuals is beyond the domain aboutwhich an observer can bargain into consensus with other observers! Therefore, it is beyond the domain ofconversations in which critical observers come to agreements by splitting up the world in agreed(’objective’) and disagreed (’subjective’) parts. To speak of an observer’s ’subjective experiences (of aparticular phenomenon)’ only makes sense if an objective counterpart of these can be found, about whichconsensus exists. But for that which happens at the boundary between observer and observed«338»,no ’objective’ counterpart can be found. It cannot be shared with other observers«339». In fact, it isbeyond the domain of objects altogether!

It is this principle of the boundary as a limit, which is beyond the reach of consensus between criticalobservers, but which is at the same time the ground for the origination of any consensual domain, I think,that is at our disposal when we attempt to investigate psychotherapeutic processes in a way that pretendsto be more sophisticated than regular psychotherapy research. I am pointing at the utilization of parallelprocesses as research tools, in a way analogous to their usage in psychotherapy supervision sessions. Thiswill be illustrated with some case material in chapter 10, and I will prepare the ground for it in chapter 9.Such a utilization does not give a carte blanche or a quality warrant forany intuition or hunch theclinician or researcher might desire to follow, but it does permit us to understand why certain phenomenacannot be observed but by stopping to be a critical observer and by taking one’s intuitions seriously.

The principle of the unattainable interface gives us the argument why themethodof investigation shouldbecome as self-referential as its topic of investigation. This, indeed, is how the parallel process ofpsychotherapy supervision works: the investigators (therapist and supervisor) become involved in theinterface between themselves, on the one hand, and on the other hand their object of study (a particulartherapy session). Thenif the therapy session does possess features of organizational closure, getting intocontact with these features for the purpose of research or supervision will have to take place by the studyof the first order distinctions that are shared between the observer and the observed, comparably to theinvestigation of organizational closure in other systems.

337. Cf.: "A soul is never sick alone, but always a between-ness also, a situation between it and another existing being. Thepsychotherapist who has passed through the crisis may now dare to touch on this." (Buber, 1957, p. 97)["Nie ist eine Seele allein krank, immer auch ein Zwischenhaftes, ein zwischen ihr und anderen Seienden Bestehendes. DerPsychotherapeut, der die Krisis durchschritten hat, darf es nun wagen, daran zu rühren." (1952, p. 13)]

338. Cf.: "... there remains only to study what the world and truth and being are, in terms of the complicity that we have withthem." (Merleau-Ponty, 1968a, p. 107)

339. It is the locus of ’uniqueness’, to which I will return in section 9.2.1.

122

Accordingly, the therapy studied, as a closed system, is to be understood as performing first orderdistinctions that can be felt within a supervisory or research conversation, and studied and discussed interms of ’implicit’ or ’preverbal’ features of the therapy, ’hidden’ shared assumptions etc. As an exampleof this idea we may think of Bion’s (1961) concept of ’basic assumption’, being a qualification that thetherapist, usually in retrospect, assigns to a therapy group performance. Such a basic assumption is in factan example of a first order distinction performed by a closed system (i.c. the therapy group), and the’assumption’ attributed to it is a quasi-object as constructed by the therapist, pertaining to their mode ofperformance«340».

An observer«341», in seeking the limits of his own critical discourse, may touch upon the self-reference in an object, in a way that could not happen as long as he would preserve his critical discourse.Thus a research process can be created that has itself qualities of self-reference. I will return to this inchapter 10.

7.2.1. The ’doctor’s fallacy’ and the concept of ’health’

There is still another way of looking at these matters. It is in terms of disease and health. Both criticaldiscourses can handle the notion of disease, allowing a distinction between substance and accident. Healthhowever cannot be described in such a way. At this place I suggest the notion of health to be a limit caseof diseases, as formulated in either discourse, that is to say: not expressible in critical terms. The health ofan observed individual, then, exists in its capacity to preserve adaptation. In the case of disease, thiscapacity has been partially disrupted.

This, in fact, can be called the ’doctor’s fallacy’ (in analogy to William James’ concept of ’psychologist’sfallacy’«342»): the physician is faced with a living being whose organizational closure is damagedor disrupted; thence he finds confirmed his inclination to explain the existence, the occurrence and thecure of diseases, in the critical terms of the mechanist processes that occur in a non-closed system. Ifmechanist, open system explanations can be provided, then there is no immediate reason for the physicianto consider dark theories on organizational closure and non-mechanist conceptions. However, if illness,qua deviation from a healthy condition, is understood in terms of a disturbed organizational closure, thenit does make sense to consider mechanist explanations as oversimplifications that do not do justice to theorganization of the healthy organism. For instance, an injection of antibiotics does have its curative effectsdue to a healthy context, consisting a.o. of blood circulation, body temperature, and the whole infinite listof the other parameters that can be left out in a mechanist explanation. In short, it is thedeviationfromhealth that is explainable in terms of mechanist processes, not the illness in itself as a condition of theliving organism. Likewise, also the reduction of this deviation can be described in terms of mechanistprocesses, whereas the living system’s complexity itself remains beyond consideration. This can be calledthe ’doctor’s fallacy’.

Likewise, the health of an observing individual exists in an individual’s capacity to adapt to his ownmedium, and not, as is often the case in humans, to ignore his own first order distinctions in favor ofestablished concepts. In other words, an observer’s health is a person’s capacity to open himself to hisown first order distinctions (and, hence, his capacity to construct new naive objects out of them, cf.section 4.2)«343», instead of expressing himself in terms of well-defined concepts. Sartre’s (1943)notion of ’mauvaise foi’ is related to this idea of being bound to follow laws and rules, instead of being

340. See my discussion on quasi-objects, cf. footnote 44 on page 30.

341. or a pair of therapist and supervisor, or a group or other configuration (cf. Olthof & Vermetten, 1993)

342. Cf. footnote 112, page 51.

343. This, however, does not mean that a person should not ward off painful experiences or an overflow of sensory stimulation, butit does mean that unwarranted generalizations of these self-protective measures would interfere with his health. Overgeneralization is atthe root of many neurotic mechanisms, already present at the basic level of avoidance conditioning.

123

free to invent or amend them«344».

7.3. A one-sided boundary

A one-sided boundary can be visualized as a Möbius strip. A strip of paper can be twisted once, and theends can be glued together into a loop. The paper originally had two physical sides, but, due to the twistin the third dimension, there is only one side left: the former one side extends into the former other side,without crossing an edge, and vice versa. In this section I will introduce a kind of one-sidedness that isdifferent from that of the Möbius strip. It is the idea that only one side of a boundary is defined. Hence, aone-sided boundary pertains to the scope of its own definition!

7.3.1. Ambiguities in the definition of the autopoietic system’s boundary

The idea of a one-sided boundary arose as an attempt to interpret some ambiguities. There are severalplaces in the writings of Maturana where one is informed about the absence of the inside-outsidedistinction, for the observed system itself. For instance:

"Given a closed system, inside and outside exist only for the observer who beholds it, not for thesystem." (Maturana, 1978, p. 41).

Furthermore:"... a closed neuronal network cannot discriminate between internally and externally triggeredchanges in relative neuronal activity." (Maturana, 1978, p. 46);

and:"For a living system in its operation as a closed system there is no inside or outside, it has no wayof making the distinction. Yet, in language such a distinction arises ..." (Maturana, 1987a, p. 364;cf. also Maturana, 1988, p. 57).

Also Varela (1979, p. 243) maintains that it is the observer who describes the state changes of theautopoietic system as either internally or externally originated.

On the other hand, there is the notion that the autopoietic system realizes its own boundaries (see e.g.Maturana, 1975, p. 317; Maturana & Varela, 1980, p. 81). The system’s boundaries are considered theoutcome of its operations, not as the result of the observer’s definitions.

How is this possible, if these boundaries are said to exist only in the observer’s domain of descriptions? Itis as if the observer is good enough for making a distinction between inside and outside, after which he issupposed to leave and allow the autopoietic system to produce its own boundaries, and, tacitly, maintainthis boundary between inside and outside.

7.3.2. The idea of a one-sided boundary

We are led to the idea of understanding the autopoietic system in terms of an inside as demarcated fromits outside. That which is inside is inaccessible to the observer: the domain of supposed states. That whatis outside is accessible to the observer: the domain of behaviors and interactions of the system.

The boundary of the autopoietic system that separates its inside from its outside, then, is said to exist inthe observer’s domain of descriptions. The observer who speaks about the system’s ’inside’ behaves as acritical observer: he cannot perceive this inside in the way he can perceive the system’s outside. We arehere in a realist discourse: the system’s inside is dealt with as an essence, the outside as its appearance.

344. cf. section 5.3.1, page 85

124

Therefore, it seems reasonable to characterize the boundary that separates the autopoietic system’s insidefrom its outside asone-sided: from the system’s outside there is a distinction between its inside and itsoutside; from its inside, on the other hand, there is not. This one-sided boundary exists only from itsoutside, as a demarcation constructed by the observer; it does not exist from its inside. That is whatmakes the inside so different from the outside: the observer’s definitions do not hold there. If anything,this inner side is the domain of the system’s autonomy, where the observer cannot penetrate. If ever theobserver is to enter into the system’s inside, it will be by no longer distinguishing it from the system’soutside.

systems with: inside: outside:

regular boundaries: system states environment

one-sided boundaries: no distinction between distinction betweeninside and outside inside and outside

Let us define a system’s regular boundary as that which distinguishes the system’s inside from its outside,irrespective of the dimension or quality of such a distinction. Then, clearly, an external observer can thinkthe system’s inside as separate from its outside.

Let us consider a system, of which inside and outside differ in a special way: ’inside’ denotes thattheredoes not exist a distinctionbetween the system’s inside and its outside, and ’outside’ denotes thattheredoes exist such a distinction.

In fact, a boundary thus defined is paradoxical; it is a boundary that takes its own definition as its domainof operation«345». Hence, a one-sided boundary has a self-referential organization. We find the ideaof one-sidedness in the famous ’ontological proof’ by Anselm of Canterbury (1033-1109), a realistphilosopher. He defined God to be greater than the concept through which he is defined!«346»Likewise, the boundaries of an autopoietic system are defined to be realized by the system itself, in a waythat isnot defined by an observer!«347»Notice, however, that a boundary that refers to its owndefinition, is also a boundary of our thinking, as in Wittgenstein’s foreword to the Tractatus:

"For in order to be able to set a limit to thought, we should have to find both sides of the limitthinkable" (1963)«348»

More generally, a one-sided boundary can be found at the outside of a system in which self-referenceoccurs: on the outside this self-reference can be discussed in critical (and incomplete) terms. On theinside, on the other hand, there is no self-reference.

An everyday life example of passing a one-sided boundary is the transition from being awake to beingasleep. When falling asleep one does not notice one’s slumbering away, and this slumbering can only takeplace to the extent it is not noticed!

345. Thus, satirically, I have proposed (1988) to understand reality and dream as separated by a one-sided boundary: their distinctionis a dream, the reality is that they are identical!

346. Anselm defined God as the greatest that can be conceived ("id quo maius cogitari nequit", ’IQM’). His point was that thisdefinition entails God’s existence: IQM cannot be conceived but as actually existing, or it would not be the greatest. Were it non-existent, then IQM could not be conceived as being even greater than something that exists merely conceptually (cf. de Rijk, 1981a,pp. 149ff.).

347. Indeed, Zolo’s criticism that the ideas presented by Maturana & Varela (1980) resemble scholastic issues is not too muchbeside the mark (Zolo, 1986, p. 145).

348. "Denn um dem Denken eine Grenze zu ziehen, müssten wir beide Seiten dieser Grenze denken können..." (Wittgenstein, 1959,Vorwort).

125

7.3.3. The interface and the one-sided boundary

The system’s inside is inaccessible not because of a transcendental status; it is inaccessible due to theblind spot that the observer’s own first order distinctions constitute for his perception. Are we inventing anew kantian thing-in-itself, this time named ’the system’s inside’ (cf. ’the inside of a black box’)? Itwould be a new thing-in-itself if it were inaccessible to us«349», but that is not entirely the case.

Kant’s world of things-in-themselves, we may say, is concerned with a ’beyond’ where our senses cannotreach; a world in regard of which all distinctions are made in our language, and not in the course of ourpre-reflexive functioning. A thing-in-itself would not allow any interaction with our senses. It is preciselyat the level of first order distinctions, as we saw in section 7.2, that an observer is capable having contactwith the observed system’s first order distinctions, however, at the price of his being an observer!

However, unlike Kantian things-in-themselves, a system’s organizationally closed inside, in which nodistinction is said to exist between its inside and outside, cannot be attained as long as the observer dwellsin language, that is: as long as he remains an observer. It only remains a thing-in-itself as long as theobserver wishes to reflect upon it and remains doing so, without further practical action«350». Inpractical affairs, on the other hand, such as psychotherapy, the actor (therapist) succeeds in entering ’into’this inside, as soon as he manages toforget (or ’undistinguish’) the distinction between the system’sinside and outside.

It is in this respect that the system’s inside does not fit in with a classical kantian framework. Whatmakes the interface between persons different from a classical thing-in-itself, is that it can be attained inimmediate access, as soon as, and only if, the observer leaves language, and thus ceases being an observerat all. This attainment does not exist as an observability, but rather as its opposite: as the (temporary)cessation of the observer’s ability to observe. It happens as soon as a first order consensus between twoindividuals has been established. This is to say: access to the closed system’s inside happens as theoccurrence of consensual distinctions. The observer, whom we had now better call ’actor’, touches thesystem (and possibly already did so), thoughwithout being able to perceive that which he touches; he is’in touch’ (cf. Shotter, 1990), rather than ’in control’.

In order to perceive these consensually made distinctions as they take place, the actor would need to usethem as tokens for themselves, which is impossible«351». Indeed, we have to consider the mereattemptto perceive them as doomed to failure. However, such failure may not be without interest. For,once an observer is prepared to accept the impossibility of the enterprise, he may utilize this failure as a’royal road’ out of language and perception«352»! This is also what we will encounter in mydiscussion of exploration«353»: the observer, after having become self-referentially concerned withhis own point of view, loses control and becomes an explorer. The ’forgetting’ of the distinction betweenthe observed system’s inside and outside happens when the observer’s activities become self-referential ashe examines his own position, and as he discoversthat he cannot perceive nor understand his object ofstudy.

349. As believes Kampis (1991) with respect to autopoietic systems: "The self-referential, autonomous units conceived in theautopoiesis theory are ultimately closed to themselves. They are examples for the KantianDing an sich. If self-reference and autonomyare complete, there is no window let in the system through which we can peep. More importantly, there is no possibility to define ormodify the system from the outside." (p. 390)

350. According to Varela, Thompson & Rosch (1991), the pitfall of western philosophy is that thought is not complemented withpractice. According to them, in particular phenomenology has not solved this difficulty.

351. see section 7.2

352. This is what many Zen exercises seem to be concerned with. Cf. "[Zen] aims at helping or pushing the student to breakthrough into a new integration by separating him from all logical, theoretical, disciplined patterns of understanding" (Whitaker, 1976, p.158)

353. section 9.2

126

An observer who is prepared to admit this, may find himself in an immediate contact, one that he doesnot observe any longer«354». We may consider this apassage of the one-sided boundary. Theobserver (or better: the person thus far indicated as ’the observer’) does no longer distinguish betweenhimself and that which he (only seen from an external perspective) is touching. It is the occurrence ofsuch moments of identification between therapist and patient, that, I think, is widely recognized in circlesof clinicians, however without translating its practical importance into a criticism of the methodologicalconstraints that are usually put upon scientific investigations. If we were to allow this translation, wemight end with avia negativatowards the topics of our interest, such as the events of psychotherapy.

7.4. A via negativafor understanding psychotherapy

7.4.1. Definition of the negative way

Miller (1997) excellently resumes the gist of Cusanus’ negative theology:"For Cusanus, every cognitive measure we employ will be inadequate in the case of the

finite things of our experience. A fortiori, when it comes to knowing the infinite God, the gap thatopens so confines our knowledge that it amounts to ignorance. But this is not uneducatedignorance. Understanding our ignorance regarding God involves grasping the reasons for it. Onlythen do we achieve reflective or educated, explicitly "learned ignorance."

To stress the incommensurability between our cognitive measures and the divine Nicholasproposes a regulative principle he believes obvious and never rejects: "It is self-evident that there isno comparative relation of the infinite to the finite." (I.3.h 9)«355»One way to interpret Dedocta ignorantia is as a series of creative proposals that help us reach toward the God who ever isbeyond our grasp. To follow Cusan thought is to confront one’s inadequacy in knowing God and tobe willing to entertain various proposals rather than any final story about how such matters stand."

Merleau-Ponty (1968a, p. 179)«356» introduced the idea of a ’negative philosophy’, in analogy toCusanus’ concept of a ’negative theology’. Contrary to a Sartrean approach of understanding theunreflected consciousness as a nothingness, Merleau-Ponty’s ideas on unreflected consciousness (’theflesh’ (cf. 3.1), also called ’Being’) is not so much put into the focus as an object of investigation. Ratherit is our very (failing) effort to get a glimpse of it, that is considered the only possible road to intuit theflesh. This makes Madison state that philosophy’s ’negativity’ according to Merleau-Ponty does not somuch concern its ...

".... "object" - it is not a philosophy of the subject or of Being conceived asnothingness- butrather its "methodology". Merleau-Ponty’s ontology is a "negative philosophy" solely in the sensethat for it the ultimate criterion of evidence is not, as Husserl insists, that the "object" ofphilosophical knowledge be present "in person", "in flesh and blood"; it is "negative" because it isnot a "phenomenological positivism" which culminates in an eidetic intuition, in aseeing. (....)Consciousness can think Being or the flesh only as that which resists and hides itself fromconsciousness, and as latency and transcendence as such. (....) This philosophy of Being worked outin the form of a negative philosophy does indeed therefore, by its method, resemble a "negativetheology", because for such a theology the highest knowledge of God, the ultimate Transcendence,can be obtained only when everything has been done to know what God is and when one thuscomes to know that one knows nothing of him; for it is then that one comes to know God, in hisvery proximity, as the totally other. (...) consciousness cannot think Being except by thinking itsown blindness, what it cannot see because it is that which makes it see." (Madison, 1981,

354. "Searles points out that especially in his treatment of schizophrenic patients he must first achieve with his patients an early,often preverbal measure of relatedness, or "oceanic relatedness," before therapist and patient can move on to the verbal, and then themetaphorical or abstract levels of relating that will be accomplished before termination and separation." (Schlesinger-Vaccaro, 1983,p. 93)

355. This page number refers to the first book of Cusanus’ work ’De docta ignorantia’ (1514).

356. 1964a, p. 233

127

pp. 195/6)

Why are these remarks about Merleau-Ponty’s notion of ’negative philosophy’ for us of significance? It isbecause they demonstrate a transition from the domain of second order consensual distinctions, into thepre-linguistic domain of first order consensual distinctions. It is by our (failing) attempts to objectify (thatis: make perceivable as naive objects) our own first order consensual distinctions, that we manage to leavethe domain of second order distinctions. For when we try (by second order distinctions) to perceive (andtherewith objectify) our first order distinctions, we attempt to take them as tokens, not for some newlyemerging naive object, but as tokens for themselves!

Madison:"Being is conceivable only as the "shadow" and the latency of our experience, that which is presentto us only by remaining irremediably absent from us. In a working note of May 1960, Merleau-Ponty«357»says that the junction between the touching and the touched - that is, that "in-between" which links up the two "leaves" of the flesh, the sentient and the sensible - takes place inthe untouchable, that is, in the flesh, in Being." (Madison, 1981, p. 197)

Merleau-Ponty’s ’negative philosophy’, therefore, is not striving towards some new conceptualization of aSartrean ’nothingness’; rather, it is an attempt to get rid of the very distinction between object andsubject, by thinking that which precedes this distinction. Empirical psychology has accepted andappropriated Sartre’s idea of the Subject as a ’nothingness’, beyond the scope of the observable world.Thus, empirical psychology has accepted itself as a positive science, occupied with objectifiable andmeasurable things and phenomena. The nothingness, the consciousness, the subjective side of anyintentional act, is by definition to remain behind the clouds of metaphysics. This is what modernpsychology gratefully took from Sartre, and it used this idea to justify its own positivistoutlook.«358»

7.4.2. The use of the negative way

Instead, our attempt to think the flesh, the source common to both the act of touching and that which istouched, their locus of connection - this attempt at thinking that which precedes thought is precisely thenegative way that, I think, should be used in our search for an understanding of first order distinctions(and, accordingly, an understanding of projective identification, cf. section 11.4). Instead of abstainingfrom what seems to be impossible, we may give it a try, and see what goes wrong and how it goeswrong! Thus, if an observer fails in grasping for an archimedean point, from which first order distinctionsare hoped to be perceivable, it may be useful for him to study his failures as acts of grasping that losetheir quality of grasping, as acts of indication that cease to be indicative at all, etc. It is here that the twocritical discourses come together: the incapacity to touch (cf. self-subsistence) meets the thing that cannotbe touched (cf. self-accidence). They meet in what Merleau-Ponty called the ’untouchable’, whichcorresponds to what I called the ’interface’ between observer and observed.

One further implication of Merleau-Ponty’s negative way is that it brings together ontological andepistemological considerations: the flesh qua untouchable is both a category of being and of knowing! Orperhaps better: it is a place where knowing and being are still undivided. It is the idea that experiencing isan articulation of being, not a representation of an already given world. It is here that we find also acorrespondence to Maturana’s idea that reality does not pre-exist its being distinguished«359»!

""The negative here," Merleau-Ponty says, "is not apositive that is elsewhere(a transcendent) -- Itis a true negative (...) ... Merleau-Ponty still refuses the idea of a "transcendent," that is, something

357. 1968a, p. 254

358. In the Netherlands this has become apparent in a particularly spiritual manifesto (Linschoten, 1964).

359. cf. footnote 52, page 32

128

which would have an existence in itself outside of our experience." (Madison, 1981, p. 197)

Instead, the existence of this ’true negative’ is the impossibility to catch the experience in itsunreflectedqualities; an impossibility that Merleau-Ponty also compares to the retina’s blind spot (’punctum caecum’,1968a, p. 248)«360». It is this impossibility that we will find again in my discussion of projectiveidentification, especially the notion of ’unconscious phantasies’, as used by Melanie Klein«361».

In view of this junction of knowing and being, empirical psychology can become ’radical’, in a wayWilliam James (1904) used the term for ’radical empiricism’:

"To be radical, an empiricism must neither admit into its constructions any element that is notdirectly experienced, nor exclude from them any element that is directly experienced. For such aphilosophy,the relations that connect experiences must themselves be experienced relations, andany kind of relation experienced must be accounted as ’real’ as any thing else in the system."(James, 1904, §1)

Thus, issues such as ’change’, ’continuity’ and ’discontinuity’, can be experienced, according to James,and, if they are experienced, they are as real as the more conventional objects of attention. The positionJames was proposing was truly radical in view of the (originally British) associationist tradition accordingto which ideas and images can be related to sensory impressions, but their conjunctions not. Hence, arelation between two ideas (or two images), such as a causal relation, was considered to be entirelyconstructed by the subject, i.e. not rooted in sensation (e.g. Hume’s treatment of causal relations). Nowwhen James suggests that this relation too can be experienced, then this means that James in fact departsfrom a conception of the world as hierarchically organized!«362»

This also is of interest for a ’negative way’. For not only objective facts and well-definable topics can betaken as the subject matter of investigation, but also the experiences related to the processes ofunderstanding and confusion, both as they happen to psychotherapists and patients, and as they happen toa researcher himself. By studying the investigator’s own confusions, a domain of phenomena opens upthat by definition would evade any positive approach. This is what I will do in section 11.1.

Indeed, this would be a negative way to understand the complexity of psychotherapy! Very often apractitioner does not need to go this far; a partial and temporal return to a naive discourse wouldsufficiently allow him to have access to many topics that would remain inaccessible within a criticaldiscourse. This will be dealt with in chapters 8 and 9.

However, many times the development of a naive discourse is impeded, and the therapist has to becomethe critical observer of these impediments. It is here, I think, that a negative way to understandpsychotherapy would be appropriate. This will be elaborated in chapters 10 and especially 11. I willpropose there to use the negative way in approaching those aspects of psychotherapeutic sessions, thataffect the capacity to understand in those involved.

360. We find here the ’black hole of psychotherapy research’ (cf. Goudsmit, 1990). Notice that Shotter (e.g. 1987a) is alsoconcerned with this idea, in his discussion of ’mind’ in terms of an ’imaginary’ entity.

361. cf. section 11.3.2

362. There is an interesting theoretical link with the work of Rosen. In a living system, according to Rosen, there is an entanglementof functions and relations between functions, such that a single physical process may play roles at various logical levels simultaneously.Rosen describes these entanglements by means of category theory, a branch of mathematics in which "there is nothing (...) that mandates[an] absolute distinction between sets and mappings [of sets]" (1991, p. 135). In this way, a mapping can be the effect of anothermapping, and so in a circular way. The radical thing is that in this way Rosen manages to avoid the conceptualization of living beingsin terms of a hierarchy between physical processes and their interrelations; instead, a relation between processes (modeled as a mappingrelation between sets) may itself be recognized as the outcome of a process, and, conversely, a process may be recognized as producinga relation between other processes.

8Psychiatric diagnosis between realist andnominalist discoursesThis chapter presents an opposition between realist and nominalist discourses as a major underlying fieldof tension in the debates on criterion based diagnostic classification systems such as DSM-IV. It is arguedthat, unlike disciplines of classical science, psychiatric diagnosis does not know an integration betweenrealist and nominalist discourses, as long as it does not recognize the presence of acts of exploration thatare self-referential in kind. It is also argued that the nominalist discourse, though dominant at present, isundermined by realist undercurrents. It is maintained that these undercurrents are partly due to a disregardof the importance of the diagnostician’s informal, personal engagements with the patient. Procedures thatare based upon strictly behavioral criteria, such as those of the DSM, are considered reconstructionswithin a nominalist discourse; it is recommended not to counter them by making reconstructions within arealist discourse, but rather by focusing upon the reconstructed nature of these procedures, and bycontrasting them with some properties, notably self-referential properties, of the exploration processes thatactually take place in diagnostic sessions.

8.1. Introduction

There is a problem in the current debates on psychiatric diagnosis. For a long time, to make a diagnosishas been a matter of personal (’clinical’) judgments, and accordingly clinicians did not agree very muchabout diagnoses. This situation changed when scientific endeavors stratified the usage of concepts andmeasurements. As a result, the ’personal touch’ of the observer has become less important or has evendisappeared, and diagnoses became more objective. But at what price? I will maintain that it was at thecost of a connection between theoretical knowledge and practical knowledge; more precisely: a connectionbetween the etiology and the treatment policy of a particular mental disorder in a particular person. I willmaintain that this connection can not be found as long as the ’personal touch’ of the observer is dismissedas a personal bias, or even as a personal measurement error«363».

A major aim of classical science has always been to attain objective observations and descriptions, inwhich measurement errors have been minimized, or have at least been taken into account for furthercalculations. As a result the observer as an agent who is in interaction with his object of study has notbeen acknowledged as included in the system studied, but he has been understood to be one who doesnotparticipate«364». Hence, his subjective feelings and opinions are considered as features of theresearch process, and of methodological interest at best. This is common scientific practice, based uponthe ideas of Kant, Newton and others. In classical science«365» it has become self-evident that theobserver’s personal feelings, qua features of the measurement process, are not of any substantial interest;the course of a billiard ball will not be affected by itin principle. Consequently, because the observer’sposition does not really count, classical science has been capable of closely linking (theoretical)knowledge about causal mechanisms to (practical) knowledge about effective control. For instance,knowledge of the state of the billiard ball (e.g. its speed, acceleration, spin, etc.) is immediately useful forcalculating our opportunities to change the ball’s trajectory. This observer independence has not onlybecome a quality warrant for objective knowledge, but also a warrant for a smooth interconnection

363. Cf. the term ’personal equation’ in the history of astronomy, where it was used in order to indicate the regular patterns foundin a particular observer’s measurement errors (see footnote 214 on page 81).

364. Cf. Howe & von Foerster, 1975.

365. In chapter 6 I have expanded on the absence versus the presence of self-reference in classical and non-classical sciencerespectively.

129

130

between theoretical understanding and practical application.

However smooth these links between theoretical knowledge and its practical applications may often be,they are not guaranteed, for theory and practice are two separate things. If we can skip from one to theother, without having to depend on our subjective intuitions, then this is only due to some calculus thatwe have at our disposal, one that provides us with a way to infer previous and subsequent states of theobject that we observe«366». If such a calculus does not exist for a particular domain of research,then we cannot align our knowledge about past developments with knowledge about future possibilities.The sober fact is that in the field of mental health we do not have such a calculus, however much wedesire it and believe that we are near to it. The fields of etiology (how come?) and treatment policy (whatto do?) are not united as they are in Newtonian mechanics. There may be a good reason for thiscircumstance.

I will discuss these two fields, etiology and treatment policy, in terms of the realist and the nominalistdiscourses. I will concludethat these two discourses cannot be connected without a personal engagementof the observer. This is the basic dilemma of a diagnostic procedure that intends to be ’objective’, i.e. inwhich the observer aims at disengaging and detaching himself from what he observes, as is common inregular science.

Therefore, the present chapter departs from the idea that the field of psychiatric diagnosis is beyond thescope of classical science. There is a role for self-reference to be played here, that is explicitly forbiddenin classical science: the process of diagnostic judgment itself is part of the object of diagnosis. In otherwords, the observer’s personal, subjective intuitions constitute a substantial part not only of the diagnosticprocess, but also of the outcome of this process. These subjective intuitions are indispensable forpsychiatric diagnosis, not because they are the imperfect and preliminary surrogates for objectiveassessment procedures that are underway to being discovered, but because measurement is pointlesswithout them, like it is pointless to speak of the state of a quantum particle without considering thecontext of a particular measurement«367». The diagnosisalwayspertains to the conversations andthe encounters that underpin and construct it. The object of assessment cannot even be fully separatedfrom the way in which the assessment process itself took place. Such an absolute separation was anachievement of classical science, that non-classical science did not preserve. These are some of the ideasfrom which the present chapter departs.

However, if one prefers to ignore, or even deny, the self-reference of diagnostic processes, then this is atthe price of finding a lacuna between theory and practice, or better, between explanation and treatment. Iwill describe in this chapter how two approaches that intend to remain ’objective’ by keeping the observerbeyond the object studied, are incomplete and mutually incompatible. The approach that puts priority toexplanation brings one into a realist discourse; the other approach, putting priority to treatment, amountsto a nominalist discourse.

366. This classical science assumes that a) a system is always in a particular state at a particular moment; and b) between thesestates transitions take place according to laws of nature. A critical appraisal of these assumptions behind the scientific tradition ofNewtonian mechanics can be found in Rosen, 1991 (esp. pp. 91 and 95).

367. In quantum mechanics the interaction process between the object and the measuring device could no longer be ignored, as inclassical science, because the measurement interactions determined which of the state variables of a particle could be measured, andwhich not. In a similar vein, there is in interpersonal encounters always a degree of projection and projective identification (to borrowterms from psychoanalysis), that escapes the attention of the participants, even if one of them is a trained diagnostician. This is to saythat the countertransference responses will to some extent remain ambiguous and uninterpreted. The best one can do in thosecircumstances is to report their occurrence as personal experiences, without pretending to provide a clear distinction between themeasurement process (esp. the feelings of the screener) and the object measured (the patient’s mental condition). This parallels Bohr’sideas on measurement in quantum mechanics (cf. Bohr, 1934, esp. p. 15ff.). It offers a formal ground for the use of subjectiveexperiences as empirical instruments (instead of the use of subjective experiences for want of more objective measures!).

Cf. also: "... the validity of the experimental method depended on the assumption that the effect of an experiment on thephenomena being studied could always be reduced to an acceptable and measurable extent. (...) ...bitter experience has now obligedscientists to recognize that this assumption holds good only in a certain limited range of cases and situations; and, specifically, that mostof the subject matter of modern psychology lies outside this range" (Toulmin, 1986, p. 44).

131

The two critical discourses cannot be combined in order to create one integrative image of both thebackground of a patient’s disorder and of his treatment options. Only a reappreciation of the observer’spersonal engagement would allow this. A discourse that permits such engagement can be called ’naive’.

The present chapter does not explore the phenomena of self-reference in psychiatric diagnosis; nor is itconcerned with the question what mental illness ’really is’, nor about how to engage into a dialogue witha patient, but instead it discusses psychiatric diagnosis in terms of the incompatibility between realist andnominalist discourses.

Accordingly, I will describe a realist and a nominalist way in which in psychiatry mental disorders areconstructed as objects that are beyond the domains of description of the patients. Not the patient will beof my interest here, but the descriptions made by those observing him as athird person, observed anddiscussed for the sake of theoretical explanations or therapeutic measures.

The major tenets of this chapter, then, are as follows.1. In the field of psychiatric diagnosis the realist-nominalist controversy is more basic to the debates

on current diagnostic systems than the often quoted opposition between descriptive and theoreticalstatements.

2. The DSM is a nominalist classification system, thwarted by realist undercurrents.3. Both in theory and in practice there is a ’desire for etiology’4. In practice, the ’desire for etiology’ amounts to a desire for attaining a personal understanding of

the patient. Since the DSM does not recognize ’personal understanding’ as a useful kind ofknowledge, practitioners often experience the DSM criteria of disorders as ’superficial’, and may betrapped in a realist discourse.

5. Naive perceptions are not fully replaceable by objective measurement procedures, and they shouldbe recognized as such.

I will first describe current realist and nominalist conceptions of mental illness (sections 8.2 and 8.3).Next, I will describe some ways in which the nominalist approach is being undermined by realistinfluences (section 8.4). Finally, I will discuss some reasons for these influences in terms of what will becalled ’the desire for etiology’, both in theory and in practice (section 8.5).

8.2. The realist discourse in psychiatric diagnosis: ’neo-Kraepelinianism’

In the realist discourse the clinician takes diagnosis as the description of a truly existing internalstate«368» in the patient, a state that has a distinctly pathological quality. Hence, a realistpsychiatric diagnosis can be true or false, according to whether it corresponds to this internal state (whichis regarded as the essence of the illness, as in e.g. endogenous depression). Psychiatry’s classicalparadigm is Kraepelin’s nosological system (Kraepelin, 1899), in which mental diseases are classifiedaccording to their etiological basis. Kraepelin calls the latter’Krankheitszustände’(disease states) (1899,p. I-7).

368. Alternatively, the uncommon and inconsistent behaviors of a person may bring an observer to concluding an illness or thepossession by a demon. This is well illustrated by the following poem:

"When fevers trouble my child frequently / When all the diseases / Have fallen in love with him / And all youthful diseases /Run after him / As if he was a beautiful girl / So that he has coughs and dysentery / And throat trouble and eye sickness / And his earshave pus / And his legs have ulcers / And he is bony, skinny / And his loin-string is loose / I know that this is not for nothing. /

I know that someone is behind it / I know someone has hidden / The child’s excreta in a tree fork / Or has buried his hair ornail pairing in a river bed / I know that some jealous woman / Perhaps even a close relative / Has visited the shadow trapper / Who hascaptured the child’s shadow."(Okot p’Bitek, 1966; as quoted by Van Roosmalen-Wiebenga, 1988, p. 71)

But the reader is not to assume that this type of argument be more typical of exotic cultures than of ours (cf. Menninger, 1963;see also section 8.3).

132

(mental) disorder according to the realist discourse:

abstract: concrete:

realist discourse: essence (what?) appearance (how?)(mental) disorder: disease entity symptoms,

observable behaviors

Next to Kraepelin’s etiological foundations, vast and substantial theoretical contributions to psychiatricthought were provided by other movements, such as the psychoanalytic movement. Thus the theories ofpsychiatric diagnosis became more and more a matter of etiological assumptions and at times even articlesof faith, and its practice a matter of vague ’clinical judgment’ from the clinician’s part«369».Empirical diagnostic considerations did guide treatment policies of patients, but to a relatively smalldegree.

From this situation, two kinds of way-out have been attempted in psychiatry. One was a revival ofkraepelinian realism, this time based upon some more solid looking research in biological psychiatry. Theother was a full turn into a nominalist discourse, away from etiological considerations and unspecified andvague terms, and towards a more technological approach«370».

To start with the first way-out, psychiatrists have attempted to improve the quality of their diagnoses bymaking psychiatry a more medically oriented discipline. The present day champion of this approach isundoubtedly Klerman. He launched a programme for ’neo-Kraepelinianism’, the nine basic assumptions ofwhich are put together as a ’neo-Kraepelinian credo’«371».

"The neo-Kraepelinian credo includes nine propositions:(1) Psychiatry is a branch of medicine.(2) Psychiatry should utilize modern scientific methodologies and base its practice on scientific

knowledge.(3) Psychiatry treats people who are sick and who require treatment for mental illnesses.(4) There is a boundary between the normal and the sick.(5) There are discrete mental illnesses. Mental illnesses are not myths«372». There is not

one but many mental illnesses«373». It is the task of scientific psychiatry, as of othermedical specialties, to investigate the causes, diagnosis and treatment of these mentalillnesses.

(6) The focus of psychiatric physicians should be particularly on the biological aspects of mentalillness.

(7) There should be an explicit and intentional concern with diagnosis and classification.(8) Diagnostic criteria should be codified, and a legitimate and valued area of research should be

to validate such criteria by various techniques. Further, departments of psychiatry in medicalschools should teach these criteria and not depreciate them, as has been the case for manyyears.

369. Thus, Strauss (1982, p. 8) speaks of a "vagueness and pseudo-communication that marked psychiatric diagnosis in the 1950sand 1960s". Cf.: "Decades ago, psychiatrists and psychologists often made diagnoses by clinical "feel" or "intuition". A patient "looked"depressed, or one had a "gut feeling" that he or she had a personality disorder." (Reid & Wise, 1995, p. xxvii)

370. This nominalist way-out will be discussed in section 8.3.

371. It is interesting to observe that Klerman (1978, p. 105) calls R.L. Spitzer (the chairman of the APA task force that hasdeveloped DSM-III and DSM-III-R) as one of the investigators associated with neo-Kraepelinianism. A few years later Spitzer (1982)will publicly resign from a neo-Kraepelinian school, as he declares himself to disagree with Klerman’s fifth and sixth article.

372. By this Klerman clearly refers to Szasz’ idea of mental illness as a myth.

373. By this Klerman refers to Menninger, cf. section 8.3 where I will quote from Menninger (1963, pp. 32/3) on the idea of a’unitary concept’ of mental illness.

133

(9) In research efforts directed at improving the reliability and validity of diagnosis andclassification, statistical techniques should be utilized."

(Klerman, 1978, pp. 104-5)

For Klerman the revived interest in classifications is in the service of the neo-Kraepelinian (realist)conception of illness«374». Diagnosis and classification are made for the sake of discriminatingillnesses from one another as separate and disjoint disease entities (cf. credo items #4 and #5). This ismore than a matter of conceptual refinement for practical purposes. It is also a matter of policy. The neo-Kraepelinian project aims at restoring the status of psychiatry as a solid discipline, a hard science; not asa field of troublesome and vague issues, subject of depreciation (cf. credo item #8). In order to attain thisgoal, the neo-Kraepelinians not only choose for an essentialist approach, but also want to have theessences operate as etiological grounds«375». The kind of clarity they seek pertains to thedistinction of essences behind a multitude of confusing empirical phenomena. This is what makes thisproject a matter of realist discourse.

The success of ’neo-Kraepelinianism’ has been considerable (cf. Blashfield, 1982, who even explains itssuccess in terms of ’invisible colleges’). On the other hand, Van den Hoofdakker (1989), after agreeingwith Klerman (1984) about biomedical research findings, declares himself to be much more reserved inregard of their implications for treatment. Apparently the relation between theory and practice, betweenexplanation and application, is a rather delicate one, even in the field of biological psychiatry.

It is in this respect that a watershed can be found between those who desire a classification system for thepurpose of finding a distinct etiology of these illnesses (e.g. the neo-Kraepelinians) on the onehand«376», and on the other hand those who are after classifications for the sake of practicalmatters such as treatment policies, andnot for (theoretical) explanatory reasons. Indeed, it is thiswatershed that, in the field of psychiatric diagnosis, represents the realist-nominalistcontroversy«377».

8.3. The nominalist discourse in psychiatric diagnosis: ’DSM and its vicissitudes’

The other way-out from the fuzzy realism that was reigning the field of psychiatric judgment in the earlydecades of this century, is a nominalist turn«378». Here it is the quest for a conceptual clarity, thatis given a central position. To give a metaphor, one might compare a psychiatric syndrome to a holidaytrip, arranged by a travel agency, a trip in which historical and cultural monuments are combined in avisit. Then within a nominalist discourse it is considered pointless to believe that such a trip is the hiddenbackground or essence behind the existence of the monuments visited, let alone theircause. In thenominalist discourse the psychiatric classifications are the delineations of behavioral patterns, and it would

374. Klerman (1986, p. 16) considers the work of Mayer-Gross, Slater & Roth (1954) as the initial statement of the neo-Kraepelinian point of view: "Strongly critical of psychoanalysis, psychotherapy, and social psychiatry, this book was an aggressivereaffirmation of the Kraepelinian approach."

Cf. also footnote 383 on page 135.

375. cf. also footnote 418 on p. 143.

376. in Klerman’s words: "The current revival of a scientific approach to psychiatry in general and to schizophrenia in particular haslead to a systematic attempt to define schizophrenia on the basis of a symptoms syndrome and to apply operational criteria to thesyndrome. This is part of an overall view that sees mental illness, like other illnesses, as having causes and delineations. The convictionis that in the absence of etiological knowledge, the most fruitful way to proceed is to define, on descriptive clinical grounds, the varioussyndromes." (Klerman, 1978, p. 115)

Perhaps we may consider Skodol’s interest in etiology (see footnote 406 on page 140) as an instance of this conviction.

377. Cf.: "The clinicians (...) need a system which will communicate easily, and which will assist in predicting the results oftreatment. Scientists see classification as an ultimate objective in which the "natural" order is represented and, in the case ofpsychopathology, as based primarily on etiology. These different objectives result in different values as to which studies and whichevidence are worth considering." (Katz, 1982, p. 10).

378. cf. footnote 370, p. 132

134

be pointless to believe that by their very definition these patterns would have turned into any underlying’disease entities’ that generate the observed behaviors. Instead, the definitions are calibrated by practicalcriteria, such as treatment options. Causal backgrounds are of interest only to the extent they contribute toa greater clarity of treatment policy.

(mental) disorder according to the nominalist discourse:

abstract: concrete:

nominalist discourse: meaning (how?) symbol (what?)(mental) disorder: classification, empirical phenomena,

treatment indication data, observable behaviors

When taken as a means to merely classify a patient’s modes of functioning, psychiatric diagnosis is of anominalist variety. Giving labels to a person’s behavior may be useful in order to determine the propertherapeutic treatment policy. As such, this kind of diagnosis is not true or false with respect to underlyingessences, for no correspondence to an internal state is intended. Rather, it is the practical usefulness of thediagnosis that matters, and, if in the nominalist discourse a diagnosis is said to be correct, this is only tosay that a particular diagnostic classification did in fact lead to a fruitful treatment policy. The practicalusefulness that matters here may concern overall treatment strategy, or particular tactical maneuvers inorder to recognize patterns, regularities, tricks or ploys in the patient’s behaviors. The nominalist view ofpsychiatric illness (or ’mental disorder’, as it is called in the DSM«379»), therefore, is pragmatic inkind. As a way-out different from neo-Kraepelinianism it is not interested in the purported illness quadisease entity, but rather in treatment options that can be assigned asmeaningsto the empiricalphenomena, which are taken assymbols. I will return to this in a minute.

A major proponent of the nominalist discourse in psychiatry has been Menninger, who suggested a’unitary concept’ of mental illness:

"... we propose to think of all forms of mental illness as being essentially the same in quality, anddiffering quantitatively. This is what is meant when we say that all people have mental illness ofdifferent degrees and different times, and that sometimes some are much worse, or better. (...)

The unitary concept does not dispense with the descriptive designations. These we must haveif they can be cast in a form that will not deny the essential unity of the process or obscure theunderstanding of the adaptation difficulties of the patient. The object of the process ofdiagnostication is not the collecting and sorting of pretty pebbles (although even this may be ofsome scientific value in large-scale epidemiological studies). It is, rather, to provide a sound basisfor formulating atreatmentprogram, a planned ameliorative intervention." (Menninger, 1963,pp. 32/33)

and:"... current nosologies and diagnostic nomenclature are not only useless but restrictive andobstructive.This does not mean the discarding of useful terminology or syndrome appellations.Torefer to a constellation of symptoms as constituting a schizophrenic picture is very different fromreferring to the individual presenting these symptoms as a victim of "schizophrenia" or as being "aschizophrenic." Somesymptomsare by definition "schizophrenic," but no patient is. The samepatient may present another syndrome tomorrow." (Menninger, 1963, p. 33)

Furthermore:"The notion that there were disease entities which could be discovered and defined and delimitedand confirmed by various tests - this notion set psychiatrists off on one kind of wild-goose chase.One name after another was applied to the special proprietary delimitations of some highlyarticulate or compulsive describer. (...) The failure to recognize the essential characteristics of

379. e.g. American Psychiatric Association (1987, 1994). See e.g. Reid & Wise (1995) for a brief overview of DSM’s history.

135

mental illness persisted in spite of a succession of categorical name changes. "Possession"(demonologic) became "bewitchery," "bewitchery" became "madness," "madness" became "lunacy,""lunacy became "insanity," "insanity" became "psychosis," a word many of us feel to be no morescientific than the word "bewitched."" (Menninger, 1963, p. 29)

More regretfully and less impassionedly than Menninger«380», Kendell maintains the same thing.He points out both the desirability and the improbability of finding such disease entities. Hede factotakesa nominalist position, but only because he considers realist conceptions as regrettably too far-fetched:

"... we are trying to put our [conceptual, nosological] boundaries where nature has hers" (Kendell,1975, p. 67),

and:"It is probably unlikely that any of the major syndromes of psychiatry will ever prove to be anentity in any strict sense«381», but if one did it would be a matter of unusual importance."(1975, p. 60)

and:"I hope (...) that it will be possible for everyone with an interest in clinical research to accept theneed for operational definitions of diagnostic and other key technical terms without feeling that bydoing so they are subscribing to an alien ideology«382». For my own part I regard thepublication of the Feighner criteria«383»as a major technical advance but that does not meanthat I regard myself as a "neo-Kraepelinian," or subscribe to all of the nine tenets (...)." (Kendell,1982, p. 12)

Clearly, it is the practical usefulness of terms and their operational definitions that counts forKendell«384», and hence their implications for practice. This nominalism in psychiatric diagnosisdoesnot construct the patient as a meaning assigner. Rather the clinician defines first of all himself as theone who does the meaning assignments. The ’illness’ that the clinician decides to be applicable, isassigned to the patient’s symptoms as a meaning that is assigned to symbols«385». The meaningsthat the patient himself assigns to his environment are taken into consideration by the clinician only assymptoms additional to the behavioral ones that the clinician has already at his disposal

The DSM is an attempt to construct a set of critical objects, the symbols of which are the concretelyperceivable behaviors (called ’criteria’ in the DSM jargon) about which agreements can be attained. Thepsychiatric classifications that can be composed out of these symbols are intended as a set of pre-defined

380. Notice that nothing has been maintained here about the effectiveness of Menninger’s approach. He is at times even criticizedfor having too much of a theory and too little of a practical outcome.

381. cf. also Van den Hoofdakker’s (1978, p. 251) skepticism on the possibilities to delineate ’depression’ from ’normality’ or fromother psychiatric syndromes

382. Kendell is pointing in these lines at Klerman’s (1978) nine articles of the ’neo-Kraepelinian credo’.

383. As formulated by Feighner, Robins, Guze, Woodruff, Winokur and Munoz (1972). These criteria of psychiatric disorders weresources of inspiration for those developed in DSM-III. "The Feighner criteria are, in the simplest of terms, a set of definitions of alreadyaccepted concepts which are "operational" in nature, and which permit common language and common criteria to be applied in thedescription and classification of patients and their disorders." (Katz, 1982, p. 9)

Notice that Klerman (1978, p. 113), on the other hand, considers the development of the Feighner et al. criteria as a strategy byneo-Kraepelinians to improve their realist programme.

384. Kendell (1975, p. 17) also approvingly mentions Linder’s point of view: "As Linder has put it, a diagnosis is not a description,an explanation, or even a prediction, it is simply a disguised prescription [viz. for therapeutic action] (Linder, 1965)". Also Giel (1982,pp. 98/9) emphasizes the primacy of treatment options for psychiatric diagnosis.

385. Cf. Foulds (1965, as mentioned by Giel 1982, p. 89), who speaks about the ’signs’, to which the clinician assigns his own(diagnostic) meaning.

136

meanings«386»assignable to these symbols, in order to facilitate the choice of a particular treatmentpolicy«387», not in order to explain the diseases etiologically. For that reason, they are usually notcalled ’diagnoses’. Depending on the various behaviors and experiences that the patient exhibits orreports, the clinician may conclude that the patient’s phenomena meet a sufficient number of criteria tocall them by the name of a particular syndrome, such as ’dysthymic disorder’. Likewise, events from thepast are not considered as essences or ’background knowledge’, but only as behavioral criteria like thepresent-day ones. For instance, DSM-IV defines ’antisocial personality disorder’ in rather present-daybehavioral terms, with some reference to behavioral patterns before the age of 15 years, whereas DSM-III-R extensively explored the patient’s behavioral patterns before that age«388».

Such a diagnostic classification, then, is not a statement about what is ’really’ the case. Rather, it is asummary of a behavioral pattern or syndrome, that can be used a label, as a pragmatic pointer to aparticular treatment or treatment policy. For example, ’dysthymic disorder’ implies a treatment that isentirely different from ’major depressive disorder, single episode’, or from ’adjustment disorder withdepressed mood’.

Due to the fact that various realist schools contested in the past about the explanation and ensuingtreatment of psychiatric phenomena, the nominalist discourse met the need for a commonground«389», a ground about which no disagreements existed on matters of theoretical orideological kind«390». Thus a nominalist discourse could arise even if the researchers’ originalintentions were realist. The empirical investigations in diagnostic classifications may or may not havebeen initiated for the purpose of underpinning neo-Kraepelinian credo articles (as Klerman maintained);once initiated as a search for ’objective facts’, they got their own dynamics according to which the neo-Kraepelinian and other realist assumptions lost their privileged position.

Thus, Spitzer (1982) publicly withdrew from neo-Kraepelinianism because he preferred being ’data

386. One of the critics of DSM-III even fulminated that it was "the outcome of large-scale committee work, designed not so muchto ascertain facts and to arrive at the truth, but rather to reconcile different power groups and pacify semipolitical Tammany Hall-typeorganizations whose influences are incommensurate with their scientific status." (Eysenck, 1986, p. 74)

387. Cf.: "The definitions and classification of all illnesses are in a sense arbitrary; the definitions of diseases and illnesses areconventions, constructs, abstractions that do not correspond to anything in nature. They are justified by their usefulness." (Skodol, 1989,p. xv).

Cf. also:"Accurate diagnosis is the key to effective treatment."(Reid & Wise, 1995, p. xxvii)Also Giel formulates the issue of treatment policy on the basis of diagnostic classifications: "... that the diagnostic label does

not pretend to give an exhaustive description, nor an individual emblem. To the contrary, it is concerned only with particular features orproperties that make [the client] belong to a group of others, so that, as a result, general statements about his problem become possible."(Giel, 1982, p. 44, my translation)

In line with this is the fact that many present day mental health institutes (and health insurance companies not in the least) useDSM classifications for deciding the type of procedures to be followed (and paid), whether and how a person should be referred, whatkind of treatment he may need, what kind of procedures should be avoided, etc. In short, the DSM is used as a policy tool in a medicaltechnology environment (cf. Gutsch, 1988; Perry, Frances, Clarkin, 1990; also Millon, 1986, pp. 66/7). See also section 8.4.

388. E.g.: "You’ve said that you ran away from home and stayed out overnight before you were 15[Before you were 15, did yourun away from home and stay out overnight?]Was that more than once?(With whom were you living at the time?)"(Spitzer, Williams, Gibbon, First, 1990, p. 208)

389. Cf.: "Psychobiologic research, which had had new life breathed into it by the introduction of the major tranquilizers in the1950s, was in serious need of objective, scientific classification systems for the mental disorders." (Katz, 1982, p. 9);

and: "Psychiatry’s crisis was caused by the repeated demonstration that the reliability of psychiatric diagnoses was horrifyinglylow, that key terms like schizophrenia were being used imprecisely and indiscriminately, and that in consequence fruitful research wasseriously handicapped." (Kendell, 1982, p. 11)

390. cf. Giel: "... to attain a classification of behaviors and experiences that fit into each [theoretical] model and that have apredictive or prognostic value with respect to the outcome of the observed disorder..." (1982, p. 63, my translation) ["... te komen toteen classificatie van gedragingen en ervaringen die binnen elk model passen en een voorspellende of prognostische waarde hebben voorde afloop van de gesignaleerde stoornis"]

137

oriented’«391»to some of Klerman’s articles of faith«392». Conversely, Spitzer approvinglyquotes (his) DSM-III«393»:

""In DSM-III there is no assumption that each mental disorder is a discrete entity with sharpboundaries (discontinuity) between it and other mental disorders, as well as between it and NoMental Disorder." This is an empirical issue and the available evidence supporting discontinuityamong the functional disorders, such as Affective Disorders and Schizophrenia, is far less thancompelling." (Spitzer, 1982, p. 592)

In order to prevent the type of diagnosis that relies heavily upon theoretical concepts, and even worse:upon ill-defined and vague concepts, and in order to prevent the quarrels between followers of one orother theoretical school the DSM intended to offer a taxonomy of disorders that was to be a purelydescriptive system of syndromes, clusters and patterns of behaviors. Thus, the nominalist discourse thatunderlies the DSM is defended as something that prevents ideological debates and unresolvable empiricalissues, as well as unprovable convictions about mental life. In fact the DSM has been tailored for thisparticular task.

For example, it was by emphasizing the poor empirical status of the concept of ’neurosis’, and itsprimarily theoretical signature, that Spitzer and his co-workers managed to keep it out of the DSM-III «394». For the DSM was to become a platform of notions with a minimum of theoretical (i.e.:disputable) burden, and a maximum of empirical (i.e.: undisputed) status. Skodol (1989, p. 45), one of theleading authorities on DSM, for example calls the DSM-III a principally atheoretical diagnosticsystem«395». However, though the theoretical-descriptive controversy has taken much attention, it isto my opinion not basic to the battles in psychiatric diagnosis«396». Basic is the incompatibilitybetween explanation and treatment, which does not come to a fruitful solution.

391. cf. the fragment from Kendell (1975, p. 21) which I will quote below in section 8.4.4, (p. 140) in which he calls his ownposition ’wholeheartedly empirical’.

392. cf. footnote 371 on page 132

393. Cf. also Reid & Wise (1989, p. 13) who admit (with respect to DSM-III-R) that: "The current state of diagnostic art, and thenature of patients themselves, precludes clear-cut borders between closely related syndromes and disorders, and sometimes betweennormalcy and psychopathology"

394. Cf.: "...neurosiswas an etiological rather than a descriptive concept. It assumed (...) an underlying process of intrapsychicconflict resulting in symptom formation that served unconsciously to control anxiety. However, there was no empirical basis forassuming the universal presence of such conflict in those disorders that had traditionally been termedneurotic." (Bayer & Spitzer, 1985,p. 189)

395. Faust and Miner (1986) compare such a conviction to the position of the early empiricists who followed the ideas of FrancisBacon: "Early empiricism and Baconian doctrine strove toward minimizing or eliminating presupposition, unnecessary theorizing, andhigher-level abstraction in order to make science factual and objective. The scientist was to concentrate study on things that could bedirectly observed. Data were to be collected and analyzed in as objective and rigorous a manner as possible, avoiding the biasesengendered by metaphysical assumptions, so that nature could reveal herself." (1986, p. 962)

Cf. also Wallace’s critical description of the DSM: "... the philosophy of DSM-III: that it is possible to have theory-free factsand that DSM-III has realized this feat" (Wallace, 1988, p. 138), to which Wallace comments: "both are assumptions of the firstmagnitude" (ibid.)

396. see section 8.5.1.

138

8.4. Four implicit contaminations of the nominalist diagnostic project

The phenomenon that I will be concerned with is that of a switch (hidden or not) from a nominalist to arealist discourse (cf. section 4.3.4). This entails a reification of names into things, symbols intoappearances, meanings into essences.

An everyday life example of this can be found in mental health institutes in which the DSM terminologyhas been adopted«397»by enforcing it upon the employees. Officially advanced«398»for thesake of uniformity of usage of terms between employees, it is not too difficult to recognize in this policyof the institutes an urge to have a clear and clean assignment of diagnoses (meanings) to the variousbehaviors (symbols), so that a maximum of reliability can be asserted to the institute’s environment, not inthe least to its financial support environment«399». This desire of being accountable and reliablequa institute, however, does not always correspond to the operational desires of the institute’s employees,the practitioners. The DSM culture has too often been imposed from above upon the staff members, andas a result they pseudo-adapted to the new jargon, meanwhile sticking to their old considerations ("doesthe patient have a major depressive episode or does he have an adjustment disorder with depressedmood?").

The nominalist diagnostic classifications may easily become (and in fact have been) contaminated byhidden realist undercurrents. I will discuss several of these, ranging from rather benign to rathermalignant, if the reader will permit me this usage of terms. The most ’malignant’ contamination, whichcomes close to identifying the realist and the nominalist discourses, is a good example of the ’classicalsolution’«400»of science, as it has proceeded most fruitfully since the Enlightenment. In terms ofthe above mentioned metaphor: these contaminations are like the ways in which a historical or culturalmonument becomes appropriated and ’recreated’ by the tourist industry. But a more substantial answer tothe ’why-bother-question’ will be given (see section 8.5.1). First, however, let us look at some of thevarious contaminations that may happen to a strictly nominalist approach to psychiatric symptoms. I willpresent them in a sequence of decreasing innocence.

8.4.1. Contamination: diseases as gestalts

In the first place, due to consensus between clinicians about the diagnostic criteria of particular mentaldisorders, these disorders become gestalts, the elements of which (the symbols) disappear out of sightafter the gestalt formation has taken place. Thus the disorders come into existence as new naive objects.As a result, clinicians feel free to speak about a particular personality disorder as something that aparticular person ’has’. As such, this kind of reification of concepts is to some extent unavoidable andalso rather innocent, but its innocence may be the precursor of more serious ways of switches intorealism.

397. in the Netherlands since ± 1988

398. Cf. for example the way in which a major Dutch mental health association (NVAGG) defends the usefulness for practitionersof the DSM-IV: "the internal communication is improved between the person who does the initial interview and the person who doesthe treatment: standardization of diagnostic procedures yields a unity of language and facilitates transfer. There is a better externalcommunication, between [the mental health institute] and the persons who referred the patients" (NVAGG, 1995, p. 5)

399. cf. Wilson, 1993, p. 403: "The [psychiatric] profession (...) was asked to become more accountable for its practices."

400. see section 6.1

139

8.4.2. Contamination: conceptual elegance as facticity

In the second place, elegance of ideas may be mistaken for confirmation by empiricalevidence«401». Elegance is ana priori property of conceptual systems, entailing such features asconsistency, parsimony, clarity, etc. Elegance, therefore, is a formal quality of a concept or of thepresentation of a concept; it is not a matter of facticity, and, like a teapot that is elegant in its design, anelegantly defined concept may turn out to be highly troublesome in practice. On the other hand, thevalidity of a conceptual system is ana posterioriproperty of the concept, dependent upon empiricalfindings. As a matter of wishful thinking, therefore, conceptually elegant constructs may become believedto be ’real’. To put it metaphorically, it would be pointless to believe that the characters in a novel mustbe (or must have been) really existing human beingsbecausethey are described so vividly. However, as acase of reification it is still rather innocent.

8.4.3. Contamination: statistical clusters as disease entities

In the third place, the status of empirical evidence may be misunderstood. Nominalist diagnosticclassifications have been constructed by consensus both on the definition of terms and on theinterpretation of empirical data. These data are usually processed by statistical tools such as clusteranalytic techniques. The clusters thus found are just patterns of data bearing a name, but one mightmistake them for the empirical proof of the material existence of a particular disease. Here reificationtakes a more serious form. By emphasizing the absence of theoretical (e.g. psychoanalytic) terms in thedescription of these syndromes, a certain realism may nevertheless slip in unwittingly. A particulardysfunction or disorder is considered as a really existing disease entity, and the more so to the extent thatempirical data can be effectively clustered by some technique or other. The statistical cluster has becomea new kind of disease entity«402».

8.4.4. Contamination: meanings as essences, symbols as appearances

As stated above, a nominalist diagnostic classification, say ’bipolar disorder’, is to be understood as ameaningwhich for treatment purposes has been assigned by clinicians to a patient’s behaviors (symbols).Though these symbols are meant as diagnostic criteria for the decisions about which treatment is fittestwith respect of the patient’s modes of functioning, they could be mistaken for the appearances (symptoms,expressions) of some disease state (essence). It is precisely this confusion that is of our interest here,because it is crucial in understanding the connection that we expect to exist between theoretical (causal)explanations on the one hand and on the other hand meaning assignments (such as practical applications).It is the fourth and final type of contamination to be mentioned here.

The history of science contains many examples of abstract concepts that start as names and labels, andthat eventually are turned into essences. Such has happened for example with the concept of ’virus’,

401. At this place it may be noted that Hirs (1987), in a PhD thesis on the construction of standard classifications, found an inherentlimitation to our capacity to make classifications exhaustive both in regard of their intensions (conceptual demarcations) and theirextensions (enumeration of empirical referents). Should a classification of concepts give priority to what is meant by its designers inabstracto, or to what these concepts refer to in concreto? We may notice that this dilemma is the same as that which underlied Bohr’sprinciple of complementarity, viz. a tension between observation and definition (cf. Löfgren, 1990a). See also section 6.2.2.

An example of the dilemma can be found in Menninger’s (1963) extensive appendix on psychiatric classifications throughoutthe ages. He shows how Kraepelin, in the fifth edition (1896) of his book on psychiatric nosology, gives a relatively systematicpresentation of psychiatric diseases, consisting of only four general types (acquired mental diseases, mental diseases resulting from amorbid constitutional predisposition, psychopathic conditions (degeneracy insanity) and developmental inhibitions), each of which hasseveral subtypes. In his sixth edition (1899) however, these conceptually elegant types and subtypes have been replaced by a less elegantbut practically clearer system of thirteen general types, with less subtypes. This, indeed, may be considered as an empiricalimprovement, though at the cost of theoretical elegance. Here conceptual elegance receded in favor of empirical facticity.

402. cf. Klerman’s remarks on the delineations and causes of illnesses (as in footnote 376 on page 133)

140

originally meant as referring to an abstract agent with a toxic or pathogenic effect uponorganisms«403». The ’virus’ eventually turned out to be a concretely observable biochemical entity;it did substantiate«404», so to say. The DSM concepts, however, have explicitly not been designedfor providing explanations«405», at least not according to the initial aims of their makers. As this(originally) nominalist classification system took shape, however, its disorders have unnoticedly becomeregarded as essences (disease entities), and causal connections with their appearances (symptoms) havebeen sought in a variety of ways.

Indeed, the assumption that the diagnostic criteria can be taken as symptoms, i.e. that the symbols can betaken as appearances«406», corresponds to a basic tenet ofclassical science, one that is only validif a calculus provides the connection between previous and subsequent states. However, it is anassumption that does not hold in the field of diagnosis and treatment of mental problems.

Kendell discusses this assumption in terms of an inherent limitation of our conceptual abilities:"Our present outlook is so wholeheartedly empirical that we find it difficult to credit how an earliergeneration could have talked of diseases being ’discovered’ like so many golden sovereigns on abeach, or have imagined that there were a finite number of them waiting to be identified. Yetalthough we know these things perfectly well, we have still not rid ourselves of all the old Platonicassumptions. Claims are still made even now that this or that syndrome is a ’disease entity’. (...)Similarly, the suggestion that patients with, for example, both schizophrenic and affectivesymptoms ought perhaps to be regarded as genuine interforms half way between schizophrenia andmanic depressive illness still meets with expressions of shock or incredulity,as if they must reallyhave one or other illness even if their symptoms are atypical. Part of the problem lies in thetendency of all concepts to become ’reified’ and for familiarity to lead to their origins in humanimagination being forgotten. The concepts of physics are as prone to this tendency as those ofmedicine and it may be that we are up against some of the inherent limitations of our conceptualabilities." (Kendell, 1975, pp. 21/2, italics added)

Instead, Kendell wishes to take diseases only as ’useful concepts’ (1975, p. 22 and pp. 64/5) for thepurpose of making treatment decisions, while not getting trapped in matters of reification«407».

403. a "contagium vivum fluidum", as Beyerinck called it in 1898 (GNLE, vol. 24, 1979, p. 445)

404. But not always do concepts, initially meant as labels, substantiate into explanatory units. For example, in somatic medicine thestatus of some names of somatic disorders, such as ’neurocircular asthenia’ or the more recent ’myalgic encephalopathy’, is stillundecided.

405. unlike Kraepelin’s classification system (e.g. 1899, p. I-13)

406. Skodol, after his remarks (1989, p. xv) on the constructed status ofall diseases (see footnote 387 on page 136), in fact feedsour suspicion by stating: "... the ultimate purpose of a classification is to further understanding of the causes of disorders and theprocesses involved in their development and persistence." (1989, pp. xv-xvi).

Indeed, the search for causal mechanisms behind mental disorders, though in itself a perfectly legitimate field of research, ishardly compatible with the constructed status of these disorders! Cf.:

"Once we have processed the clinical experiences successfully to the extent that we are able to arrange disease groups withparticular causes, particular phenomena and particular courses, then it will be our taskto penetrate into the essence of the individualdisease process." (my translation) ["Ist es uns gelungen, die klinischen Erfahrungen soweit zu verarbeiten, dass wir Krankheitsgruppenmit bestimmten Ursachen, bestimmten Erscheinungen und bestimmten Verläufe aufstellen können, so wird es unsere Aufgabe sein, indas Wesen des einzelnen Krankheitsvorganges einzudringen.] (Kraepelin, 1899, p. I-5; my translation).

407. See footnote 391 on page 137. Notice, however, that even Kendell’s nominalism is not entirely consistent here. Immediatelybefore the fragment quoted just above (1975, pp. 21/2) he writes:

"... it is well nigh impossible for us to regard illnesses as having any sort of independent material existence. To our generationit is self-evident that diseases, tuberculosis as well as schizophrenia, are nothing but man made abstractions, inventions justified only bytheir convenience and liable at any time to be adjusted or discarded." (p. 21)

But to qualify both schizophrenia and tuberculosis as man made abstractions is not only too generous to constructivism, it isalso misleading. For if both are man made abstractions, they are so of an entirely different kind. Tuberculosis is precisely a diseaseentity of the variety that Kendell deems unattainable for most psychiatric disorders (cf. the quotation given on p. 135, from Kendell,1975, p. 60). It has a well defined causal agent, and also a clinical course as well as a treatment strategy that are strongly related to theproperties of this causal agent. By comparing schizophrenia with tuberculosis, the old Kraepelinian realism works its way even throughKendell’s otherwise consistent nominalist discourse! Indeed, it was Kraepelin’s idea to compare psychiatric illness with tuberculosis (as

141

8.5. The desire for etiology: theory and practice

Even if Kendell is mistaken with respect to what he calls the inherent limitations of our conceptualabilities, there are apparently some counterforces at work against the nominalist project; against givingconceptual and empirical clarity the absolute priority over unwarranted causal explanations. I willdistinguish two of these counterforces: a theoretical and a practical desire for etiology.

8.5.1. The theoretical desire for etiology

Reid & Wise (1989, p. 5) suggested the possibility for (by that time) a future DSM-IV to undergo atransition from a descriptive classification system to an etiological one. They hoped«408»that suchan ’etiological turn’ would yield results as objective and descriptive as were the original intentions of theDSM designers«409».

Likewise, Millon writes:"The fact that the [DSM-III] Task Force bypassed etiologic variables so as to minimize dispute andto facilitate acceptance of the DSM-III by clinicians of diverse orientations is no reason to abandonthe search. The DSM-III criteria will facilitate etiologic as well as other research endeavors, andthe data these bring forth should be considered in the reconstruction of syndrome categories.Thereis no intrinsic reasonwhy etiologic variables cannot play as important a role in psychopathologicalclassification as do descriptive and inferred clinical properties." (1986, p. 66; italics added)

To my understanding, however, such intrinsic reasons do exist. The principal of these would hold thatthese etiological variables cannot act as mere ’descriptive criteria’, but will act as essences, and thereforewill contaminate the idea of a descriptive classification. The ’syndrome categories’ about which Millonspeaks would no longer be meanings assignable to symbols, patterns recognizable in behaviors, but ratheressences behind appearances, that is, illnesses seen as causal processes behind the manifest symptoms.

Abstinence from theoretical and etiological considerations had been a successful way to unite otherwiseconflicting and opposing parties. For that reason it has been vitally important for the DSM adherents tokeep the descriptive emphasis and to resist theoretical engagements«410». Equally, it wasconsidered useful to dig up and identify hidden and implicit theoretical assumptions, in order to replacethem by more empirically grounded notions. Hence, the DSM favored and was sustained by debates thatmoved around issues like the implicit theoretical assumptions behind the DSM, such as assumptions

related by Van den Hoofdakker (1978, p. 226), who refers to the 8th edition (1913) of Kraepelin’sPsychiatrie). A similar usage of the’man-made’ argument has been given by Klerman, for the very opposite, realist, sake:

"However, to say that the medical model or the concept of mental illness is a social construct is not to say that it is a myth orthat it is invalid. All social constructs are not myths and they are not necessarily untrue. After all, "the rights of man," the electron, andthe university are also social constructs. They are not facts given in nature, but rather are complex ideas developed by historical forcesand legitimated by consent. The concept of illness is not arbitrary but reflects areas of shared consensus, embodying truths arrived at byrules of evidence". (Klerman 1978, p. 107)

In this way, Klerman attempts to present his neo-Kraepelinian realism as a matter of consensus construction.

408. a hope that Faust & Miner (1986) would certainly call ’Baconian’, cf. footnote 395 on p. 137

409. cf.: "... our understanding of etiologic factors is not yet sufficient to employ them in most DSM-IV diagnoses." (Reid & Wise,1995, p. xxvii)

410. In fact any way of gathering data for problem identification presupposes a particular theoretical point of view (cf. Mahrer,19;89, p. 151). Theory and data cannot be disconnected, as is so often suggested. Rather, as Faust and Miner (1986) put it: "Theorizingand inference have perhaps been reduced [in DSM-III] but have been eliminated nowhere, and regardless of the level of analysisapplied, definitions and terms come with a complex network of stated or unstated auxiliary assumptions." (p. 964)

142

concerning cultural or racial background, gender identity, etc.«411», or about the oppositionbetween psychodynamic and descriptive approaches«412». However, the tension between nominalistand realist discourses was not resolved in the mean time. For in order to survive, the DSM had to remainas atheoretical (’descriptive’) as possible and hence to disregard as much as possible any etiology thatrelied upon theory.

But the principal question was begged: cannot one operate in both discourses simultaneously andcoherently, linking etiological backgrounds from the past, via present behavioral phenomena to futuretreatment policies«413»? Is not this also what patients in psychotherapy sessions are asking from us:to find the causes of their problems in order to deduce the appropriate solution policies? It may have beenfor the purpose of attaining this integration of discourses that the neo-Kraepelinians proclaimed that "thefocus of psychiatric physicians should be particularly on the biological aspects of mentalillness."«414»For in that way some coherence between etiology and treatment practice could behoped for.

Likewise, the search for coherence between explanation and application is made a predominant feature ofVan Praag’s (e.g. 1990) quest for a functional psychopathology. It differs from Klerman’s neo-kraepelinian approach in that Van Praag chooses his units of investigation at a much more elementarylevel than that of the disease entities. Thus Van Praag manages to abstain from many controversies inpsychiatric nosology. In his approach it is attempted to smoothly link the two discourses.Psychopathological symptoms are taken as ’manifestations of an underlying psychologicaldysfunction’«415», and psychotropic drugs are to be administered according to the inferreddysfunction, and not according to nosological classifications. The same is maintained for the utilization ofpsychotherapies«416». Thus, knowing the background of what is the case is purported to entail clearpharmacological or psychotherapeutic measures (that is: to imply logically these measures, not merely toconnect them by cookbook!). Thus the question how to deal with the symptoms (i.e.: the question of theirmeaning) is linked to the dysfunction underlying the symptoms (i.e.: their essence).

However, apart from biological approaches like this, psychiatry is left with a cleft between explanatoryconcepts and treatment implications. The question is not: "how is the solution which is valid for biologicalpsychiatry to be generalized to other domains of mental health care, in particular to psychotherapy?"; butrather: "why isn’t such an elegant integration available for these other domains, an integration that guidesthe practitioner from explanation to prediction/control, and vice versa?".

This question is begged also by the emphasis modern scientific culture puts upon data gathering. In fact,

411. All these discussions are useful in themselves, and even very revealing at times (e.g. Ruijgt, 1991; Walker, 1987) butnevertheless contribute to the fallacy that the diagnostic classification system should be purified from misconceptions and othertheoretical remnants, in order to become truly objective.

412. e.g. Frances & Cooper, 1980. See also the account by Bayer and Spitzer (1985) on the scientific and political battles fought forthe sake of inclusion of psychodynamic considerations into the DSM-III. As the psychodynamic ideas were considered of too littleempirical caliber, they were rejected as grounding the DSM, for their ""inclusion inDSM-III could be viewed as an official endorsementof one school of thought"" (Spitzer, as quoted by Bayer and Spitzer, 1985, p. 194). For instance, etiological notions on internal entitiesthat were said to cause neurotic behaviors were considered too theoretical, and at times even were subtly made ridiculous (cf.: "The taskforce accepted some of Rockland’s descriptive changes, but found without merit his dynamic and etiological recommendations, eg, thatobsessive compulsive disorder represented a "regression to anal conflicts." These proposals were characterized as "simplistic andparochial," as derived from a "set of inferences that have been documented only by anecdotal material"" (Bayer & Spitzer, 1985,p. 190)).

413. cf: "The question, it seems to me, is can we have both? Can we maintain an active respect and investment in descriptivecriteria and reliable data, while including an equal respect for the subjective experiences of psychiatric disorder, for theory andpostulated underlying processes? Failure to use both orientations is perhaps the most unscientific approach of all." (Strauss, 1982, p. 9).

414. This is Klerman’s (1978) sixth article; cf. section 8.2 above.

415. Van Praag (1990, p. 5)

416. Van Praag, Asnis, Kahn, Brown, Korn, Friedman & Wetzler (1990, p. 731)

143

all modern researchers in the field of psychiatric diagnosis consider themselves objective, factual, dataoriented and empirical. All proclaim to be concerned with what is ’really there’, beyond the reach ofideology«417». Therefore, the question is not, as the DSM adherents have it, whether or not one hasimplicit common sense assumptions or other assumptions, presuppositions or prejudices that bias one’sempirical work. Each of us has them, and they can be made explicit in due time. The question thatmatters is not whether one is data oriented, objective, factual, empirical etc. What matters is whether oneis data oriented for the sake of theoretical explanations and underlying principles, or for the sake ofpractical applications«418»! The problem that remains to be solved, therefore, is:how could wecoherently connect the etiologies of psychiatric disorders to their treatment policies?I will deal with thisquestion, not by solving it straightforwardly, but by taking a look at how practice itself manages to dealwith the lacuna between the two critical discourses; and I will find that it is by allowing itself to deviatefrom the realm of critical discourse, at some appropriate moments, and by accounting for self-referentialsituations in terms of the clinician’s own subjective impressions and experiences.

8.5.2. The practical desire for etiology

The diagnostician who uses the DSM is expected not to need explanatory considerations in order todetermine a particular diagnostic classification«419». A patient whose problem is classified as’agoraphobia’ is treated by means of behavior therapeutic procedures; a patient with ’major depressivedisorder, single episode’ receives a treatment based upon psychotropic drugs; a patient with ’bereavement’receives a conversational treatment in which the grief is worked through according to the patient’sreflective capacities, etc. Though the classification does not and cannot certify an optimal treatmentpolicy, its only goal is to pragmatically stratify the decision making procedures that precede the treatment.Nevertheless, practitioners are said to like explanations«420». In this section I will argue for analternative understanding of this liking, namely that it is not so much the explanation that is of thepractitioner’s interest, as the contact with the patient.

8.5.3. Using the patient’s background for explanation versus personal understanding

Practitioners often find the here-and-now descriptive criteria of the DSM too narrow. With respect to thetype of problem that the practitioner encounters, he often seeks for a connection with a particulardevelopmental stage (e.g. a correspondence between neurotic conflicts and the oedipal stage, or acorrespondence between the defense mechanism called ’splitting’ and the stage of separation andindividuation«421»). Information about the patient during that particular developmental stage is thenconsidered highly relevant for understanding his problems.

But is the integration of explanation and application something that practitioners in the field of mentalhealth, both diagnosticians and psychotherapists, are really looking for? Is it the case that they are, like

417. They are looking for "more objectively grounded empirical systems of classification" (Katz, 1982, p. 10). Even if one operatesfrom a psychodynamic or other ’theoretical’ position, one is usually more than interested in the discovery of ’hard’ facts and how theyfit in with the cherished hypotheses.

418. Cf.: "The clinicians (...) need a system which will communicate easily, and which will assist in predicting the results oftreatment. Scientists see classification as an ultimate objective in which the "natural" order is represented and, in the case ofpsychopathology, as based primarily on etiology. These different objectives result in different values as to which studies and whichevidence are worth considering." (Katz, 1982, p. 10).

419. Cf. e.g. First, Spitzer, Robert, Gibbon, Williams (1994) and Spitzer, Williams, Gibbon, First (1990).

420. Likewise, Karasu & Skodol (1980) argued in favor of the introduction of a sixth axis for the DSM on which psychodynamicdevelopments are to be classified, maintaining that such an axis would be more in line with the needs of the practicing psychotherapist(see also footnote 406 on p. 140).

421. These are obviously correspondences of a high theoretical (psychoanalytical) constitution. (cf. Pierloot & Thiel, 1986, pp. 18-20)

144

researchers, faced with a choice between the advancement of practical outcomes, or to that of theoreticalexplanations? The answer, I think, is: no. The practitioner is not at all interested in any distinctionbetween realist and nominalist discourses. He is far more interested in the attainment of apersonalunderstanding of a particular patient’s past, in order to root upon this personal understanding his futuredecisions on treatment options. This personal understanding is a matter of ’getting in touch’ (to speakwith Shotter, 1990) with the subject matter, so that, out of this ’touch’ a treatment can be envisaged. Thepractitioner uses elements from both critical discourses according to his needs, without bothering aboutproblems of compatibility between discourses. In other words, what matters for the practice of diagnosticclassification, as well as for the practice of psychotherapy, is not so much a critical discourse, as adiscourse in which the practitioner is able to get into contact with his subject matter, yielding a type ofunderstanding that can be characterized as personal, due to the practitioner’s engagement. For it is througha process of ’getting in touch’ that the practitioner comes to terms with the patient and his problems, andbecomes capable of ’grasping’«422»them. Getting into the history and origination of the patient’sproblems, then, is not so much a matter of chasing after an etiology, as it isa search for personal contactand personal understanding, with the purpose of creating a shared horizon«423», and, hence,together creating naive objects.

For example, a 35 years old male university teacher comes and presents his problem in terms of’feeling dead, feeling dumb, not being convinced that there is anyone interested in his presence’.Furthermore, he complains of his wife being annoyed by his manners, which, as he admits, lack thequality of desire for her. He says: "I don’t know what desire is, though I like having intercourse attimes". In order to better understand these complaints, the diagnostician would explore his past, hisyouth, and look for points of connection. Then it is found out that this man was a child of richparents who preferred not to spend too much time with him, leaving him most of the time with amaid who was to take care of him. He is willing to mention the existence of strong feelings whenhe thinks of his early youth, but, he mentions in passing, these are of nobody’s concern.

Here the clinician may experience the social appeal that the patient is implicitly demonstrating by showinghis loneliness and his sense of getting stuck. The clinician might discuss these feelings, but, first of all, hewill feel some of what the patient is conveying: perhaps a sense of pity or compassion, a sense ofdespondency about this man’s early years. What exactly it is that he will experience, cannot bedetermined in advance. But my point is notwhat but insteadthat he will be touched somewhere,emotionally affected, andfeel the patient’s actual problem as his own, be it for a brief moment. This isthe phenomenon of projective identification (Ogden, 1982). It is here that the clinician ’gets in touch’ andmay perceive the patient’s problem as anaive object.

In other words: when in a conversation a shared perspective is attained, and when the clinician is involvedin the patient’s problems through projective identification or otherwise then a naive discourse opens upbetween the two conversation partners.

Another example may be useful. An unmarried woman in her early forties complains of not beingable to cope with (as she calls it) "social situations". She doesn’t want people to become involvedwith her, and at her work site she attempts to avoid informal encounters, especially those situationsin which persons may come too close to her. After having been forced to work at a different site ofher company, she collapses, as she cannot keep other people at a large enough distance. As aconsequence, some persons show their concern, both personally, such as colleagues, andprofessionally, such as the chief of her work section. But these expressions of concern make hereven more uptight. During the first screening conversation the clinician asks whether hisquestioning is also experienced as coming too close. "No", the patient says, "because you don’treally need something from me for yourself". "Would it work on your nerves if I asked you to do

422. cf. von Foerster’s tesseract perception, section 4.2, esp. footnote 135 on p. 58.

423. ven if this shared horizon of personal contact is afolie à deux, if reality testing is neglected for the sake of consensusmaintenance, even then the interpersonal understanding will yield at least some insights that cannot be replaced by a more detached anda more empirical approach.

145

me a favor, e.g. working in my garden?" the clinician asks. "No" says the patient, "but you wouldif you told me you knew [follows the name of her work section head] personally". "Oh yes", theclinician replies, "that would be most unpleasant". The conversation then continues with a searchfor other, perhaps earlier situations known to the patient in which she was faced with people whocame too close. The patient mentions a few, and eventually relates some very early events in whichshe had repeatedly been terrified by her father in a way that seems quite menacing also to theclinician.

It is easy to infer from this vignette the existence of an early trauma that is to account for the patient’scurrent problems with persons who come too nearby. But the point is that this patient’s story about herfather evolved out of an exploration of what would be an unpleasant intrusion within the presentinteraction with the clinician. This is not merely a matter of subtle interviewing techniques; it means thatthe clinician is introduced to some of the patient’s backgroundsthrougha personal encounter with thepatient, especially through their shared fantasy about how frightening an acquaintance of the clinician withthe patient’s head would be. It is in this shared fantasy that they develop a naive object, and it is fromthis naive object that the clinician not only obtains entrance into the patient’s history, but also learns toappreciate this historyfrom the naive object as a point of departure!

Indeed, to perceive the patient’s problem as a naive object, while seeing through (from) the presentphenomena back into the patient’s past, is a widespread practice«424». Suchhineininterpretierenissurely different from unprejudicedly finding a (causal) explanation of a particular trauma, but, unlikecritical discourses, it does provide an integrated view that the practitioner can embody.

8.5.4. A misplaced preference for realist approaches

In the practice of a diagnostic process a great appreciation is given to the patient’s individual history,especially to the origination of the problems, whereas in a nominalist classification system thesedevelopmental notions are largely disregarded, in favor of current behavioral features (the ’criteria’), orthey are constructed as ’past criteria’«425».

It is most tempting, but erroneous, to understand the practitioner’s appreciation of developmentalinformation in terms of a preference for realist positions. His disapproval of a strictly nominalist approachdoes in itself not imply any realist sympathies. A major difference with a realist position is that thepractitioner does not regard the patient’s background, his development and pathogenesis as abstractentities. That is to say, he may be quite capable of doing so, but his diagnostic and/or therapeutic tasks donot push him into that direction. Rather, they require his commitment to developmental and historicalissues, his capability to have them appear in the immediate present, as features of the patient that can begrasped and that also affect the practitioner by their immediacy. This is also how in a team ofpractitioners good diagnostic decision making takes place: by working towards a naive perception of theproblem, the latter is felt to become one that is alive and immediately present to those participating in thediscussion.

Notions about the patient’s background, development and pathogenesis often even exist as implicitfeatures of the practitioner’s performance, and they are useful in bringing him into contactwith thepatient, instead of having him reasonabout the patient. Gadamer’s idea«426»that the spectator of a(theatre) play participates in the play, applies here! For the spectator not only enables the actor to performmeaningfully, but also actively engages in following the movements of the actor, thus supporting the actorto bring into existence the character he is playing.

424. One is reminded of Wittgenstein’s ideas on looking into the future when watching the sky (see footnote 132 on p. 57).

425. cf. footnote 388, p. 136

426. Cf. Gadamer, 1986b, p. 28 (1977, p. 37). See the quoted fragment from this work, above in section 3.2, (my) page 45.

146

Now when a practitioner is thus involved, he will not be too willing to have his performance guided bybehavioral criteria (as those of the DSM), especially if it is prescribed that theprocessof diagnosticdecision making is to follow the course of a ’structured clinical interview’«427». In those cases it istempting for practitioners to oppose these prescriptions by showing howsuperficial«428»thebehavioral criteria are, and by claiming instead that criteria at a much more profound level of humanperformance are to be considered, viz. those that constitute the deeper levels, indeed the essences, of themental disorders. It is here that those who advocate less superficial diagnostic criteria, are trapped in arealist discourse.

Precisely this strategy was attempted by those psychoanalysts who opposed to the construction of theDSM-III. They also ended up by countering the nominalist classification system with a realistdiscourse«429»! It was a realist discourse that the psychoanalytic opponents were advancing inorder to fight the nominalist classification builders. And they lost their battle on the concept of ’neurosis’,as they lost other battles for realist notions.

8.5.5. A process oriented argument against the DSM?

These analysts might have made a better case, had they not sought recourse to the realist discourse andclaimed that a lot of essential features were overlooked by the superficial behavioral criteria of the DSM.For these claims were all rebuffed as loaded too much with unproven theoretical assumptions. Instead,they might have done better, perhaps lost better, perhaps won, had they pointed at theprocessof decisionmaking in diagnostic affairs. For this process contains many moments that are of an explorativenature«430». Such explorations presuppose a naive observer, one who is unaware of hisobservational activitieswhile he performs them, as he is absorbed entirely by the object of study that, dueto his perceptual engagement, emerges to him as a gestalt«431».

There are two major points to be made here. First, by focusing upon the interpersonal encounter as acategory of reality, we find a new way of understanding essences, that does not lead to a realist discourse.For an essence of human performance, also of the mental functioning of a psychiatric patient, then, is tobe sought in the moments of experiences and perspectives shared between persons«432». Such kindof essence does not exist as a causal background, but rather as a realm of ’enigma’ and ’pre-being’«433», a realm of paradox and interrogation«434». This realm is not too open for

427. e.g. First, Spitzer, Gibbon, Williams (1994), Spitzer, Williams, Gibbon, First (1990).

428. cf. Bayer & Spitzer (1985, p. 189)

429. Cf. Bayer & Spitzer, 1985. This is not to say, however, that psychoanalysis itself is of a realist identity per se. Rather, it is myimpression that the battle for including in the DSM concepts like ’neurosis’ drove psychoanalysts into realist positions! However,psychoanalytic jargon at times does show up of a realist signature, as in: "internalized object relations constitute substructures of theego, substructures that are, in turn, hierarchically organized" (Kernberg, 1984), as quoted by Volkan (1988, p. 27). Schafer (1976)explicitly takes the reification of psychoanalytic concepts as an issue of his criticism.

See also Grotstein’s (1998a, p. 60) criticism of the ways in which psychoanalytic theorists have dealt with the concept of’drive’ as a source and agent of human action.

430. Reid & Wise (1995) give a two-paged chapter on ’diagnostic process’, in which one looks in vain for remarks on thesemoments of exploration. In chapter 9 I will elaborate on the concept of ’exploration’.

431. In a similar vein, Gadamer (1986b, p. 25/1977, p. 33) discusses the process of perception of a work of art in terms of theobserver’s coming to see it as a "hermeneutic identity". See also section 3.2 and section 9.4.

432. Buber (1948) presents the inter-individual interface as an ’Urkategorie’ of human reality (p. 165): "... the essential [of a realconversation] is not enacted in the one and the other participant, nor in a neutral world that comprises both of them as well as all otherthings, but between them, in the most exact sense, as if in a dimension accessible to both of them only." (my translation) ["... dasWesentliche [von einen wirklichen Gespräch] vollzieht sich nicht in dem einen und dem andern Teilnehmer, noch in einer beide und alleanderen Dinge umfassenden neutralen Welt, sondern im genauesten Sinn zwischen beiden, gleichsam in einer nur ihnen beidenzugänglichen Dimension."] (Buber, 1948, p. 166)

433. cf. Merleau-Ponty (1960a), p. 9 and 8 respectively

147

empirical investigation, as it is demarcated by the limits of our capacity to understand, but its existencecan at least be sensed by participation. The mode of participation the clinician finds himself in with aparticular patient is a source of diagnostic investigation of the clinician himself, as well as of theinteraction patterns that gave rise to that particular mode of participation.

In the second place, it is to be noted that the exploration of a particular patient’s problems amount to alarge degree to the exploration of the clinician’sown interaction opportunities with this patient, to theclinician’s capacity to experience himself in contact with the object of his explorations. This is an activityin which the process of exploration senses itself. We may call this configuration of self-related explorationplay. Play, as a paradigm of exploration, consists of playing with one’s very opportunities to play! Play isself-referential in this respect«435». And it is due to this property of self-reference that theseexplorationscannotbe formalized in terms of a procedure! In other words, processes of explorationcannot be replaced by a step by step procedure, without destroying their explorative nature. In fact theexplorative process may yield findings that could not have been attained at all by means of a step by stepprocedure«436».

These two points complement one another:if essences are situated at the interpersonal interface, notinside the patient but in between the patient and the clinician, then the process of exploring these essencesmust contain at least some aspects of self-reference!In chapter 9 I will elaborate this idea of self-referential exploration processes.

In line with these ideas, we may imagine how an argumentation may have looked that opposed theperceived superficiality of the nominalist decision trees and other classification prescriptions, withoutbeing trapped by realist notions:

"What we are doing with our patients when diagnosing and treating them can always be put interms of a step by step procedure (such as a decision tree), but only afterwards, notbeforehand«437».In our work we need to disregard at certain moments the distinction between content and form ofthe phenomena we meet, that is: between thethingsperceived and theways in which they areperceived by us. This non-distinction happens at moments of our personal involvement with theproblems at hand. At these moments there is not a step by step procedure, but afterwards, evenshortly afterwards, this personal involvement can be ignored, and the diagnostic judgment formedcan be presented to others as well as to ourselves, as if the personal involvement had not beenthere; as if the judgment had been based solely upon the presence of objective behavioral criteria,and upon the selection of a path through a decision tree. However, if there do exist correctpsychiatric diagnoses that nevertheless cannot be accounted for in terms of a path through adecision tree of behavioral criteria, then we should find ways to save them from dismissal. Thepoint is: such diagnoses do exist. For whenever a diagnostic process becomes self-referential, as forinstance when playful exploration occurs, there is no step by step procedure that is followed, norcan one be constructed that can replace the exploration process.

In other words, when we account for our diagnoses in terms of criteria like those of theDSM, the accounts we are supposed to give ignore the explorative nature of our inquiries,especially the moments we experience the patient’s problems through our own senses, as a shared

434. Ibid., p. 8. In this respect I follow Madison’s (1981) interpretation on Merleau-Ponty’s ’negative philosophy’, according towhich Being (called ’flesh’) could only be deciphered and inferred indirectly, because Being is ’the invisible of the visible’ (cf.Madison, 1981, p. 195). The type of essence at which I am aiming, then, lies in the middle of a relation between two subjects, whereeither subject by definition cannot enter, but where at least each can participate. See also my discussion of Merleau-Ponty’s concept of’flesh’ in section 3.1, sections 7.4.1 and 7.4.2, and section 11.4.

435. Cf. Gadamer, 1960, pp. 97-98/1986a, p. 92; see chapter 9, especially section 9.4.

436. The idea that every performance that actually happens in a material system can be formally described by a step by stepprocedure has become known as the Church-Turing thesis. I already mentioned above some limits to its validity (cf. section 6.2).

437. I am referring here to the distinction between ’reconstructed logic’ and ’logic in use’, or, equivalently, between ’justification’and ’discovery’ (cf. Kaplan, 1964; Reichenbach, 1938). Critical discourses can be said to exist within a context of justification, naivediscourses within a context of discovery. This will be discussed more extensively in sections 9.1 and 9.2.

148

event; they ignore our personal contact with the patient, the projective identifications and otherintricate patterns between the patient and us. In short, they ignore that for the practitioner therearises an indistinguishability between the things perceived and the ways in which they areperceived! For that reason we find nominalist diagnostic classifications too superficial, and in orderto complement them we are in need of terms and concepts that do pertain to the interactionphenomena mentioned, especially in order to relate our own subjective reflections to our momentsof identification with the patient, which are often moments of naive perception.

Nominalist diagnostic classification procedures (especially the structured clinical interviewprotocols that come with the DSM) aresimulations, not models«438», of real diagnosticprocesses. They ignore the informal and unformalized interpersonal events that make up the essenceof the diagnostic task; they depict the diagnostician’s work as if based solely upon the applicationof a rule for manipulating objective criteria.

Our criticism, however, does not advocate the introduction of terms and concepts that referto a ’deeper’, more internal level of personal functioning of the patient, but it does argue for thedevelopment of terms and concepts that allow us not to equate our work with a simulation of it.First and for all, we need a recognition of the fact that parts of our diagnostic tasks cannot be madeexplicit, in terms of a clear procedure. They are acts of playfulness and wonder, and for that reasonthey are concerned with our own fascination of the moment. It is through these moments that weare able to connect a patient’s past as we have come to understand it, with the opportunities weenvision for his treatment.

The nominalist classification (DSM) has been developed in order to eliminate one type oferror: the acceptance of a wrong diagnosis due to an overestimation of subjective impressions. Thishas been and still is a widespread error. But another type of error is overlooked: the rejection of acorrect diagnosis due to an overestimation of objective procedures. For objective criteria cannotalways be provided for correct diagnoses! Reality of diagnostic practice is more complex than canbe formalized. In many instances psychiatric diagnoses do apply to a particular patient (forinstance, they are not mere projections of the diagnostician), even if it is impossible to formalize orto reconstruct them as the conclusions of an argumentation."

For instance, the concept of ’neurosis’ would orient the clinician, beyond any realist interpretation, to lookfor manifestations of the patient’s problems from three different sources: in terms of how the patientbehaves with the clinician, in terms of how he reports to behave with others, as well as in terms of earlyexperiences (cf. Malan, 1979). Indeed, the concept of neurosis would orient the clinician towardsconflicting themes in and between these various sources, and also towards moments of fusion andconfusion between them. As a result, the practitioner can become witness of early experiences, as well ascurrent relationships with other persons, factually situated beyond the conversation setting, like a spectatorbecomes involved in the play he sees unfold on the scene of a theatre, and become capable of (naively)looking through the eyes of the characters of the play«439».

Afterwards, e.g. at the time of reporting about the patient, the diagnostic performance can be recaptured interms of a particular procedure. Then the practitioner becomes critical again, looks back, and attempts toappreciate what has happened. The exploration process can be reconstructed afterwards, as if it had beenonly a search through critical objects, such as patient behaviors and their meanings, or their explanations.Any justification of the clinician’s findings will certainlynot be about the diagnostician’s own subjectiveassessments of the patient, his countertransferential responses, set aside his difficulties of keeping his

438. A truly rigorous treatment of the differences between models and simulations can be found in Rosen (1991). His major point isthat in a simulation there is no mapping relation between the process that generates the simulation phenomena and the real process thatgenerates the phenomena simulated. In a model, on the other hand, such a relation does exist. Similarly, the structured clinicalinterviews of the DSM do not reflect any of the exploration processes that are generative of diagnostic judgments in practice. Instead,they are fully focused upon the logical construction of diagnostic outcomes, and they suggest an atmosphere of collaborative inquirybetween diagnostician and patient, that ignores any of the interpersonal tensions that are intrinsic to this type of interviewing, and that,in fact, very often make up the most precious source of information to the clinician.

439. Notice that both the practitioner and the theatre spectator engage in ’looking from’ their actual experiences into another domainof events (for the former: e.g. a particular patient’s childhood; for the latter: e.g. the imminence of a particular disaster). See section 4.2on Shotter’s notion of ’knowing from’.

149

mind clear during the sessions. Rather, justifications will take the shape of theories, even theoriesconstructed ad hoc, about the importance of particular early experiences, or about the existence of internalmental entities.

The practitioner encounters the patient, and is able, from this encounter and due to it, to perceive thingshe cannot justify in critical terms. Likewise, when scientists attempt to reformulate these encounters interms of information processing, the gestalt qualities of the naive objects disappear.

Thus, one might be inclined to call naive perception a bridge or an integration of realist and nominalistdiscourses, but that would not be a correct formulation: it is precisely due to theabsenceof criticaldistance that a practitioner may combine past and future, history and treatment, in his naive approach ofthe patient. By becoming naive, the observer is begging the problem of combining the two criticaldiscourses, rather than solving it. If we are to find an integration of realist and nominalist discourses,then, we will have to take into account the particular importance for practitioners of their naiveperceptions! In other words, if a bridge between realist and nominalist discourses is possible, it will notconsist of a panacea called ’naive discourse’, or even worse, an escape into naive discourse, but rather ofa critical use of naive discourse, one that takes into account that naive discourse can supply moments ofconjunction between realist and nominalist discourses,without these moments themselves being thepractitioner’s supreme way of performance.

Artists too make use of a capacity to integrate both naive and critical perceptions, during their acts ofperforming or creating. Arieti (e.g. 1975) has described this in terms of a synthesis between primary andsecondary processes, which he called ’tertiary processes’. Primary processes go by associations, analogiesand metaphors. Secondary processes follow ’rational’ rules of argumentation, such as those of predicatelogic. Typical of a synthesis between the two processes is that the artist manages to have one withoutlosing the other, e.g. a poet may dwell in metaphoric language, without losing sight of its metaphoricstatus.

In a similar vein, one may consider also for psychiatric diagnosis a critical use of naive discourse as acase of tertiary processes, where a practitioner combines his moments of immersion and engagement witha more critical awareness of the patient’s behaviors. In fact the practitioner may manage to use a naivediscourse as a hinge between two otherwise incompatible discourses«440». A full simultaneity ofnaive and critical performance is impossible, but at the time of accounting for one’s diagnostic findings, itis possible, afterwards, to recognize, next to more critical considerations, the role of one’s own personalengagements.

This can be done more or less substantially. A least substantial recognition would be the mere mentioningof the fact of naive discourse as having occurred in one’s considerations. A more extended critical use ofnaive discourse would entail a more detailed description of the facts of the exploration processes. Finally,a most comprehensive way of critically using the occurrences of naive discourse would entail an explicitreflection upon the meanings and other properties of the exploration processes, to the extent of enteringinto reflections on the currently unfolding explorations«441», and to describe how and in whichrespects the naive discourse facilitates points of view otherwise unattainable. This may put into the focusthe self-referential qualities of the object dealt with. In order to document this, it may be possible todeliberately lead one’s audience (e.g. the readers of an account) into an experience that cannot becommunicated but by having the audience have it qua experience«442».

440. Buber (1952, pp. 10-11/1957, p. 95) describes this ’hinge’ in terms of a crisis of method (see also my quotation from his workin section 9.2.2).

441. Here the account may obtain aspects of self-reference. Then a so called parallel process may occur, in which the clinician tosome extent relives and rebuilds the sessions’ self-reference. If that happens, the parallel process may even turn into a useful tool forempirical investigations into issues usually beyond the scope of research (cf. chapters 9 and 10).

442. An example of such a treatment of one’s audience is given in footnote 472, below on page 163.

150

8.6. A scientific position of non-detachment

In the diagnostic process a lot more happens than can be formulated in terms of rules and executableprocedures. This happens more often in matters of expertise (cf. Dreyfus & Dreyfus, 1986). It is preciselythe hidden, tacit knowledge (cf. Polanyi, 1966) of the practitioner that offers a hinge between realist andnominalist discourses. The practitioner’s tacit knowledge is beyond the dichotomy between theory andpractice: it is not theoretical reflection, nor is it a straightforward application of general knowledge. Itdoesnot distinguish between what I called abstract and concrete aspects of objects, nor does it distinguishbetween past and future developments. It is only concerned with the moment of actual contact (cf. Shotter,1990) between subject and object. As such it does not take the critical distance that is so typical of therealist and nominalist discourses. In short, this tacit knowledge of the diagnostician comes very close towhat Gadamer meant by’phronesis’«443».

If some calculus is to be found or made, by means of which in the field of psychiatric diagnosis theoryand practice can be connected, then such calculus must take into account the interpersonal processes, theself-referential aspects of the exploration processes, the tacit and unarticulated tasks that take place at thecontact boundary between patient and diagnostician. In short, the measurement processes themselves, notonly their outcomes, will have to be included in the calculations and the inferences. Modern physics hasnot managed to solve such problem«444», and it seems wise to try first another trajectory, viz. to besatisfied with less than mathematical rigor and logical strictness, and to accept a connection between thepatient’s past background and his future treatment, that is less explicit and precise in its descriptions ofself-referential events in the practitioner. This is in fact what practitioners, in the fields of diagnosis andtherapy, have been doing: to abstain from calculus, and to account for the self-referential structures interms of ’subjective’ impressions and experiences«445».

The field of psychiatric diagnosis, therefore, has the opportunity to reappreciate the practitioner’s personalinvolvement as a tool for expressing relations between measurement and the object measured, that cannotbe put in exact terms.

8.7. In conclusion

The two critical discourses have turned out to be thus far incompatible in the field of psychiatricdiagnosis. An observer (diagnostician, therapist), when in a critical discourse, is concerned either withfinding explanations (etiology), or with devising applications (treatment policies) for a particular mentalproblem. Linking the two together seems to be possible only through the recognition, more or lessextensive, of a mode of performance which I called ’naive’, in which the observer’s own personalcommitments are not discarded.

I also maintained that the debates about the value of the DSM have been couched in terms of adescription-theory opposition, whereas the real problem is that, for the sake of consensus attainment,theoretical background notions, especially those about etiology, have been sacrificed. I suggested that abetter case for the inclusion of theoretical notions might have been made, if the prescriptive, procedureoriented, character of the DSM were criticized by means of some insights from phenomenology and self-

443. Gadamer (1960/1986) uses this term for ’prudent action’, a mode of tentatively proceeding, e.g. in matters of jurisdiction,where the actor does not blindly apply rules or techniques, but rather weighs the unique configurations in which he operates. Amaximum of context sensitivity is observed, with the aim of doing justice to the uniqueness of a situation (e.g. the situation of aparticular legal affair). Buthow exactly the actor is to operate, is a matter of logic-in-use, not specifiable in advance, and onlyreconstructible afterwards. In chapter 9 I will pay more attention to this logic-in-use (section 9.1) and to phronesis (section 9.4).

444. Cf. also Pattee, 1996. Spencer Brown (e.g. 1972, see also footnote 265 above) and Varela (1975) did make attempts atformalizing self-reference, but these did not substantiate into an empirical discipline.

445. It is of interest, to notice that modern points of view on complex biological systems do exactly the reverse: they attempt to dealwith the problem of subjectivity in terms of self-reference, be it called ’autopoietic organization’ (Maturana & Varela, 1980), or’semantic closure’ (Pattee, e.g. 1982), or ’closure under efficient causation’ (Rosen, 1991).

151

referential systems. For the diagnostic process is to a large extent of an explorative nature, containingmany moments where the exploration process investigates itself. These moments of self-reference maylead to what may be the ’essentials’ of the diagnostic conversation: the moments of personal encounter,where prespecified method fails. These moments, however, cannot be accounted for by means ofreconstructions in terms of a step-by-step procedure (they cannot be translated into non-self-referentialevents), and it is the recognition of this anomaly that will contribute to a scientific culture that is capableof making critical use of naive discourse. I will turn to this now.

9Against behavioral technologyThe previous chapter already mentioned the issue of self-referential exploration processes and theimpossibility to formulate an explicit step-by-step procedure that can replace such exploration processes.The present chapter is written in order to elaborate that idea.

Accordingly, I will argue that if the Church-Turing thesis«446»does not hold for self-referentialprocesses, and if playful exploration is also a self-referential process, then it is impossible to formalizeplayful exploration, such as performed by a therapist at his moments of naive perception, in terms of astep by step procedure. Put differently, some parts of psychotherapeutic work, both in diagnosis and intreatment, cannot be performed as ’applied science’«447»; they cannot be accounted for strictly interms of a path through some decision tree of patient’s behavioral criteria, and may nevertheless be valid.(The reverse, of course, does not hold: if a diagnosis or a treatment cannot be accounted for in terms ofobjective behavioral criteria, then this in itself does not warrant any validity!)

I will also present some of Gadamer’s ideas on play and ’phronesis’. These concepts are useful forunderstanding the kind of confusion of logical types that I consider basic and essential topsychotherapeutic exploration processes.

9.1. The context of discovery and the context of justification

Reichenbach (1938) distinguishes two contexts of speaking about the activities in which scientists mayengage. One he calls the ’context of discovery’ and the other the ’context of justification’:

"There is a great difference between the system of logical interconnections of thought and theactual way in which thinking processes are performed. (...) Epistemology does not regard theprocesses of thinking in their actual occurrence; this task is entirely left to psychology. Whatepistemology intends is to construct thinking processes in a way in which they ought to occur ifthey are to be ranged in a consistent system; or to construct justifiable sets of operations which canbe intercalated between the starting-point and the issue of thought-processes, replacing the realintermediate links. Epistemology thus considers a logical substitute rather than real processes. Forthis logical substitute the termrational reconstructionhas been introduced«448»"(Reichenbach, 1938, p. 5)

And hence:"The way, for instance, in which a mathematician publishes a new demonstration, or a physicist hislogical reasoning in the foundation of a new theory, would almost correspond to our concept ofrational reconstruction; and the well-known difference between the thinker’s way of finding thistheorem and his way of presenting it before a public may illustrate the difference in question. Ishall introduce the termscontext of discoveryandcontext of justificationto mark this distinction.Then we have to say that epistemology is only occupied in constructing the context ofjustification." (pp. 6-7).

The context of justification is introduced as one in which we are bound to present our arguments and acts

446. cf. section 6.2 and footnote 436 on page 147

447. The basic source of inspiration for this part of my argument has been Van Strien (1975), who proposes to recognize thepractitioner’s performance as a separate category of ’thinking in practice’, different from ’applied science’.

448. Reichenbach refers at this place in a footnote to Carnap’sDer logische Aufbau der Welt(1928).

152

153

as rigorously as possible; the context of discovery is introduced as one in which our factual thinking andacting takes place. These, according to Reichenbach, are entirely different domains.

"The act of discovery escapes logical analysis; there are no logical rules in terms of which a"discovery machine" could be constructed that would take over the creative function of the genius"(Reichenbach, 1951, p. 231)

In line with these ideas, the prescriptive methodologies of our mainstream scientific culture have focusedprimarily upon the construction of contexts of justification as the hallmark of scientific work, leaving thecontext of discovery aside as artistic, irrelevant or ’pre-scientific’. This is particularly apparent in the useof ’pilot studies’, usually meant as qualitative non-rigorous explorations of a particular field, preceding thepurportedly ’scientific’ stage in which generalization, prediction and verification are among the featuresthat make ’science’ preferable to ’art’«449».

Kaplan (1964), on the other hand, elaborating Reichenbach’s distinction between the two contexts, doesconsider logic applicable to the context of discovery, and introduces a distinction between ’logic-in-use’and ’reconstructed logic’. It is the former which he believes to be at work in discovery contexts. ThusKaplan attempts to give the discovering activities a certain rehabilitation. They too are considered to bedescribable in certain logical terms. But Kaplan also claims that this logic-in-use must be reconstructiblein terms of a context of justification«450». In this respect he shares the optimism of early AIresearchers such as Newell and Simon, whose work he mentions. Likewise, Hanson (e.g. 1965) was evenmore involved in elaborating such a ’logic of discovery’ as something which was no less scientific andvaluable as the established justificatory logic (the ’hypothetico-deductive analysis’).

The optimistic idea that there is a logic of discovery in human problem solving and in other practicalactivities in which people are involved, a logic that can be reconstructed and used as a tool forjustifications or legitimations of these activities, is based upon the assumption that an actor, irrespective ofthe nature of his activities, may step aside and describe his actions. Stated otherwise, it presumes thateach logic-in-use canin principle be reformulated as a procedure or recipe that could be performed byanother actor. We recognize here the ideas presented above in terms of classical science! For as soon aswe have a classical model at our disposal, we are capable of using it from a detached position, i.e., aposition from which we can maintain a clear view of our object of study, not hindered by our ownengagements with it.

This, in fact, is the basic question of any technology: can we reformulate some acts in such a way thatsome pre-specified outcome can be obtained repeatedly? Whenever this question can be answeredaffirmatively, control by means of a procedure (such as an expert system) is possible. Wouldn’t it makesense then to suppose that the acts of a psychotherapist too can be fruitfully reformulated as atechnological procedure? Yes, it would make sense, but the supposition would not necessarily be correctfor that reason. In fact I will assert its incorrectness in section 9.2.3.

A context of justification, therefore, is one of critical observers and critical investigators who are

449. E.g.: "It seems extremely important to make a clear distinction between psychotherapy artisanship and the science of behaviormodification. (...) Psychotherapy as an art, then, represents whatever is being done (the artisan’s behaviors towards his patients and hiscognitive attempts to organize and explain his domain of human behavior change) by practicing clinicians with patient in hospitals,clinics, private offices, patient homes, and the like, throughout the world. Its practitioners apply vaguely formulated principles, withgreater or less expertise, producing various degrees and kinds of results. (....) The important difference at the scientific end of this art-science continuum is the explicit focus on and intent at controlled observation and measurement, generalization, explicit formulation,prediction, and verification. It is a movement away from the private world of the clinician’s armchair to the public, replicable, andcontrolled world of science." (Kiesler, 1971, pp. 38-39)

450. "Let me follow the popular though loose usage and refer to all such faculties as "intuition". Then the point I am making is thatintuition also has its logic-in-use,and so must find a place in any adequate reconstructed logic. There is surely a basic differencebetween intuition and guesswork (...).

The difference seems to me to lie in this: What we call "intuition" is any logic-in-use which is (1) preconscious, and(2) outside the inference schema for which we have readily available reconstructions. We speak of intuition, in short, when neither wenor the discoverer knows quite how he arrives at his discoveries, while the frequency or pattern of their occurrence makes us reluctantto ascribe them merely to chance" (Kaplan, 1964, p. 14). (Italics added)

154

concerned with the construction of classical models! To the extent an investigator or actor is engaged inthe application of models for the sake of practical purposes, I will call his way of performing theprocedural mode.

9.1.1. Psychotherapy as applied science

Applied science, when concerned with applying the results of scientific investigations, is usually mainlyconcerned with steering and controlling. Controlling for external influences, maintaining invariant theexperimental setup, planning one’s acts, predicting and monitoring their effects are well knownconstituents of this type of applied science. These activities take place according to a strict scenario orprogramme, and can be justified as such. There is always a body of general knowledge available, whichserves as the model for the object studied. The qualification ’applied’ first of all pertains to the ways inwhich this model is used, especially to the ways in which the general model’s outcomes are interpreted(decoded, in terms of Rosen’s ’modeling relation’«451»). For instance, it would be a case of’applied science’ to use the general model’s calculations about the cannon’s position and angularorientations, in order to actually bring the cannon into the desired state.

Applications, then, make use of procedures which are intended tosolve problems, that is, to attain goals.Goals or desired final states can be formulated in precise terms in advance, and more or less clearevaluation criteria can be given for determining whether the goals have been attained. An application issuccessful if the results calculated in advance are actually obtained in practice. The kind of knowledgethat is used in this type of applied science makes use of general knowledge, it isnomothetic.To the extent that the applied scientist does not question his models in terms of how they correspond toreality, but is only interested in their usefulness for practical purposes, we may speak of atechnology.

For the field of psychotherapy we may classify behavior modification and cognitive therapy techniquesunder the heading of technology. These therapeutic styles depart from a corpus of procedures and rulesthat can be adapted and applied to individual problems of individual patients, once a proper diagnosis hasbeen made. They depend strongly on the availability of observable behavioral criteria (cf. section 8.3),which are considered the states of the problem to be dealt with (e.g. a particular phobia). The therapy isconsidered as a problem-solving trajectory of gradually altering problem states. Hypothesis formulationand testing may take place both during a treatment and between treatment conditions. In terms of an art-science polarity (cf. footnote 449, page 153) this would be considered ’science’.

The evaluation of the therapeutic outcome can be performed separately from the proper therapeuticactivities, in that the former does not take place as a constituent part of the latter. Like the diagnosis,evaluation consists of a measurement performed upon the patient, that is itself not part of the therapeuticprocess (e.g. measurement of social anxiety or of sexual impotence in a person).

However, more often than not the practices of these styles of work deviate from their paradigms, andsome unforeseen and not prescribed mutations and variations occur in the execution of the procedures, ina way that secretly makes the behavior therapist resemble his less procedurally oriented colleagues whosework is better to be described in terms of ’exploration’.

9.1.2. Psychotherapy as exploration

When we focus upon the discovering activities of science, the investigator is not so much involved withtesting and evaluating, as withproblem defining. Within a context of discovery no programme orprocedure is pre-given and no algorithms are available. Even to the extent that heuristics can beformulated, their implementation requires a lot of improvisation and instantaneous decisions, as in thepractice of behavior modification programmes.

451. cf. section 6.1.2.

155

Goals or desired final states can be formulated only in global terms, such as ’identify the problem’, ’havethe patient understand his situation’, ’allow the patient to accept himself more fully’, etc. These goals arenot so much the goals of a problem solving process, as the general emblems of a particular outcome thatone hopes to attain, without knowing in advance how this outcome will look or what it will be.

Being empirical in this context does not so much mean rigorously testing one’s hypotheses, as openingoneself to the newness of the phenomena, so that the unexpected is given maximal opportunity to appear.It is the process of getting rid of existing conceptual schemes, expectations, thoughts, hypotheses, etc. Infact, being empirical in this sense amounts to a loosening of the constraints«452»put upon theproblems at hand. Let us use the termexplorationto denote these activities.

The kind of knowledge that is used in this type of applied science is usually calledidiographic. Thepractitioner departs from, and looks for, descriptions of unique particulars. This uniqueness is usuallyunderstood as the ways in which these particulars are embedded in their contexts. A context-freedescription of them is considered pointless if not impossible. A generally formulated procedure could beapplied only by ignoring their uniqueness. The activities within this explorative mode cannot beformulated as particular instances of general rules.

For the field of psychotherapy we may classify therapies such as psychoanalysis, client-centered therapy,etc. as taking place in a context of discovery, different from technology. These styles are described interms of activities that arenot so much intended to solve problems, as to define them. Arriving at aproblem identification is not so much a matter of moving through a search tree. No algorithms areavailable, and a special status is given to the attainment of insight in the patient (as well as in thetherapist). What it is that will become clear and self-evident by the time the therapy is terminated, is notyet known in advance. In terms of the art-science polarity this would be considered ’art’.

In respect of client-centered therapy, Zimring (1974) writes for example:The therapist does not assume the client’s goal. If the client’s presenting problem is an inability tostudy, the therapist does not assume that the therapy is successful when this is remedied" (p. 121)

To the extent that the therapist does have an idea about the desired goals of the therapy, he isaccomplishing a problem-solving rather than a problem-defining task. To that extent also is he bound torigorously test his hypotheses and expectations, and to evaluate the obtained outcomes, and to describethese activities in a context of justification.

During the explorative activities the evaluation of intermediate therapeutic effects cannot be performedindependently of the therapeutic process itself, contrary to the behavioral technologies. Instead, thetherapy itself is here the only paved way to evaluate its own outcome. No validindependentmeasurementprocedure is possible, for no goals have been established that could have been used as the criteria formeasurement. Any attempt to develop a measurement independently of the therapeutic process itselfcrashes upon validity problems of the variety "is this (measured) change of patient functioning a goodindicator of his progress?". The desired effects of the therapy cannot be defined in advance; clarity ofproblem definition is an outcome rather than an initial state of the therapeutic process. Only at the end ofsuch an explorative therapeutic process do we know how this therapy has been fruitful and why. No validcriteria independent of this process are available. The therapy develops its own measurement procedure, asit unfolds.

452. of the variety: ’any solution x should satisfy the conditions a through c’.Cf.:

"Fundamental to Erickson’s approach is his view that patients have problems because of their learned limitations. The object oftrance is to relax these learned limitations of the patient’ usual frames of reference to permit the vast reservoir of their unrecognizedpotentialities to operate. Freed from the common sets, biases, and inhibitions of consciousness, learning can proceed on an autonomousor what is conventionally called an unconscious level." (Erickson, Rossi & Rossi, 1976, p. 298).

Cf. also Watzlawick (1984, p. 328) about an escape from language: "... the experience of breaking through (...) whoseimmediacy defies description". This is a phenomenon extensively studied also by Zen and Yoga masters.

156

A different way to put this is that the explorative process in a psychotherapy has the character of creatingnew elements, such as concepts, experiences, decisions, which could not be foreseen in advance. Hence,the therapeutic exploration process isself-modifyingin the sense of Kampis (1991): "no formal system canbe strong enough to transform the information content of the component-producing processes, typical oflife and mental activity, into a set of algorithmic rules that describe the process by means of state-transition rules" (p. 236). And it is due to this fundamental impredictability that we cannot formulate inadvance which behavioral criteria will be useful for measuring the outcome of therapies! Or, putdifferently, those criteria might be formulated, and they are formulated in fact, but by the time they areused for measuring how patients have improved, they do not measure to which extent their validity hasexpired! Hence, a truly sensible research method should investigate how useful and valid the scheduledmeasurement criteria are. This, in fact, would amount to a negative approach.

9.1.3. An asymmetry between explorative and procedural modes

In a procedural mode there is always a strict distinction between the body of knowledge and itsapplication. Or, in terms of section 6.1.2, the decoding function is disjoint from the model, and arepresentation of it can be added to the model as a separate unit. On the other hand, in an explorativemode, such a clear distinction is lacking. The decoding function cannot be separated from themodel«453».

In the practice of psychotherapeutic activities it is often difficult to distinguish between the generalknowledge about the patient, and the applied knowledge, the particular information about this presentpatient, that makes the psychotherapist decide to make a different intervention, or to intervene later, orwith less emphasis. Furthermore, the ’diagnostic phase’, in which the practitioner might wish to determinehow the general knowledge is to be decoded for this particular patient, mixes up inextricably with the’treatment phase’. To recognize in the patient a particular contextual pattern and to treat the individualwithin this pattern go hand in hand«454». More generally, we find in these explorative processes acoinciding of a process (e.g. the therapeutic process) and that which is produced by this process (e.g.insight in one’s personal dynamics; changes of one’s experiential capacities).

To explore means to possess and to exercise a sensitivity to novelty, which is a capacity to suspendmeasurement and classification. Novelty can appear only to the extent that conceptual schemes can be putaside. Only then can new features of the phenomena studied be found and can the object of investigationappear as unique. This sense of uniqueness disappears as soon as general classificatory statements aregenerated or taken recourse to.

The practical involvement ensued by the explorative activities is not the application of some implicit (or’unconscious’) general knowledge based rule! For instance, there is no such a thing as a formal model orconceptual scheme of the therapist’s explorative expertise«455». However, one might desire todetect such a scheme by means of interviews and observations). In projects aimed to formalize medicalexpert judgments in terms of problem solving heuristics protocols, for example, the medical expert isconsidered to ’have’ the constructed protocols already somewhere ’inside’ his abilities, and hisperformances are considered as executions of them, and therefore as indications of the internally storedprocedural knowledge.

453. Cf. Toulmin (1992) who, within a different line of reasoning, maintains that in the last decades pure and applied science havebecome more and more indistinguishable and inextricable.

454. This has implications for the kind of ’insight’ that can be acquired during a therapy: not so much insight of a taxonomic kind,as one that is mixed up with the therapeutic process itself.

455. To the extent that the expert’s knowledge can be formalized and treated as if already existent as a procedure ’in’ his mind, it isnot an explorative kind of knowledge that is found. Even his search strategies, to the extent that they can be formalized, consist ofcategories and schemata. However, his explorative capacities reside in his abilities to suspend these categories and schemata, wheneverhe ’feels’ he should do so. Implicit criteria for the occurrence of this suspense might be found in the expert’s performance or histhought, but this would only displace our problem: when and how does the expert decidenot to follow the rules?

157

In spite of the enormous rise of artificial intelligence, this latter point of view is misleading«456».To construct a system of general rulesas-if the actor had been executing these in his practicalexplorations, is one thing. This is inherent to model building. But to believe that the rules as used by themodel are those actually used by the actor, is quite a different thing. Indeed, such approaches attempt toreconstruct the actor’s explorative actions in a context of justification, by disregarding their explorativecharacter, as if the actor already had a knowledge about what to do, and then did so accordingly. But infact he did not know in advance what to do, and, more importantly, he did something of which the logic-in-use, contrary to Kaplan’s claims, cannot be reconstructed (for reasons to be discussed in section 9.2.2).

In explorative enterprises the actor is not armed by any predefined structures such as laws or availabletechniques, or at least, to the extent the actor operates within a context of discovery, he abstains fromthese equipments. To reconstruct his actions as if he did use them, as if he were a ’kind of’ technologistexecuting a procedure, is to misunderstand his actions as actions that areonly possible as long as they gowithout a pre-given structure. This holds for both applied and pure inquiries. Let us explore the idea ofexploration in more precise terms.

9.2. Exploration as a self-referential activity

9.2.1. Uniqueness

A major difference between the procedural and explorative modes resides in the conception of theuniqueness of the objects dealt with by the actor. In what I have called a procedural mode, the kind ofuniqueness that is taken into consideration has been presented very eloquently by Eysenck:

"It is quite undeniably true that Professor Windelband«457» is absolutely unique. So is myold shoe."

and:"To the scientist, the unique individual is simply the point of intersection of a number ofquantitative variables" (Eysenck, 1952, p. 18)

According to Eysenck’s definition, uniqueness of an object is a property of the object; it is nothingspecial, and certainly not a reason fornot subsuming the unique individual under general laws orsubjecting it to treatments derived from them. Applications in the procedural mode are concerned withphenomena that are ’unique’ in this sense. What kind of uniqueness may then be meant if, due to it,explorative activities are fundamentally different from those rule-guided activities that fit in so well withthe requirements of applied classical science?

It is that the uniqueness of the observed objects is emphasized and unfolded by their being explored. Theprocess of exploration unveils them as embedded in a pattern in which their uniqueness is becomingapparent. Thus, the contextual relations that are discovered are brought forth by the exploration processes.

This, however, does not yet make the objects observed unique. What does make them unique is not aparticular property of them (which Eysenck would consider a mere intersection of variables), nor theirbeing brought forth by the observer (as are all observed objects), but rather their specific context in whichthey are embedded. For it is that the observer’s explorative activities are themselves also part of thepatterns that they uncover! When we thus understand the uniqueness of an object as situated in the net ofcontextual relations with its environment, in which the observer’s own explorative activities themselvesalso partake, then this uniqueness is not so much due to some quality of the object itself, as to the factthat the observer, by becoming explorative, has engaged in aself-referentialnetwork of activities. Theuniqueness of a particular object, then, consists of the exploration of the observer’s own interactionalopportunities. It consists of the observer’s capacity to experience himself in contact with the object of his

456. The reader is reminded of my quotation from Winograd (1981), see my footnote 184 on page 74.

457. who was the first proponent of the nomothesis-idiography polarity

158

explorations. We may call this configuration of self-related explorationplay«458». Play, as aparadigm of exploration, consists of playing with one’s very opportunities to play! Play is self-referentialin this respect. We recognize here, in this notion of uniqueness, that which I described in section 7.2 asthe observer’s attention to the interface between himself and his object of study.

Accordingly, the psychotherapist is monitoring his own responses to the patient, he is steadily wonderinghow a particular patient is affecting him and how this patient manages to do so. If a particular patient istriggering a sense of sadness in the therapist, the therapist is toplay with this experience, in that he has toadmit the feeling as his own feeling, taste it, see how it can be masked by other experiences, look for itagain, as if his new experience of sadness were an infant with which one plays peek-a-boo! It is throughsuch a playful handling that the patient obtains a particular unique meaning to the therapist.

Clearly the uniqueness of an object, understood in terms of the exploration of the interactionalpossibilities in which the observer engages, isnot expressible as an intersection point of variables. Ratherit has as much to do with these variables as with the observer’s acts of value-assignments to thesevariables. Indeed, the observer’s very act of value-assignment is indistinguishable from the value that isassigned! This is ’quite undeniably’ beyond professor Eysenck’s frame of reference, but for us it cannotbe surprisingly new; it the concept of naive objects, as already introduced in chapter 3.

The uniqueness of a particular object resides in the ways in which the actor, as a player, finds himselfable to play with this object, i.e., to explore it. The actor/player/explorer, for that reason, is actuallyconcerned with his own activities, meanwhile enjoying the uniqueness of the object he is exploring. Theobserver who is performing explorative activities in the field of his interests, finds himself involved inactive participation in this field, due to which a detached observation is impeded. His involvement entailsinteractions with and manipulations of his topic of interest. They constitute the uniqueness of theobserver’s own exploration process. That which makes an object of exploration unique, therefore, is theuniqueness of the exploration process itself that the actor performs! In the exploration process nodistinction exists betweenthat which is explored andthe way in whichit is explored. Exploration is self-referential. And it is through such a self-referential exploration process, in which the therapist explores hisown experiences while interacting with the patient, that the therapist may come to perceive the patient interms of these experiences, or better: through these experiences. In terms of second order distinctions: thetherapist’s (sad) experiences are the tokens (first order distinctions) through which the patient arises to thetherapist as a naive object.

9.2.2. Self-reference and the archimedean point

Given a particular procedure to produce a particular product or to attain a particular goal, that whichescapes from the description of this procedure is the factual application (interpretation, execution) of thisprocedure. The process of execution goes beyond the explicit awareness of the actor. It is usually ’tacitknowledge’ (Polanyi, 1958). The description of the application process does not make part of theprocedure that it applies. This is inherent to what Löfgren has described as the ’linguisticcomplementarity’ (cf. section 6.2.3).

A description Dp of a procedure P doesnot contain a description Di of an application process APi (whereby running APi, Dp is applied so that, as a result, P is performed). In its turn, Di can be applied only by anew application process APi+1 that does not equal APi (where by running APi+1, Di is applied, so that APiis performed). Therefore, any description Dj of APj will require an as yet not described application processAPj+1 in order that APj is performed.

If a finite description can be made of the application process, as in technology, then this descriptioncannot be part of the description of the procedure that is applied, but instead, it can be appended to it (cf.section 6.1.2). On the other hand, if a finite description of the application process is not possible, then it

458. I already presented some ideas on play in section 3.2, and I will return briefly to this topic in 9.4.

159

is impossible to distinguish strictly between a procedure and its application, and the performance is self-referential in some aspect! However, this does not mean that an adequate self-referential description ofsuch performance would be possible. This is the regular case in matters of craftsmanship. The mastershows the pupil the crucial skills and properties of the material, and conveys their ’feel’, without any needto give a fully explicit description of application processes. That is, the master orients the pupil to theproduct as to a naive object. An adequate explicit description of this ’feel’, however, would require aninfinitude of additional information about procedures, techniques, and criteria for using or not using them.

A possibly smarter way to encode the ’feel’ of the master’s performance would be to make a description,such that properties of the subject matter described could be sensed in the way in which the description isconstructed. Indeed, if one wishes to describe the impossible, then write a poem«459»! A poemmight convey to an audience the ’feel’ of the master’s work, without, of course, digressing intotechnological smalltalk. I will return to this idea in section 9.4 and in chapter 10 on parallel processes.

For the present moment, let me state that the ’design’ for a work of craftsmanship is permeable toproperties of its application; the design can be ’redesigned’ due to unforeseen circumstances during theprocess of construction. If there is an interpenetration of a procedure and its application process, thenthere is no way to make a blueprint (or program) for the craftman’s performance.

To the extent that an activity is explorative it is concerned with itself in a way that cannot be theexecution of a procedure. Many aspects of exploration could be caught in terms of rules that are followedby the actor, and each time that such a rule is formalized, the acts it covers can be performed as part of anewly constructed procedure. But there is always something in the explorative activities that escapes frombeing formalized. Here Rosen’s ideas on the position of semantics applies«460».

Likewise, there will always be an opportunity to ’pull out’ some part of the craftman’s performance andstretch it into a rule, but the explorative qualities will appear too thin to be caught.

Exploration requires to engage with the topic of exploration. As a consequence, there is no longer anexternal point of view available from which the actor can watch his topic of attention. Observation hasceased as a critical and detached activity.

Technology is possible only after the actor has entered a nominalist discourse, where symbols serve asstable observable objects. But the only thing that cannot be brought into a critical discourse, so that itcould be controlled from an archimedean«461»point of view, is the way in which the actor relatesto this archimedean point. Precisely when such control is attempted, this point loses its quality of surveypoint, and it is the attitude to one’s archimedean point, to one’s sources of certainty, that becomes of focalinterest when one attempts to apply a certain technique to oneself.

459. One is reminded of Jim Croce’s (1973) eternal strophes:

Amaj7 C#m Bm7 E7Well, I know it’s kind of late, I hope I didn’t wake you

Amaj7 C#m Bm7 E7But what I got to say can’t wait, I know you’d understand

D D#dim C#7 F#m / DEv’ry time I tried to tell you the words just came out wrong

Amaj7 E D Amaj7 / ESo I’ll have to say I love you in a song

etc.

From: the song "I’ll have to say I love you in a song", album "I Got A Name", ABC Records ABCX-797 (December 1973) Lyrics(these lines have been gratefully combined from: http://imsnispc01.netvigator.com/~wlleung/ihavetos.txt and http://www.jim-croce.com/lyrics/igan.html#side2)

460. quoted above in footnote 264, page 96

461. Cf. Descartes’ second meditation (1641, AT VI, p. [24]/1973, p. 149). See also section 5.2.1, esp. footnote 195, on page 77.

160

A physician would not be able to detach sufficiently from himself and obtain a distance required for beingprofessional. Kafka (1969, p. 211) puts it thus:

"Physicians cannot adequately treat members of their own families because they are not ’objective’enough; because, let me rephrase this, they cannot treat members of their own families sufficientlyasnon-living objects."

More immediately, a surgeon or a dentist would not be able to be a doctor to himself«462». It istrue for all situations in which we attempt to deal with our own body as if it were an external object fromwhich we are able to detach, and as long as we succeed in detaching from ourselves and keep anarchimedean point of view. Metaphorically speaking (and literally true as well): one could nowise useone’s own index finger to point at itself.

But what about the psychotherapist who treats himself therapeutically, like Freud who performed his self-analysis? Precisely because it is possible to discover in oneself motives and meanings thus far unthoughtof, there isno detachment from oneself, but instead an access to oneself. This is only possible to theextent that one doesnot apply a procedure; to the contrary, it is possible only to the extent one stopsdoing so. The things discovered during a process of self-exploration are recognized as one’s own, not asbelonging to some third person, whom we observe in a detached way. In terms of the metaphor: the indexfinger, while attempting to point at itself, has stopped to point at all«463». This holds for eachattempt at self-exploration. Not so much a transparency to oneself is required, comparable to Descartes’idea of the Cogito’s self-transparency, as the very absence of an archimedean point! Freud’s self-analysiswas possible only ifnot understood as a method, a strict procedure, performed from an external controlposition. As the performance of a technique it would have been impossible. As an exploration of one’sown experiential field, on the other hand, it is very well possible and not at all reserved to greatminds«464». Indeed, it is a basic capacity of living beings to sense oneself. But it is only within acritical discourse that we are capable of putting this capacity into words, and describe it as self-referential.

The relation to one’s archimedean point also becomes of interest, in matters of personal encounter, wherethe events at the interface between the two persons invite the practitioner to leave his critical discourse,his detached scientific position, and instead to give up the firm grounds he thus far believed to possess inthe certainties of his scientific method. Buber puts this thus:

"Until, in certain cases, a therapist is terrified by what he is doing because he begins to suspectthat, at least in such cases, but finally, perhaps, in all, something entirely other is demanded of him.Something incompatible with the economics of his profession, dangerously threatening, indeed, tohe regulated practice of it. What is demanded of him is that he draw the particular case out of thecorrect methodological objectification and himself step forth out of the role of professionalsuperiority, achieved and guaranteed by long training and practice, into the elementary situationbetween one who calls and one who is called. The abyss does not call to his confidentlyfunctioning security of action, but to the abyss, that is to the self of the doctor, that selfhood that ishidden under the structures erected through training and practice, that is itself encompassed bychaos, itself familiar with demons, but is graced with the humble power of wrestling andovercoming, and is ready to wrestle and overcome thus ever anew. Through his hearing of this callthere erupts in the most exposed of the intellectual professions the crisis of its paradox. Thepsychotherapist, just when and because he is a doctor, will return from the crisis to his habitualmethod, but as a changed person in a changed situation. He returns to it as one to whom thenecessity of genuine personal meetings in the abyss of human existence between the one in need ofhelp and the helper has been revealed. He returns to a modified methodic in which, on the basis ofthe experiences gained in such meetings, the unexpected, which contradicts the prevailing theories

462. Cf. also Menninger (1963, pp. 366/7): "Most doctors abhor taking medicines".

463. This type of transition has been presented in section 7.3.3 in terms of a ’one-sided boundary’.

464. The famous paradox of the barber who shaves all men in the village who do not shave themselves then reduces to: if he shaveshimself detachedly (professionally), he doesn’t shave himself undetachedly (unprofessionally), and vice versa.

161

and demands his ever-renewed personal involvement, also finds its place." (Buber, 1957, pp. 94-5)«465»

Technology cannot handle self-reference. Self-reference deprives the technologist from his criticaldiscourse and from all footholds he believed to possess. Hence, if it makes sense that psychotherapycontains exploration as a necessary ingredient, which is a self-referential activity«466», then wehave found a ground for arguing that a classical model of the therapist’s actions would necessarily beincomplete and inadequate. Conversely, if a psychotherapeutic procedure could be constructed withoutself-referential parts, then such a procedure would be dubitable as to its opportunities for exploration bythe therapist«467».

9.2.3. Improvisation and rules

Explorative activities are necessarily not predictable, nor planned or under the control of a strict scenario,because they are concerned with precisely that which cannot be brought under such a regime: the positionof the person who does the controlling. For that reason it does not make sense to regard them as’subjective’: only if some ’objective’ procedure would be available to replace it, would the objective-subjective distinction make sense. Now, faced with the impossibility to offer an ’objective’ alternative tothe kind of self-referential attention that is the case in exploration, it does not make sense to call the latter’subjective’.

An actor who is involved in exploration is not so much in need of a calculus as of an ability to improviseand use his ’intuition’. Though this has been stated often in regard of psychotherapeutic activities, it isusually considered as a testimonium paupertatis by those who would like to reformulate these activities instrict procedural terms. But in fact, improvisation is a proper way to act whenever the proceduresavailable and their subject matter abound with unforeseeable properties (cf. Kornwachs, 1986; cf. alsoWaddington, 1977, p. 145). This holds also for the field of psychotherapy, where both the patient’sexperiential field and the therapeutic process itself are not stable.

Indeed, improvisation is the only possibility in view of the self-reference of the explorative process: thetherapeutic process, exploring itself, has to adapt continuously to its own developments. Though atherapist needs to know when and how he should perform a particular technique or rule, he should alsoknow how to adapt it, when to drop it and when to disregard it. This latter kind of knowledge, however,cannot be part of the applied rule itself, or laid down in a meta-rule, for at each moment of the therapysuch a rule or meta-rule would itself have to be adapted, dropped, reinvented or discarded (dependent onthe contingent features of the ongoing therapeutic process). A rule cannot indicate its own exceptions nor

465. "Bis [ein Psychotherapeut] in einem bestimmten Fall, in bestimmten Fällen über das, was er tut, erschrickt, weil ihn dieAhnung überkommt, daß, zumindest in solchen Fällen, aber letzlich vielleicht in allen, etwas ganz anderes von ihm gefordert ist, etwasder geläufigen Berufsökonomik Unangemessenes, ja die geregelte Berufsausübung zu gefährden Drohendes. Nämlich daß er zunächstden Fall aus der methodengerechten Versachlichung ziehe und selber, aus der in langer Lehre und Übung errungenen und durch sieverbürgten professionellen Überlegenheit tretend, in die elementare Situation zwischen einem anrufenden und einem angerufenenMenschen eingehe. In Wahrheit ruft der Abgrund nicht die zuverlässig funktionierende Aktionssicherheit, sondern den Abgrund an, dasheißt, die unter den durch Lehre und Übung errichteten Strukturen verborgene, die selber vom Chaos umwitterte, selber mit denDämonien vertraute, aber mit der demütigen Macht des Ringens und Überwindens begnadete und immer neu so zu ringen und zuüberwinden bereite Selbheit des Arztes. Aus dem Vernehmen dieses Anrufs bricht in dem exponiertesten der geistigen Berufe die Krisisseiner Paradoxie aus. Der Psychotherapeut wird, eben wenn und weil er Arzt ist, aus der Krisis in die Methodik zurückkehren in einemodifizierte Methodik, in der, von dem in solchen Begegnungen Erfahrenen aus, auch das Ungewohnte, das den herrschendenDenkungsweisen Widerstrebende und den stets erneuten personhaften Einsatz Heischende seinen Platz findet." (Buber, 1952, pp. 10-11)

466. Cf.: "Ultimately, psychoanalysis is a self-referential field: all we have is the data of our own experience and our experience ofthe patient’s experience." (Levenson, 1988, p. 143)

467. Instead, exploration recedes here to the level of design of the effective procedures! Such an ’experimental-clinical approach’ "ischaracterized by its emphasis on methodology instead of by a specific theoretical orientation. (...) Behavior therapy would best proceedunencumbered by parochial adherence to one specific theory. In the author’s view, behavioral theories and therapies should be based onexperimental evidence with real clinical patients rather than on orthodoxy or faith." (Emmelkamp, 1986, p. 385-6)

162

the contingencies of its implementations, but a good practitioner can find these, comparably to how acraftsman can get a ’feel’ (cf. section 9.2.2) for his product.

The absence of an archimedean point requires types of mathematics other than calculi. Explorativeactivities are not so much concerned with producing something already known in advance, as with therecognition of features that have not been planned or programmed, and that establish their own rules asthey emerge. This happens for example in the kind of systems described by Joachim Mowitz and myself(1989) in which geometrical forms are created by growth processes that define their own rules of growthduring the process«468».

Furthermore, if the reason why psychotherapy cannot become a technology is due to the central role ofexploration in it, which is self-referential, then the kind of activities that take place in psychotherapy canbe put in procedural termsonly to the extent that they miss the (self-referential) pointaround which thepsychotherapeutic endeavor is pivoting.Gustafson (1986) writes:

"Doctors like standard technique. They like experiments less. But Alexander and French propose tomake every case an experiment! Freud had this attitude in the early and middle phases of hiscareer. Ferenczi was always up to his grand experiments; Reich was also. Indeed, perhaps all of thewriters described in this book take this experimental attitude about psychotherapy. Certainly, weshall see it in the following chapters about Sullivan, Winnicott, Balint, Gedo, Havens, and theMilan team. The modern, well-known brief psychotherapists, Mann, Malan, Sifneos, and Davanloo,had some of this experimental attitude, but all were ambitious as well to define a standardtechnique of their own.All became defenders of their own dogma.The experimental attitude makeslife as a psychotherapist much more interesting." (p. 44, italics added)

Clearly, this ’experimental attitude’, as Gustafson calls it, resembles that what I discussed in terms of’play’, and any proceduralization of it (the construction of a ’standard technique’) would destroyit«469».

The same idea put in other words:"Do not let focusing, or reflecting, or anything else get in between. Do not use it as an in-between.Do not say: "I can stay here because I have my reflecting-method, I have my ping-pong-paddle, soyou cannot get me. You say something? You get it back." There is a sense that we are armed, yousee. We have methods; we know focusing; we have credentials; we have doctors. We have all thisstuff and so it is easy for us to sit there with stuff in between. Do not let it be in between; put itout of the way. You can have at least as much courage as the client has. If not, I would beashamed of myself, with all the stuff that I have, if I still cannot really look when this person can.So I want to be there in that same way.

That - I think - is the first job we have. And on the question what we client-centered peopleneed to do now, on this, also, I think the first thing we need to do is to communicate that attitude.That is so necessary in a field that is becoming more and more "professional," which is to sayuseless and expensive." (Gendlin, 1990, p. 206)

The often found fact that therapists actually perform differently from what they claim about theirperformance, now can be understood from a new perspective: if they were to describe correctly what theywere doing as therapists, it would not be therapy at all! If they do therapy, then of necessity they cannotdescribe their performances correctly! Their descriptions are necessarily incomplete. But thisincompleteness is not due to practical circumstances, and for that reason contingent; to the contrary, it isdue to a formal property of the activities, viz. their self-reference. For that reason, we do not only need adescription of what a therapist actually does, but also one that describes of itself how and why it isincomplete! But that is a truly hard problem: the therapist "does not know what exactly he ’knows’ in

468. in contrast to cellular automata, that are run on pre-defined grids in euclidean spaces and on pre-defined time intervals

469. Whitaker (1976, p. 154) even calls the role of theory ’constipating’. Cf. also Haley (1978).

163

knowing that"«470». We find here the necessity of using a negative way: the therapist, inaccounting for his actions in a session, should explore and put into words his problems to convey the gistof the session. "You (researcher) will not understand what I am talking about" can be made a usefulsource of information, in this respect (cf. section 7.4.2).

When London (1964, pp. 156ff) considers psychotherapy necessarily as a matter of manipulation andcontrol of a patient by a therapist«471», he ignores the improvising quality of the therapist’sperformance, as well as the self-relatedness of this performance. London maintains that psychotherapistsare manipulative by definition, and that it makes less sense to debate the existence of these manipulations,than to take care of their proper, ethical usage. But at this point the contra-control advocates, of whichRogers’ thinking (cf. Rogers & Skinner, 1956) has been a paradigm, seem to have missed a vitalargument. ’Facilitating’ (non-controlling) actions by a therapist may or may not have therapeutic effectsupon the patient, but this does not turn ’facilitation’ into a procedure or an otherwise controllablevariable! It would be a procedure only if it could be executed as such, quod non. Being facilitating parexcellence requires from the therapist his renouncement of the archimedean point, the point from whichcontrol is possible. The uniqueness of an individual person (i.c. the psychotherapy patient) only makessense as a ground against therapeutic manipulation, if this uniqueness is understood in terms of a playful,explorative perception by the observer (i.c. the psychotherapist), not in terms of an intersection ofvariables (cf. Eysenck on uniqueness, see section 9.2.1). The ’facilitating’ therapist can be facilitating onlyto the extent that he is able tofeel how a particular rule or technique should be adapted to the particular(’unique’) circumstances of a particular therapeutic situation. This feeling of the implementation featuresis a feeling by the therapist of himself, being the instrument. The assumption that a particular action (i.c.:’facilitation’) can in principle always be ’refined’ into a procedure, is in fact a case of the Church-Turingthesis (cf. section 6.2 and footnote 436 on page 147). Its value finds a logical limit in the self-referentiality of explorative actions.

9.3. The problem of psychotherapeutic procedures

A particular consequence of the self-referential relation between the topic of attention and the way inwhich the application process takes place, isthat the topic is not clear for those who are involved!Thereis a necessary vagueness around this self-referential process. This vagueness is due to the impossibility togive an exact description of a self-referential process.

Due to this self-reference, the way in which the therapeutic process takes place is intimately linked withthe topics of attention in the therapeutic conversation. The ’things’ that have been discussed within atherapeutic setting are to a large degree not translatable into other, regular, conversations. They cannot bereported easily to an external person. That which is usually lost when reporting on the therapeuticconversation, is the latter’s self-reference: in the reporting situation there is generally no link between thatwhich is reported and the way in which it is reported, whereas in the therapeutic situation such a link doesexist«472».

470. To use for a second time these lines from Gadamer, TM p. 92. Levenson (1988, p. 135) also discusses the phenomenon ofknowing what is taking place in a therapy, but not being able to put it into words.

471. See Dies (1985) for an elegant presentation of some pro-control and contra-control advocates in the field of psychotherapy.

472. In a talk given in a conference at Bergamo in 1990, where I gave my paper (Goudsmit, 1992b), I have illustrated this point bydemonstrating to the audience that a self-referential message loses its meaning, so that, as a result, the topic of shared attention isdissolved. In order to demonstrate this, I made use of the fact that the (English spoken) lecture was translated simultaneously intoFrench and Italian. With the kind permission of the translators, whom I informed in advance, I have pronounced sentences like "Forexample, ladies and gentlemen, in order to illustrate what happens during self-reference, you are made advertent of the fact that at thismoment most of you are listening to the translation of my English spoken text, because the translator, who is sitting there in the rearend of our room, is speaking to you via the headphones that you are wearing. And right now, as you are listening to the voice of thetranslator in your headphones, you may hear the translator tell you: ’it is me, the translator, who is talking to you now, me, thetranslator, sitting behind you in this room; the speaker is talking about me!’". The translators, who in fact were just translating mywords, were thus put into a position in which they could no longer be a transparent medium. In this way I stepped out of the languagegame that was current at the conference. As a result, the audience got confused and was no longer able to discern between form

164

To give an example, in a therapy of weekly sessions it becomes apparent that the patient has problems toremember the gist of the sessions throughout the whole week. After some four days he has encountered asufficiently large amount of persons who do not treat him as respectfully as the therapist, that he loses hissense of self-respect that the therapist would have liked to ’grow’ in the patient. When this comes to bementioned by the patient ("somehow by wednesday I can no longer remember what you have said, andhow you said it"), the therapist decides to call this forgetfulness the effect of an imaginary ’wednesdaytherapist’«473». It is this wednesday therapist that personifies the patient’s own modes ofdisrespecting himself. During the sessions that follow, a confrontation between the (real) therapist and thewednesday therapist is arranged, and the various forces that the patient feels to be affecting him areexplored. Now the point is: what should this patient tell his wife when he comes home from the session;what is it he has been talking about with the therapist? The difficulty is not that he would not be able toexplain to her who the wednesday therapist is (though she might find it a strange thing to talk about);rather, the difficulty is that the real therapist, unlike his wednesday colleague, is both the one who iscompared to the wednesday therapist, and the one who does the exploration! And it is precisely thisdoubleness that makes the situation self-referential: for it is the real therapist who enables the patient tocompare him with the imaginary wednesday therapist; the latter would not allow such a comparison.Beyond this therapeutic setting any comparison between the two therapists would be a prosaic comparisonof features; it would lack this self-referential property, and with it the crux of the session.

Most research endeavors in psychotherapy are focused upon the construction of treatment protocols andprocedures that can be described clearly, and that accordingly can be transferred between therapists, i.e.,without too much interference of variables that are related to personal, therapist related ways of applyingthe procedures. Thus, for phobic disorders a great many therapist independent treatment procedures havebeen developed, which are extraordinarily successful. However, for those therapeutic interactions where,due to self-reference, a precise procedure cannot be specified, this clarity has been lacking.

Thus, Egan (1975), in a widely used and often reprinted textbook, aims at contributing to an increase ofthe reliability of helping processes (p. 2). Accordingly, he presents a system of ’helping skills’, each ofwhich can be learned and practiced. Exploration, and especially self-exploration are mentioned asindispensable helper skills, but their self-referentiality is not hinted at in any way. Thus, the specificationof these explorative skills remain troubled with a quality most incisively expressed in Lewis Carroll’slines: ""Where shall I begin, please your Majesty?" he asked. "Begin at the beginning," the King said,gravely, "and go on till you come to the end: then stop."" (Alice’s Adventures in Wonderland)

Most psychotherapists, though fully confident of the therapeutic usefulness of their own performances asthey unfold during therapeutic sessions, are not able to ward off critical comments by scientists and fellowtherapists, that their modes of performance and their decision criteria cannot sufficiently be accounted forin terms of clearly describable procedures (cf. footnote 449, page 153). They often feel bound to admitthat their professional activities can be reformulated, at least in principle, in procedural terms. Thoughtherapeutic practice is heavily dependent on the attainment of immediate contact with thepatients«474», many psychotherapists do not like to admit that therapy sessions contain momentsbeyond their control«475».

There is a vast amount of authors in the field of psychoanalysis«476»and client-centered

(translator) and content (speaker) of the lecture, until I took up again the regular format of the lecture. Precisely this, I maintained,happens in psychotherapeutic encounters, in which features of the therapeutic interaction (form) are articulated (as content) into theconversation, in a way that cannot be reconstructed in a ’plain’ report, without making it self-referefential in turn.

473. cf.: "Most analyses have surrogate analysts lurking somewhere on the fringe, providing a symmetry to the process. If we arethe mother, there may be out there a father" (Levenson, 1994, p. 16)

474. cf. footnote 354 on page 126

475. cf.: "... psychoanalysts abhor unclarity." (Levenson, 1994, p. 21)

476. e.g. Peterfreund, 1983

165

therapy«477», that attempt to describe the therapeutic enterprise in terms of procedures, tactics andstrategies. My present point is not that these descriptions are without value, nor that their content is notinteresting, but rather,that the gist of their ’procedures’ is that they cannot be appliedrigorously!«478»And that means that, viewed from a critical position, they should be understood asself-referential acts, the descriptions and justifications of which are necessarily incomplete or distorted. Inorder to account for their performances, practicing therapists often seek recourse to a cartesian theory of"reasoning by analogy", as an instrument of justifying a phenomenon that I would prefer to describe interms of the ’third position’ ortertium, beyond the object-subject dichotomy!

Reasoning by analogy means that"... pour que la projection [des données de mon expérience intime en autrui] soit possible, pourqu’elle ait lieu, il faudrait que je me fonde sur l’analogie entre les expression de physionomiequ’autrui m’offre et les différents gestes physionomiques que j’exécute moi-même." (Merleau-Ponty, 1951, p. 21)«479»

This way of thinking is nearly irresistible in its simplicity, once one is (unlike Merleau-Ponty) convincedof the absoluteness of an object-subject dichotomy. The therapist, by having particular subjectiveexperiences, can assume that the patient has similar subjective experiences, due to the general rule thatthere must be some analogous information processing in both.

We find this clearly illustrated in the most famous (1912) of Freud’s so called papers on psychoanalytictechnique. First, Freud recommends in this paper

"not directing one’s notice to anything in particular and in maintaining the same ’evenly-suspendedattention’ (...) in the face of all that one hears" (pp. 111/112).

This, Freud claims, prevents the therapist from selectively perceiving the materials that the patientpresents, as well as from dismissing those materials the meaning of which may become clear only muchlater.

We may easily recognize in this recommendation the clarity of mind and the ensuing accessibility to one’sown unconscious, that which in this chapter has been called ’exploration’. This attitude avoids theselection of particular topics of conversation and objects of attention. In our terms: it avoids theconstruction of second order distinctions, in favor of first order distinctions.

Furthermore,"[The different rules I have brought forward] are all intended to create for the doctor a counterpartto the ’fundamental rule of psycho-analysis’ which is laid down for the patient. Just as the patientmust relate everything that his self-observation can detect (....) so the doctor must put himself in aposition to make use of everything he is told for the purposes of interpretation and of recognizingthe concealed unconscious material without substituting a censorship of his own for the selectionthat the patient has forgone." (p. 115)

Thus far Freud’s recommendations seem to fit in well with the ideas presented in chapter 7 on theinterpersonal interface as a limit for critical observation.

The therapist is to use ’his unconscious (...) as an instrument in the analysis’ (p. 116), and he is to havean awareness of possible resistances and complexes in himself, so that his access to his own unconsciousis optimal: "It may be insisted (...) that he should have undergone a psycho-analytic purification ..." (p.

477. e.g. Wexler (1974); Lietaer, Rombauts, van Balen (1990)

478. Instead, they should be performed with ’phronesis’ (I will introduce Gadamer’s term ’phronesis’ in section 9.4).

479. "... in order that the projection [of my intimate experiential data into the other] would take place, it would be necessary that Irely on the analogy between the physiognomic expressions that the other offers me and the different physiognomic gestures that Iperform myself." (my translation).

166

116) So far, a ’third position’ could have been meant by Freud.

However, immediately after the lines from page 115 quoted here, Freud’s argument takes a sudden turn,when he elaborates on the term ’transmitting’:

"To put it in a formula: he must turn his own unconscious like a receptive organ towards thetransmitting unconscious of the patient. He must adjust himself to the patient as a telephonereceiver is adjusted to the transmitting microphone. Just as the receiver converts back intosoundwaves the electric oscillations in the telephone line which were set up by sound waves, so thedoctor’s unconscious is able, from the derivatives of the unconscious which are communicated tohim, to reconstruct that unconscious, which has determined the patient’s free associations"(pp. 115-6).

Thus, Freud seems to be saying, the therapist’s unconscious is not in immediate contact with the patient’s,as in first order consensus, but, instead, it is possible to reason by analogy and make valid inferencesabout the patient’s unconscious.

Freud describes a mode of therapist explorative performance (evenly-suspended attention) that hemighthave described in terms of such a third position. Why does not he do so? To put it in his own terms:because he wanted to put it in a formula! And, we may add, reasoning-by-analogy is a formula, a logicalprocedure that can be executed, but the evenly-suspended attention is not. Instead, it is the abandonmentof a critical position, the loss of an archimedean point, the absence of control.

The major advantage of offering a formula is that it provides the reader with an intimation of certainty interms of a procedure that has to be followed. Accordingly, a personal intuition, obtained as a result of ashared first order distinction (attained through evenly-suspended attention), can be reconstructed andpresented as the result of a logical inference. After the reconstruction the therapist’s accounts look moresolid.

For to introduce a new discipline to an unfriendly audience, and to argue that there is a necessaryconfusion between the procedures followed and their application processes, is a secure way to being takenless seriously than a discipline that is described in more procedural terms. Precisely this was the dilemmathat major theorists in the field of psychotherapy wanted to avoid. However, I think, the confusionbetween procedures and their application should be considered as an empirical phenomenon inherent topsychotherapy, and it should be studied accordingly. One way of doing so is by means of a ’parallelprocess’, in which a ’double mimesis’ can occur. I will discuss this in chapter 10. Furthermore, I think,the field of psychotherapy should take this confusion as its hallmark, instead of attempting to conceal it!

9.4. Gadamer: phronesis and the aesthetic non-distinction

Above in section 9.2.2 I mentioned the usage of poems as a mode of expressing the impossible. Could weuse poems as non-classical models for the self-referential qualities of psychotherapies? I would not thinkso. But we might use some aspects of the ’logic of poems’. In this section I will discuss this ’logic ofpoems’ as an ’aesthetic non-distinction’, and in chapter 10 I will propose that we can build a researchsetting that bears this property.

The act of exploration, as it is called in this text, has been elaborated by Gadamer (1960/1986a) a.o. in histreatments of game/play (both ’Spiel’ in German«480»), art, and practical wisdom (phronesis).

480. I follow Palmer (1969, pp. 171, 175) in that we might better translate here the German ’Spiel’ as ’game’. Palmer remarks:A game, as a playful activity, "... is a self-defining movement of being into which we enter. (....) Our participation in the game

brings it into a presentation, but what is presented is not so much our inner subjecivities as the game: the game comes to stand, it takesplace in and through us." (p. 172).

On the other hand, "... the play achieves its meaning only as a presentation. Its real meaning is a matter of mediation. It doesnot exist primarily for the players but for the viewer. The play is as hermetically sealed and self-sufficient as any game, butas play itpresents itself aseventto the viewer." (p. 173)

167

Gadamer’s ideas on playing, art and phronesis are highly congruent with the more formal considerationsregarding the limitations of self-descriptive systems, such as those elaborated by Löfgren«481».

We encounter the idea of self-relatedness in Gadamer’s treatment of ’phronesis’. This term is given byAristotle as denoting a social and ethical practical wisdom, and it is opposed to ’techne’ (we wouldnowadays use its derivative term ’technology’), that proceeds according to a certain plan or blueprint. Incontrast to techne, Gadamer maintains that phronesis takes shape only as it is applied, e.g. in jurisdiction.Phronesis is a continuous effort to deal with general rules as well as with particular situations:

"... but the moral knowledge that is directed towards these guiding principles is the sameknowledge that has to respond to the demands of the situation of the moment". (Gadamer, TMp. 287)«482».

This double orientation cannot happen through standardized techniques and explicit procedures, and it is amatter of personal courage for the practitioner to accept that.

In phronesis, man disposes of himself in a way different from the ways in which he disposes of otherthings. Gadamer maintains that the kind of knowledge involved must be different accordingly. He followsAristotle and calls it ’a self-knowledge’ (TM p. 282)«483».

"The self-knowledge of moral reflection has, in fact, a unique relation to itself" (TMp. 288)«484»

It is tempting to impute to Gadamer here a concept of self-reference of the kind that does not allowreformulation in terms of an executable procedure. This temptation is enhanced after reading:

"Where there is a techne, we must learn it and then we are able to find the right means. We see,however, that moral knowledge always requires this kind of self-deliberation. Even if we conceivethis knowledge in ideal perfection, it is the perfection of this kind of deliberation with oneself(euboulia) and not knowledge in the manner of a techne" (TM p. 286)«485»

Furthermore:"Moral knowledge is really a knowledge of a special kind. It embraces in a curious way bothmeans and end and hence differs from technical knowledge. That is why it is pointless to

481. Unfortunately, on a previous occasion in which I dealt with this relation between self-description and Gadamer’s hermeneutics(Goudsmit, 1992b) I had not yet come across Weinsheimer’s (1985) discussion of Gadamer. His elaboration offers some vital supportfor my present purposes. He writes:

"[Gadamer’s] antipathy to artificial-language schemes blinds him to the coincidence of his views with those of the prominentmathematical logicians we have considered. The reasons for this antipathy are clear. Constructing an artificial language implies thatlanguage is a tool like any other. (....) This instrumentalism implies that we stand outside language. It is no less manipulable than theobjects it makes manipulable. Artificial language thus exemplifies the dominion of subjectivity, which puts everything at our disposaland under our control.

But this everythingis, as we have seen a serious oversimplification; and Gadamer knows that it is. Indeed, that method neverexhausts truth is his thesis. What he does not acknowledge is that the advocates of method know this too. (...) All natural science, asobjective, is not itself what it studies. (...) Either an objective science is incomplete because it excludes itself from what it studies; or, ifit claims to include itself, it is self-contradictory, since the science (as objective) provides its own exception. Either objective andincomplete, or complete and inconsistent - that is the dilemma of natural science, whether from the point of view of hermeneutics orthat of mathematical logic.

But are not the human sciences stretched on the this same rack? In them, the researcher is himself what he studies. The humansciences are therefore not objective, but neither are they even potentially complete." (pp. 56-8)

Clearly, the idea of linguistic complementarity fits in very well with this description.

482. ".... das sittliche Wissen, das auf diese Leitbilder gerichtet ist, ist das gleiche Wissen, das die fordernde Situation desAugenblickes zu beantworten hat." (WM p. 305)

483. ’ein Sich-Wissen’ (WM p. 299)

484. "Das Sich-wissen der sittlichen Überlegung besitzt in der Tat einen einzigartigen Bezug auf sich selbst." (WM p. 306)

485. "Wo es eine Techne gibt, muss man sie lernen und dann weiss man damit auch die rechten Mittel zu finden. Umgekehrt sehenwir aber, dass das sittliche Wissen stets - und auf unaufhebbare Weise - solches Mitsichzurategehen verlangt. Selbst wenn man sichdieses Wissen in idealer Vollendung denkt, ist es die Vollendung solcher Beratschlagung mit sich selber ([euboulia]) und nicht einWissen in der Art der Techne" (WM p. 304)

168

distinguish here between knowledge and experience, as can be done in the case of a techne. Formoral knowledge must be a kind of experience ..." (TM p. 288)«486»

The differentiation in techne between knowledge (cf. universals) and experience (cf. particulars) is notmade in the case of phronesis. We are reminded of Vico’s ’imaginative universals’ (cf. section 3.3). Theway in which moral knowledge embraces both means and ends is in that it doesnot allow a description ofthe means (method) irrespective of the end. Or, in other words: in phronesis, the ’application process’, asI have called it in section 9.2.2, is considered too much part and parcel of the ’means’ (i.e. the procedure)to be ignored in a description of the actor’s performance. Only at the price of eliminating this applicationprocess from our attention, can a context-free description of the ’means’ be given. In phronesis, however,this is impossible. Procedure and application process overlap!

In a similar vein, with respect to playing games Gadamer writes:"Play«487» is really limited to representing«488» itself. Thus its mode of being is self-representation" (TM p. 97)«489»

and:"The mode of being of play does not allow the player to behave towards play as if it were anobject. The player knows very well what play is, and that what he is doing is ’only a game’; but hedoes not know what exactly he ’knows’ in knowing that" (TM p. 92)«490»

The game exists as a self-presenting activity in which the player is involved in playing. This involvementmakes him ’not knowing’ what the game is. At least, a detached observation of the game by the player isnot possible as something apart from the player; and the player does not notice this, since there is nopoint of view from which he could perceive it«491». Rather than a player who plays the game, thegame is considered as becoming the subject of the playing, not its object (TM pp. 95/6)«492». It isan activity to which external, detached, control by the player does not pertain.

With respect to works of art, especially musical or dramatic, Gadamer extends his ideas to ’play’, i.e. theperformance of a piece of art:

".... that the non-differentiation of the interpretation from the work itself is the actual experience ofthe work" (TM p. 107)«493»

486. "Das sittliche Wissen ist wirklich ein Wissen eigener Art. Es umgreift in einer eigentümlichen Weise Mittel und Zweck undunterscheidet sich damit vom technischen Wissen. Eben deshalb hat es auch keinen Sinn, hier zwischen dem Wissen und der Erfahrungzu unterscheiden, wie das sehr wohl bei der Techne angängig ist. Denn das sittliche Wissen enthält selbst eine Art der Erfahrung insich ..." (WM p. 305)

487. see footnote 480 on page 166, on the translation of the German ’Spiel’

488. perhaps better: ’presentation’ (cf. footnote 115 on p. 51).

489. "Das Spiel ist wirklich darauf beschränkt, sich darzustellen. Seine Seinsweise ist also Selbstdarstellung." (WM p. 103)

490. "Die Seinsweise des Spieles lässt nicht zu, dass sich der Spieler zu dem Spiel wie zu einem Gegenstande verhält. DerSpielende weiss wohl, was Spiel ist, und dass, was er tut, ’nur ein Spiel ist’, aber er weiss nicht, was er da ’weiss’" (WM pp. 97-98)

491. Cf.: "The game (...) has its own nature, independent of the consciousness of those who play it." (Palmer, 1969, p. 174)

492. WM p. 102cf. also: "This is the eternal origin of art that a human being confronts a form that wants to become a work through him. Not a

figment of his soul but something that appears to the soul and demands the soul’s creative power. What is required is a deed that a mandoes with his whole being: if he commits it and speaks with his being the basic word to the form that appears, then the creative poweris released and the work comes into being." (Buber, 1970, p. 60)["Das ist der ewige Ursprung der Kunst, daß einem Menschen Gestalt gegenübertritt und durch ihn Werk werden will. Keine Ausgeburtseiner Seele, sondern Erscheinung, die an sie tritt und von ihr die wirkende Kraft erheischt. Es kommt auf eine Wesenstat des Menschenan: vollzieht er sie, spricht er mit seinem Wesen das Grundwort zu der erscheinenden Gestalt, dann strömt die wirkende Kraft, das Werkentsteht." (1983, p. 16)]

493. "... dass die Nichtunterscheidung der Vermittlung von dem Werke selbst die eigentliche Erfahrung des Werkes ist" (WMp. 114)As said, I would prefer the translation ’non-distinction’ for ’Nichtunderscheidung’

169

Furthermore, in regard of (aesthetic) play, also an external observer’s position is affected in a way that he,too, becomes involved:

"Rather, openness towards the spectator is part of the closedness of the play" (TMp. 98)«494»

What does this mean? Wouldn’t it be most reasonable to distinguish the work of art from its performanceor execution? It is at this point that Gadamer forcefully opposes what has become named the romantic-subjectivist conception of aesthetic awareness:

"My thesis, then, is that the being of art cannot be determined as an object of an aestheticawareness because, on the contrary, the aesthetic attitude is more than it knows of itself. It is a partof the essential process of representation«495»and is an essential part of play as play" (TMp. 104)«496»

Again: what can be meant by ’is more than it knows of itself’? We met this also in the above quotation(TM p. 92) on games. How can it be that a work of art is not what the romantic-subjectivist conceptionhas taught usbecauseour awareness of it is incomplete? Whence this ’because’? It is from the fact thatart, as aesthetic play, is not to be situated in a domain separate from the observer. It exists in a way thatthe observer is necessarily part of it; necessarily, that is: no critical distance is possible without losing thework of art. Thus, Gadamer presents as his alternative:

"... by opposing to aesthetic differentiation, the properly constitutive element of aestheticconsciousness, ’aesthetic non-differentiation’. It has become clear that what is imitated in imitation,what is formed by the poet, represented«497»by the actor, recognised by the spectator is tosuch an extent what is meant - that in which the significance of the representation lies, - that thepoetic formation or the achievement involved in the representation are not distinguished from it.When a distinction is made, it is between the material«498»and the forming«499»,between the poem«500»and the ’conception’. But these distinctions are of a secondarynature. What the actor plays and the spectator recognises are the forms and the action itself, as theyare intended by the poet. Thus we have here a double mimesis: the writer represents and the actorrepresents. But even this double mimesis is one: it is the same thing that comes to existence ineach case.

More exactly, one can say that the mimic representation of the performance brings to being-there what the written play actually requires. The double distinction between a drama«501»and its subject matter«502»and a drama and performance corresponds to a double non-distinction as the unity of the truth which one recognises in the play of art" (TMp. 105)«503»[italics added]

494. The German text states it somewhat stronger: "Das Offensein zum Zuschauer hin macht vielmehr die Geschlossenheit desSpieles mit aus." (WM p. 104). One might think of a translation like: ’partly makes up’, instead of ’is part of’.

495. perhaps better: presentation (cf. footnote 115, page 51)

496. "Die These ist also, daß das Sein der Kunst nicht als Gegenstand eines ästhetischen Bewußtseins bestimmt werden kann, weilumgekehrt das ästhetische Verhalten mehr ist, als es von sich weiß. Es ist ein Teil desSeinsvorganges der Darstellungund gehört demSpiel als Spiel wesenhaft zu." (WM p. 111)

497. cf. footnote 115

498. Stoff

499. Gestaltung

500. Dichtung

501. Dichtung

502. Stoff

503. "... daß wir der ästhetischen Unterscheidung, dem eigentlichen Konstitutivum des ästhetische Bewußtseins, die’ästhetischeNichtunterscheidung’entgegenstzen. Es war klargeworden: das in der Nachahmung Nachgeahmte, vom Dichter Gestaltete, vom SpielerDargestellte, vom Zuschauer Erkannte is so sehr das Gemeinte, das, worin die Bedeutung der Darstellung liegt, daß die dichterischeGestaltung oder die Darstellungsleistung als solche gar nicht zur Abhebung gelangen. Wo man doch unterscheidet, wird von der

170

The work of art is considered’Stoff’ (translated as ’material’ or ’subject matter’) that is written down byits creator as’Dichtung’ (translated as ’poem’ or ’drama’), and interpreted by its performer as’Aufführung’ (performance). The drama and the subject matter are said to find their truth in theirbecoming indistinguishable; the same is said with respect to the drama to be interpreted and itsperformance. Indeed, this is a state of affairs that cannot be understood as a technical achievement! Forhow could, technically speaking, the work of art, as subject matter created by its composer, beindistinguishable from the drama as written down; how could this work as written down becomeindistinguishable from the work as read (performed) from its written version?

The very notions of writing and reading imply that the written version (the code) existsindependently, andis demarcated from that of which it is a description (by a writing operation), as well as from that of whichit is a prescription (by a reading operation). These demarcations are defined to occur in time and to yieldentities that are ontologically distinct by these very operations. These are some of the assumptionsinherent to the concepts of reading and writing. But Gadamer suggests a different concept of temporalityin matters of aesthetics. The indistinguishabilities between the work-as-created and the work-as-written, aswell as between the latter and the work-as-performed, are only possible if we radically discard the notionsof reading and writing from the respective domains in which games and plays/works of art exist. If thereis an aesthetic non-differentiation between ’Stoff’, ’Dichtung’ and ’Aufführung’, and if thisindistinguishability is crucial for the very existence of the work of art, then ’Stoff’, ’Dichtung’ and’Aufführung’ cannot be considered as separately existing entities, distinguished in time by the operations’write’ and ’read’ respectively.

But whether this aesthetic non-differentiation is the case, is a matter of attitude. It is precisely thedifference between what I have called in chapter 4 the difference between ’naive’ and ’critical’ observers!It is the critical observer who is able to make a distinction between a substance and its accidents, andlikewise to distinguish between a work of art and its performance«504»; or between a theme and theway in which it is poetically written down by the artist, or between the official rule of a game and theway in which the game is actually played. All these distinctions are possible, but only by a criticalobserver, not by a naive observer who is involved in a play (irrespective whether as an actor or as aspectator).

Thus understood, playing (games) exists due to the player’s capacity to abandon critical observations andto entrust himself to the game«505». We may ’playfully’ call this abandonment of the criticalperspective an act of ’un-perception’.

The reader is reminded of what Winnicott (1951, p. 15) has described in terms of ’being lost in play’: thechild does no longer take a distance from the object of his play; it becomes a ’transitional object’, i.e.: an

Gestaltung ihr Stoff, von der ’Auffassung’ die Dichtung unterschieden. Aber diese Unterscheidungen sind sekundärer Natur. Was derSpieler spielt und der Zuschauer erkennt, sind die Gestalten und die Handlung selbst, wie sie vom Dichter gemeint sind. Wir haben hieralso einedoppelte Mimesis: der Dichter stellt dar und der Spieler stellt dar. Aber gerade diese doppelte Mimesis isteine: was in dereinen und in der anderen zum Dasein kommt, ist das gleiche.

Genauer kann man sagen: die mimische Darstellung der Aufführung bringt das zum Da-Sein, was die Dichtung eigentlichverlangt. Der doppelten Unterscheidung von Dichtung und ihrem Stoff und von Dichtung und Aufführung entspricht eine doppelteNichtunterscheidung als die Einheit der Wahrheit, die man im Spiel der Kunst erkennt." (WM pp. 111/2)

Notice that above in section 3.4 on page 51 I gave a quotation that immediately follows the one given here.

504. We are reminded of Yeats famous line "how can we know the dancer from the dance?", from the poem ’Among schoolchildren’ (also to be found at http://www.iscm.ulst.ac.uk/isc/Yeats/AmongSchoolChi). The complexity of the non-distinction referred toin this line has been discussed a.o. by Culler (1982, p. 246).

Cf. also: "The paradigmatic dance experience is one of pure movement. It is a solo dance in which no story is being told and inwhich movement stands for nothing but itself. The dance is a purely kinetic happening; it is a dance in which I am moving for the purejoy of movement. (...) There is, in effect, a union of myself and the dance. It lives through me at the same time that I live and breathe init. The density of the dance and the density of my being thus come together: we meet in the sheer physicality of movement which isboth me and the dance. (...) I am nothing more than the movement and the movement is nothing more than me. How indeed tell thedancer from the dance?" (Sheets-Johnstone, 1986, p. 235-6)

505. Cf. also Verden-Zöller’s (1989) remarkable experiences in the field of mother-child play.

171

object that is neither internal nor external to the player. Rather, the transitional object is said to reside in athird domain.

Comparably, a work of art can be appreciated and recognized as a ’hermeneutic identity’ due to theattainment of a naive perception by those who are committed to it as observers«506». For theobserver of a work of art it is this ’un-perception’, that enables him to incorporate already perceivableobjects (called the ’concrete objects’ in 4.2) and enter into a naive, i.e. aesthetic perception, throughwhich the work of art exists.

9.5. The ’hermeneutic identity’ of psychotherapy

The truth that, according to Gadamer, the poem or drama (’Dichtung’) and the subject matter (’Stoff’)find in their becoming indistinguishable, is a truth that would not fit in the scheme of section 6.1.1 (page90). It is neither a ’real truth’ nor a ’proved truth’. We might call it a ’semantic truth’. It is this kind oftruth, I think, that contains a ’logic of poems’ (as I called it above), and it is useful for us when dealingwith self-reference. The performance of psychotherapists cannot be described as a procedure plus anapplication process that is distinct from it, but, like the drama (’Dichtung’) and the subject matter(’Stoff’), their indistinguishability may be part of the ’double mimesis’: the poetic formation (’dichterischeGestaltung’) and the (actor’s/performer’s) achievement of presenting it (’Darstellungsleistung’) may allcome together.

Thus, if we compare the therapist’s work with the poet’s, we may understand the ways in which he givesshape to his performance as the application of a procedure, but also as the ’Dichtung’ of a particular’Stoff’ «507». For instance, the therapist knows that he should investigate the patient’s resistancesfirst, but he decides to postpone this, as he finds the patient too upset to do any reflections. Instead, heoffers some more accommodating remarks, which bring the patient to comment on something else. Or, togive another example, the therapist decides that a particular issue would be most relevant to discuss withthe patient at the next session; this is a kind of blueprint or ’Stoff’ for the next session. But when thepatient enters and starts with an entirely different topic, apparently fully unrelated to last session’smaterials, the therapist may decide not to break in and first explore for himself (and perhaps also togetherwith the patient, at a later moment) what connection there might be between the previous and the currentsubject matters; this is ’Dichtung’. It is a matter of what I called (section 8.5.5) a critical use of naivediscourse: the therapist decides to suspend his thoughts, and have himself first confused by the patient’snew story in order to explore the new materials freely.

Thus, there is some ’pure’ procedure, still unapplied, that I compare with a basic theme that serves as thepoet’s ’Stoff’. Furthermore, a therapist may relate about his performance with this patient, for instance toa supervisor or to an investigator, in such way that the latter, like an audience, may receive the story andperceive the ’what’ and the ’how’ of the presentation as a unity, without (critically) distinguishingbetween them.

506. Cf. the emergence of ’poetic’ objects according to Vico (section 3.3). See also footnote 431, page 146.

507. I recently came across some historical remarks on Mozarts opera ’Idomeneo’, of which the construction process has beendocumented fairly well in the composer’s letters to his father Leopold. It is remarkable to which extent Mozart took into account thevarious strong and weak qualities of the musicians in the cities where this opera was to be performed (notably Munich and Vienna). Asthe famous conductor Harnoncourt (1980/1984, p. 37) puts it concerning his interpretation of the opera: "... that Mozart, as with otheroperas of which there are known to be later versions, additions or changes, always sought and achieved for each performance the bestpossible dramaturgical context and musical balance. We cannot, therefore, simply join up the "best bits" from the various versions; allour notions of "improvement" have always proved futile. Mozart cannot be improved upon; he is invariably right."

172

Gadamer’s categories: comparable to:

Stoff therapist’s problem definition & procedure

Dichtung/ therapist’s application processdichterische Gestaltung

Aufführung/ therapist’s presentation toDarstellungsleistung supervisor/investigator

The presentation ’performed’ by the therapist, then, i.e. his story about the therapy session, may bearproperties of the therapy session spoken about. This phenomenon is widely known as the ’parallelprocess’. It is widely ignored in circles of scientific investigations, but it is a major instrument forsupervisors. More about it in chapter 10.

10Observing therapeutic events through parallelprocessesParallel processes in psychotherapeutic interactions have been described extensively in predominantlypsychoanalytic literature (Ekstein & Wallerstein, 1958; Caligor, Bromberg & Meltzer, 1984; Sachs &Shapiro, 1976; Searles, 1955; Arlow, 1963; Doehrman, 1973). Briefly stated, it amounts to thephenomenon that a patient may induce in his therapist several kinds of experiences, to the effect that thetherapist starts to treat the patient or himself in ways similar to how the patient treats others, or himself.Likewise, as a repercussion of the therapeutic relationship, the therapist may start to treat others (e.g.supervisors, colleagues, etc.) either similarly to how he feels himself treated by the patient, or similarly tohow he treats the patient, thus triggering in these other persons responses comparable to those by thepatient.

The parallel process will be proposed in this chapter as a paradigm for research attempts to ’get inside’, toenter into a therapeutic process. As such, an investigator of psychotherapeutic processes is in a positionsimilar to the observer of an organizationally closed system: he is faced with a one-sided boundary, and inorder to pass it he must ’forget’ it. This is possible by getting involved in a parallel process, in which theresearch process obtains characteristics derived from the therapeutic process that is investigated.

Psychotherapy, when concerned with discovering the uniqueness of an individual patient, necessarilyentails the therapist’s personal commitment and his explorative self-concern. This personal involvement isthe only thing that cannot be reformulated in procedural terms. Though many of the activities ofpsychotherapists could be reformulated in the strict procedural terms of a context of justification,exploration, as their core activity, would resist such a reformulation. If a psychotherapist is to maintainexploration in his repertoire, it must be by embedding it in activities that mightotherwiseperhaps bebetter performed by the rigid application of a well describable procedure, but whichnow are notperformed in that way. All therapies, not only the so called uncovering types of psychotherapy, such aspsychoanalysis and client-centered therapy, but also the behavior therapies, contain explorative parts.

If these explorative parts are to be investigated empirically, then it is too appealing to look for proceduresthat logically reconstruct (cf. section 9.1) them, as a ’logic of discovery’. Indeed, for any explorativetrajectory, once it has been accomplished, a procedure can be formulated retrospectively that exactlysimulates it, or perhaps even one that obtains better outcomes in less time. Thus, a behavioral technologyhas in fact become possible, to the extent that ’standardized’ therapies were found to yield even betterresults than ’tailor-made’ ones (cf. Schulte, Künzel, Pepping, Schulte-Bahrenberg, 1992). The point is, ofcourse,that a procedure alone would not enable the therapist to explore the problem; but once thelaborious job of exploration has been done, the erroneous paths become apparent, afterwards. On the otherhand, if explorations are not required, for instance when the problem definition is exhaustively clear, then’standardized’ therapies do not need to be complemented by an explorative component«508».

How can we empirically investigate the therapeutic processes qua explorations, without imposing uponthem some procedural format? Any valid answer to this question should pertain to the way in whichexplorations are brought forth as self-referential processes (cf. chapter 9). We may empirically finddifferences here. The self-referential processes involved in play and exploration may go more or lesssmoothly. If they go less smoothly, then there is a need for an audience thatsupportsthe closure to

508. Comparably, once a mathematical expression has been formulated, a computer can calculate with it more quickly than a humanbeing. The latter, however, is indispensible for interpreting the semantics of a particular situation, in order to produce the syntacticexpressions that the computer requires. Cf. my quotation from Rosen (1991) on semantics (see footnote 264, page 96)

173

174

develop. Gadamer’s ideas on the role of the audience (of a play or of an aesthetic performance) gain anew, clinically relevant, dimension here«509»!

If the therapist can freely explore his own exploration processes, play with, what I called, his veryopportunities to play (cf. section 9.2.1), then the object explored will easily gain a hermeneutic identity(cf. sections 3.2 and 9.5) of its own. On the other hand, if the therapist cannot freely play with thesituation, then this means that his opportunities to utilize his own experiences, to explore them and to playwith the situation encountered, are diminished. There will be less of a hermeneutic identity of thetherapeutic process studied. Instead, the therapist will look for opportunities to involve others, (acolleague, a supervisor, or possibly an interviewer) into his attempts to attain a self-referential explorationand see the hermeneutic identity of the therapy he is dealing with. As a result, the supervisor orinterviewer may feel invited, or even pushed, to sustain a particular naive object, explore it, endure in themean time the therapist’s incapacity to do so. This is what happens in the interview to be presented belowin section 10.2.

I will proceed as follows. First, I will discuss in section 10.1 the phenomenon of the parallel process as aspontaneously occurring feature of a research process. I will focus upon the utilization of thisphenomenon for passing the therapeutic system’s boundaries. Then, in section 10.2 I will show how theparallel process can be used as a tool for looking at therapeutic processes where exploration is notperformed entirely freely.

10.1. Utilizing the parallel process for the passage of one-sided boundaries

An autonomous system’s boundary can only be defined from the system’s outside, there where theobserver is. It is the environment’s boundary with respect to the system, but not vice versa! At the otherside of this boundary there is an absence of distinction between inside and outside, that impedes theobserver to sensibly point at connections without describing quasi-objects«510»(cf. section 7.3.2).

The dichotomy between ’inside’ and ’outside’ can be elaborated with respect to therapeutic systems. Myparticular aim will be here to show how one can be advertent of one-sided boundaries of autonomoussystems, within research of psychotherapeutic processes.

When a therapist has himself guided by a client system«511»and thus comes to experience andperceive actually the issues spoken about, it becomes possible for him to pass a one-sided boundary. Hethen gets imprisoned in the client system’s network of constructed facts, in such a way that he can see nolonger how he participates in the construction of the problem. Instead, he experiences how understandable,or how logical, self-evident or unavoidable the client system’s naive objects are. Beyond this boundary acritical problem definition or diagnosis, conceptualized beforehand, does no longer hold, if one did nottake into account that within the system such a critical position is absent. For after having passed thesystem’s boundary, after having become a participant of it«512», the therapist loses sight of itsboundary and his ways of reflecting become constituent parts of it. Then it can be a hard job for atherapist to get outside the system again, like it can be an impossible task for a patient to escape from a

509. cf section 9.4 and footnote 494, page 169; cf. also section 11.2.2

510. i.e., objects that do not exist as such for the person observed (see section 2.2.3 for an introduction of the concept ’quasi-object’)

511. irrespective whether this system consists of one single individual, a (married) couple or a family

512. Obviously, it is often no problem to participate in a system (e.g. an association), while maintaining one’s clarity about thesystem’s boundary. Such a system, howver, is not autonomous in the way meant here.

175

closed (family) system«513». On the other hand, it can also be very difficult for an externalobserver to approach an autonomous system in such a way as to get inside and to contribute to itsconservation.

These two movements, from inside to outside and vice versa, are passages of the one-sided boundary, andhence they are in no way trivial acts. How do they take place? Without too much effort we can give anabstract answer, using our definition of a one-sided boundary. The passage from inside to outside happensby distinguishing again ’inside’ from ’outside’. The passage in the other direction happens by losing orforgetting the difference between ’inside’ and ’outside’. This brings us to a peculiar difference betweenthe two directions of passage. It is only with explicit awareness that one can succeed in the directioninside-outside. In the other direction, however, the passage is only possible if it happenswithoutawareness! One does experience the passage, but one is not advertent of it.

However, the reader may object here that these abstract answers do not yield any directions to thepracticing therapist. Instead, let us look for a more practicable clarification. I will proceed as follows. Insection 10.1.1 I will present an example in which a family demonstrates its enormous capacity to absorbnot only the therapist, but also his colleagues who observe him from behind a screen. As a result, theyhad serious difficulties quitting the system. Here I discuss the passage from inside to outside. Next, insection 10.1.2 I will give an example of a therapy researcher who enters into a therapy that is hard toaccess. Here I discuss the passage from outside to inside. In both examples the role of parallel processeswill be apparent.

10.1.1. The passage from inside to outside

10.1.1.1. Case

The presenting problem of a family (father, mother, son (13), two daughters (11 and 15)) is that the son isso intractable to his parents. It becomes quickly apparent to the therapeutic team that the parents do notsufficiently support each other when constraining the boy. Thus, his behaviors could be a means, used bythe parents, to express their feelings of anger against each other. Furthermore, it becomes quickly clearwhat has to be done: the parents should be taught how to support each other in their acts of showing theboy his limits; and they should be taught how to deal with their conflicts in a different way. So far fortheoretical insights. Now practice. It turns out that the therapist does not feel capable to discuss with thefamily the issue of ’putting limits’, without also feeling responsible, in an inappropriate way, for thefamily’s problems. This is not how the therapist wants it. To talk with the family, especially with theparents, about constraining the boy, would mean to the therapist that he should also constrain the parentsand put a limit to their behaviors. The therapist is slowly developing a sense of being incapable of solingthe family’s problems, in a way comparable to how the members of the family feel.

513. cf: "I wasstuck in the situation, just as he was. There was no resolution, just this exasperating, anxiety laden,not clearthinking processgoing nowhere. I felt that I could not move or alter the situation to make cohesion out ofmy confusion, which mirroredor was bent to the shape of the therapist’s confusion. It was a if I was in this bind with these feelings and perceptions in me and I hadhad to resolve this situation for myself." (Caligor, 1984, p. 26-7)

Arlow describes this in terms of "how therapist and patient may share certain fantasy wishes in common. Such a situation (...) isone of the principal causes of the so-called "blind spot."" (Arlow, 1963, p. 582)

To give a clinical vignette from my own therapeutic work: A patient in his early thirties is working through the pains ofseparating from his family of origin, especially from his siblings, with whom he had always had a sense of allegiance of the variety: "bysticking together, we have found the best solution possible for us in view of our terrible parents". It is into this domain of living that thetherapist is introduced in a long series of weekly therapy sessions, to the extent that the therapist eventually experiences a) a stronginjunction to save this patient from the quagmire he is drowning in, in addition to b) a full incapacity to do more than a weekly attemptat his reanimation. It is here that the therapist, finally, himself comes to experience this patient’s sense of suffocation. Later in thetherapy the patient mentions that his problem with the sibling group is that he cannot save the others, though he desperately would liketo, because they cannot share with him the idea that the world is larger than their ’best solution’: this idea is external and alien to this’best solution’; on the other hand, if the patient reenters the sibling group and joins them in order to convince them from within, heloses sight of what he wanted to convey to them concerning the external world, and they have little work in explaining his ideas interms of his madness.

176

The therapy sessions consists of talking about the problems, however without coming to grips with them.During the sessions the son is usually looking into other directions, in a sulk, playing with his watch bandor something similar. He withdraws as much as possible from the conversations. Nevertheless, one timehe mentions in passing that he does not believe in talking. To talk, he says, is something entirely differentfrom doing things. He doesn’t give a dime for talks. On the other hand, the parents’ preference for talkingfits in very well with the verbal medium of the therapy. Also the therapist extensively uses words.

Then, the team of observing colleagues, sitting behind the one-way screen and watching the therapy,decides to make an intervention. Observer1 envisions to make a remark of conciliation, like "there isalways a lot of mutual acceptance in a family, for otherwise the family members cannot have conflictswith each other". Observer2, on the other hand, envisions to say "why does the therapist treat the parentswith so much consideration?", being convinced that some of the family’s conflicts are due just to a lackof mutual acceptance. But by the time she has to perform her intervention, she does not manage, unlikeObserver1, to voice her words; she gets entangled in her own sentences«514», and, to make thingseven worse, she is interrupted by her colleague, observer1, who makes an approving and appreciatingremark about what she is saying. But she feels obstructed by this approving interruption, just because shedid not agree with him about his words on mutual acceptance.

In the after session discussion it becomes apparent that the therapist feels as if he is expected to take overfrom the mother the task of taking care of the children. Mother has installed in him part of her despair, soit seems, and during a major part of the session she leans back, without entering into the conversation. Wefind here an instance of projective identification (see chapter 11). The therapist’s identification with herhas also become apparent in the session, in a moment of slip, when, speaking about the married couple,he says to the father ’one of us two’, which he immediately rectifies with ’one of you two’. After thesessions the therapist recurrently expresses his concern with the great contrast he experiences between thetherapeutic sessions and the after session discussions with the team: "now it is all clear to me, but nexttime I will be back in the [therapy] room and I will be stunned again."

Furthermore, in the after session discussion observer2 reproaches observer1 of having interrupted herunnecessarily. She is also surprised about her own vulnerable response to his behavior, which she usuallydoes not have. It becomes apparent that both observers disagree about their positions: observer2 believesthat they basically disagreed, whereas observer1 believes that what observer2 did say was in fact in supportof his own opinion. It should be noticed here that the ’conflict’ between the two observers has the shapeof a one-sided boundary: at the side of observer2 the fundamental differences between the two observersdo exist; at the side of observer1 they don’t exist! Can we utilize this circumstance for a betterunderstanding of what is wrong with the therapy?

We find that both the therapist and his observing colleagues are not capable of addressing the therapeuticsystem with respect to the issue of ’putting limits’. As clearly as one can conceive things in advance, soconfusing these things turn out to be when one has the family face them.

10.1.1.2. Elaboration

By the mere fact of having conversations with he family, the therapy has become an operation against theboy. To ’talk about’ is not an innocent activity; it is part of the family’s problem, and the therapistunwittingly commits himself to one of the parties«515». Just like the therapist, also the observingcolleagues have become participants of the autonomous process that initially they observed from theoutside. At the other side of the one-sided boundary it is no longer possible to reside at a metapositionwith respect to the system, that is not also part of the system! The therapist and his observing colleagues

514. cf. Olthof & Vermetten (1993)

515. An interesting way out of such a dilemma has been described by Ogden, when he decided to offer the patient silences "whichserved as an interpretation that could not have been made in words because of the perversion of language and thougth that was beingenacted in the analysis" (Ogden, 1994b, p. 229).

177

do not realize the degree to which they have become partial and controversial by offering ’talking’ as acure. Their talking and analyzing does not offer an escape to a more external position. To the contrary, itsustains the problem. The very attempt at communication at a metalevel does not effectuate themetacommunication as desired; instead, it confuses levels, so that, as a result, all verbal communication,regardless of its logical type, can be dismissed as ’another case of talking’«516». It is in thisdisappearance of distinction between types, that the ’inner side’ of a one-sided boundary can berecognized.

Observer2 gets entangled in her own sentences within the session, but afterwards, beyond the session, sheis well capable of explaining what she had had in mind to say and how she felt impeded by her colleague,Observer1. This is interesting for three reasons. Firstly, for what she had wanted to tell the family and thetherapist; secondly, for the interaction between the observers; and in the third place because of thedifference between inside and outside the system, that becomes apparent.

The inner side of the system, materially existent as the ongoing session in the therapy room, can becharacterized by a lack of distinction between logical levels, in conjunction to a conviction, in thoseinvolved, that an escape from the system is still possible, e.g. by metacommunication! On the other hand,the outer side of the system, materially existent as the conversations with the therapist (after the session),and between observers(behindthe screen) can be characterized by a full availability of this distinction.Hence, we are dealing with a one-sided boundary.

There is something that Observer2 can say only beyond the therapy sessions: that some conflicts of thefamily do exist because of an absence of mutual acceptance; that she felt highly obstructed by Observer1

’s approving interruption. The position of the parents (’talk about the problems’) is represented here byObserver1 (’mutual acceptance’); and the boy’s position (’no talking’) is represented by Observer2 (’nomutual acceptance’). From an external point of view it might still be possible to maintain that the systemis characterized by a taboo, viz. that the problem cannot be solved by talking about it (or, for theobservers: the taboo of the non-acceptance).

These taboos, as well as the circumstance that real metacommunication, and accordingly a real escapefrom the system’s inside, are no longer possible, are precisely beyond the understanding of thoseinvolved. They are, so to say, quasi-objects for them, and only an external observer (also our twoobservers after their return from the therapy room) can see them.

But qua description, it is a descriptionfrom the outside, and as such it does not do justice to theabsenceof distinction between logical levels at the system’s inside, nor to the fact that this absence was notnoticed either by the family members or by the therapists, however well trained and non-naive the latterwere! Their attempts to oversee the system by talking about it did contribute to the system’s autonomy, asthey did not notice that their very talking was controversial, nor that there was also a conflict abouttalking; nor that, in this way, the way out of the system had become part of the system. This could berealized (by the professionals) only by analyzing the clash between the two observers.

In an unequivocal way the observing team’s role becomes apparent here, viz. to create a context thatsufficiently differs from the therapy session to allow a search for that which cannot be said within thetherapeutic system. That is to say: once a parallel process spontaneously occurs within the performance ofthe observation team, then it can be used as a ’laboratory’ for attempting to get out of it, in order that thetherapist may better succeed in getting out of the original (therapeutic) process.

The thing that we are looking for is not only a matter of articulating the things that couldnot be saidwithin the therapy sessions and that could be said beyond them (by the therapist or the

516. Such a situation reminds us the ’double bind’, as defined by Bateson, Jackson, Haley & Weakland (1956, pp. 206-7).

178

observers)«517», but is also based on the manifestation of this difference. The insights alone are notsufficient; what is needed, when outside the system, is that the therapist (and similarly the observers) isable to feel his own incapacity to think and understand when within the system,and to connect to thisincapacity, i.e. to endure it, explore it, practice it as a choice option. An example of this way of dealingwith the therapist’s own incapacity to think can be found in Joseph’s (1988, p. 146) description from apsychoanalysis:

"It seems that the way N was speaking was not asking me to try to understand the sexualdifficulties or unhappiness, but to invade me with despair, while at the same time unconsciouslytrying to force meto reassure myself that it was alright, that interpretations, now empty of meaningand hollow, were meaningful, and that the analysis at that moment was going ahead satisfactorily."(italics added)

Here too, messages at a metalevel (i.c. the therapist’s interpretations) had become idle, unnoticeably, anddid no longer offer a real retreat from the system. But by noticing that, this therapist became capable ofexploring it. In order to notice it, however, the therapist has first to accept the absence of clarity as afeature of the situation he is in, and to stop looking for a different point of view, as if the clarity had onlymoved and were to be found back in that point. In section 11.6 I will return to this issue, and discuss it interms of Merleau-Ponty’s (1968a) concept of ’Being’ as something truly negative, i.e. as a real absence,one that is not a presence ’elsewhere’.

10.1.2. The passage from outside to inside

The ideas presented in the previous section hold for each person who a) is at a position external to anautonomous system about which he maintains something that cannot be said from within the system, andwho b) attempts to participate in the autonomous system. If a researcher desires to attain the system’sinside«518», he will have to act such as to have something happen to him artificially, thatspontaneously also happened to the therapist and the two observers of my last example. In that examplethis would mean: to forget the difference between talking and doing, between fighting about the way offighting and fighting about something else. More generally it would mean: no longer distinguishing thesystem’s boundary. This is what a researcher can attain by having the process of his investigations (i.c.:the interview) coincide with the object of the investigations; that is: he has to make his investigations self-referential.

10.1.2.1. Case

The patient is a young man with a most negative self concept and most negative expectations with respectto other persons. The therapist who is being interviewed is an experienced psychoanalyst. Before theinterview he has clearly demonstrated his lack of confidence in psychotherapy research in general, as thisalways leaves the essences beyond consideration, so he believes. The interviewer does not have theprofessional background of the therapist; he is only an apprentice in psychotherapy, by that time. In theinterview that took place the interviewer became more and more desperately aware that he did not manageto become a participant and hence could not really understand what had happened in the therapy sessions.The therapist, from his part, brings the interviewer home in unequivocal terms that he is not equippedwell enough, for various reasons, to attain a more proper understanding. The interview goes so difficultly,and the interviewer is treated on bare facts of such a barren quality, that he feels most uncomfortable andmentions this at several moments in the interview.

(...)

517. This situation resembles one in which a patient refuses to talk about himself, and can only say things about other persons. It isthen of interest to the therapist to explore which projections the patient is making. Likewise, the opposition between the observers canbe understood as a kind of projection from the observed system.

518. I do not use the term ’observe’ here. An autonomous system’s inside does not allow being perceived as an object.

179

t: Well, what I could relate is how a young man time and again dashes against his anxiety that he isnot acceptable, not good enough, that it is due to himself that he will always be rejected byeveryone who is significant to him. Other people also can have this problem, of course, but for thisboy’s life it is pivotal. And in the previous hours it has also become apparent that he has not beenjudged passable by his father, as he feels it,that he was never good enough, for various reasons,but also that he had been very disappointing with respect to his mother’s expectations, that thismother had committed suicide because of that, when he was about eight years old. At least, that iswhat he thinks of it; that mother had had a car accident, she had crashed against a tree, and it hadnot been clear whether or not she had done so on purpose. But he takes it as a proof of how he hasfailed as a child and how disappointing he has been. It is all pretty complicated and he believes itstubbornly. Also when he is going a little better, this anxiety always pops up. Also in thetransference. Within soon, he knows, he will do something disappointing to me, so that I will say:now it is all over. And because he knows this, he is bound to always remember: "I must bevigilant, for this is only for a limited time; I must not have too much confidence, not disclose toomuch of myself. For then I will be dismissed. And if I would not really disclose myself, then Iwould not become really dependent, and perhaps I can manage my dismissal. But if I were to makemyself dependent, well, then of course he will dismiss me, and then I will be left holding the baby.

i: so the penalty for making oneself dependent is considerablet: yes. And I use to discuss with him that he believes that whenever he discloses himself, I will see

how bad he is and how wicked.i: that is what he is afraid of?t: yes. And that I will dismiss him for that reason. Well, then it is pretty complicated. I am working

on that, slowly. It is a laborious process.(...)

The interviewer puts his central question:i: What I am actually (...) aiming at is (...) to tryto find out where according to you the dynamics of

the process can be found. So, the relevant things that happen in this therapy, where would yousituate them and what do they look like? For instance, a particular session, like today’s one. Wouldyou please tell something about it?

t: I think it is the moment where silence falls, the external reality disappears, to talk about that. Andthat this boy feels, looks at me, and feels that he is having contact with me. And at the same timehe feels that I will also reject him for sure. Look, when he looks at me, and he doesn’t give meanything to talk about but his own feelings, then this pivotal anxiety comes to the surface, therejection. That’s what I see as the dynamics.

i: and you say that happens when the external reality disappears?t: the talking about the things that also bother him, more externally. When that has been told, and

there is a short silence, then he feels immediately the contact with me and he should say .... I doask for it as well, what is going on inside him. That is such a dynamic moment, the transferencegets a more prominent place, becomes closer to his consciousness. It is then that the feels like ’hewill surely abandon me’.

i: did you do something at the moment that the external reality disappears and transference isincreasing?

t: no, I don’t do anything special. I listen to him, I ask what he is thinking of. I didn’t know whetherhe was still thinking of something external or not. Well yes, and then I try to discuss with him howthings are, why he thinks so and why he is so sure that I will abandon him. Do you get what Imean?

i: Yes, but I would only would like to hear some more coloration. I do understand things, but theyare just somewhat too abstract for me at the moment.

(...)t: uh uh. Yes, and I am afraid that you want too much. For I know much more about him, as he deals

with these feelings, from all those preceding hours.

At this place the interviewer’s problem to get an adequate insight is presented by the therapist in terms ofa lack of information. Notice also that what the interviewer cannot comprehend well enough, according tothe therapist, concerns the patient’s feelings, not the particular dynamics at the onset of transference.

180

Therefore, a certain switch is taking place in the interview. Attention is shifted from the dynamics of theonset of transference to what is the subject matter of the conversation at those moments: ’the patient’snegative self-experience’.

t: and of course we mused on this already many times(...)i: what, according to you, is hard to understand for an outsider about what you both are doing? I can

imagine that you mention something to a colleague, or he tells something to someone else in hisenvironment. Those are facts. It is also happening now, actually. You tell me something, and thelistener listens and says ’uhuh’, and still he doesn’t get it, he doesn’t get the taste of it. What doyou think, what could that be?

t: I think that it is difficult to imagine and accept how vigorous such an inner conviction is, of beingbad

i: that’s right, it is hard to imagine. I can catch it, hear it(...)

The therapist relates now that it is very threatening for the patient to realize that he is making progress inthe therapy. Each time he seems to realize it, he feels threatened, for then he immediately thinks that ’itcannot be true, of course, that I do amount to something’. Then that inner conviction comes on that thetherapist will surely abandon him because he (patient) will surely do something wrong.

i: You are also saying: it is so hard for an outsider to imagine this, someone to whom you tell such athing, me in this case, such a person does not get the gist of it, even if he can handle it cognitively.Is that what you are saying?

t: Yes, because it is a terribly difficult feeling to live with, it is hardly impossible to imagine thatpeople can have such a negative feeling about themselves and be so convinced that it is true.

(...)

i: I can imagine something of your skepticism, or critical tones, that you vented in our telephone call,and you said: what should a researcher do with all this stuff

t: yeahi: now it is me who is getting the serious feeling: what should I do with all this stufft: yes, I don’t know either(...)i: what if I would like to try right now what is happening there [in the therapy session], do I bounce

against a wall, like that’s just the way it is, such a different world that I will never get at it?t: I wonder how you will do it, and uh, yeah....

(...)

i: in any case it has something to do, I think, with the things being told, the things you are tellingabout him, those are facts and as such they are necessary but, in a way, they are not enough. Theyare facts and, well, what is one supposed to do with those facts?

t: yesi: and I think that is what you are having in mind: what is a researcher supposed to do with those

facts; right?t: well, I had hoped, at least that’s what I made out of your reply, that you heard something different

from factsi: which reply?t: as I told you, I thought you tasted something that actually happens in such an hour. And you also

say that you are beyond, and that you cannot imagine that one can feel so bad about oneself; in factthat you show you realize ’such a thing does exist’

i: you are saying ’such a thing does exist’. What thing?t: that there is happening something different from facts.i: I see, but I can’t get at it. That’s what I feel

181

t: yes, for that’s impossible. You can watch two persons making music, let’s say. You can listen tothem and watch them, but you don’t belong to them, you don’t join their performance«519».

i: when I hear it, can’t I join them in enjoying it like they enjoy it, you mean that?t: No, your position is very different. You are not part of it. and if you would have had the same

profession and in a particular way, as would have a supervisor, so to say, then that person wouldbe very well able to recognize it, because he knows those experiences so well. It would makethings much easier.

i: I am not such a person, for I am far less experienced, let alone that I would have supervisorqualities.

t: yes, and it is not such much .... he does not need to be a supervisor, he can also be a colleague, butone who has some work experience with these problems; such a person immediately intuits that ithappens. He is less of an outsider

(...)

The therapist maintains that, had the interviewer been less interested in problems that concern his ownexternal position, he would have been ...

t: (...) very content with facts. In that case you would have said ’now I understand’, and that’s it.i: Is there a way to get me closer to it, or would that really require a colleague having those

professional experiences; or perhaps personal experiences?t: h’m. Well, that is than perhaps what you called my skepticism. I think it is tremendously difficult,

if you are not in this craft in this way, to intuit it or to really catch what is happening and what itis about. But I am careful, for perhaps it can be done in a sense, but I tell you: I am curious to seehow one should do it.

The interviewer feels challenged: "try it, you won’t succeed"

i: you don’t have the experience either that people can get at it?t: all of us we have to be trained for years and years, before one develops something like a

fingerspitzengefühlfor it. Therefore, how should one, without such a training background and witha different purpose ...., you see?

i: how should one just quickly imitate such fingerspitzengefühl?t: yes, it cannot be imitated, or it would not be a feelingi: exactly. No, I don’t have that. (...) But still at times I catch something of a situation in which I do

not participate myself. So I do not want to bear up entirely.t: noi: sometimes I can imagine how it is to be a virtuoso, irrespective whether it is with psychoanalysis

or with violint: yes, that’s why I am careful and I am saying: I don’t know.That’s why I am curious, you see ...i: how I will go aboutt: yes, how you will manage

the interviewer is getting annoyed

i: ....I won’t manage as a kind of climbing the Mont Blanc. I can’t do it as a sort of individualachievement. I am really bound to ask ’where do you see opportunities, for me, in this case, givenall my flaws, and yours. Where do you see something?’ and I think that sometimes it may be usefulto look at the way in which I misunderstand things, in your opinion. The loss of depth, when Imake a factuality out of a lived experience, saying things, like a mainstream researcher, of the

519. It is interesting to notice the difference between this conception of attending a musical performance, and that described byGadamer (see my quotation from Gadamer’s TM p. 98, above on page 169, related to footnote 494), who considers the spectator to bemuch more part of the performance.

182

variety ’oh yes, I understand, those are facts’. And what it is that makes the difference betweenthose facts and the experiences. Then I think it is you who is the expert with respect to thatdifference, to indicate it. (...) You have told me this and that about this man ...

(...)

i: and now that I summarize those things, you can say: ’yes, but that is not really the point’t: h’m. You mean that’s what is going on now? It is just what we have been talking about. It is the

factual part of the story, and then there is also the experiential part. And you are saying, rightlyand understandably, that that’s why you can’t fully get at; and then something is eluding you. Well,what could I say more about it, I don’t know.

(...)

Now the interviewer presents the projection metaphor:

i: You are watching something from above, for example a house, and you are saying: it is flat; butfrom a side view one says: oh no, it isn’t flat at all. So in a way, when I recount the story, then Imake it flat, I take a dimension out of it. So I am aware of that, and I think that many researchersare eager to ignore that difference and not to investigate it. And I think that that’s what ishappening very often in the field, when reports have to be written. As soon as you try to givewords to something, you distort it. Do you recognize something of it?

t: uh, yes, but that’s how it has to be; and so what? One does not need to reproduce everything, andwhy should researchers have exactly the same experiences as the topic of their investigations? Towhat purpose? perhaps it is impossible.

(...)

i: (....) What I want to know is: can I catch the difference between an upright house and a projectedhouse, in this particular case? that is what I am wondering about. I think I do have a few clues: -that something cannot be said properly; - that it has to do with my own lack of expertise in aprofession; - also that it has something to do with my own youth having been different from his; -that I am not myself in the analytic session, doing and experiencing certain things. These are a fewthings that I heard you saying. It is comparable to musicians who do something that one cannotshare with them. However, I imagine that still there must be something, even if I cannot catch it orrecount it properly, that still I could notice something in myself with respect to how it is that I getstuck, how it is that I am trying. I am also trying right now in several ways and there is always aplace I am arriving at. The way I am doing the run up, and the way I fail, do they say somethingabout what is happening? And then it is you who is the expert, so I it is from you that I would liketo hear it.

Here the interviewer is almost fully straightforward with respect to the ’negative approach’.

i: my way of getting stuck, apart from what it says about me, does it also say anything about thethings we are talking about?

t: yes, I am just thinking of those images. A picture or a scale-model of a house, that’s not the houseitself. If you don’t accept that, because you prefer the real house, then in fact you’ve got to buildthe real house anew. (...) You want to cancel the difference; here is the scale-model, there is thehouse, away with the difference.

i: I am not saying ’away with the difference’, I want to know the difference.t: what it is, that difference? uh(...)i: I cannot know the difference, because I am not in that situation. You are my interpreter, in that

way.t: yes. The difference is just that you are not in that situation.

183

(...)

i: That which I am doing in a conversation with you and a couple of other experienced persons is tofind out whether in spite of those restrictions I can still find something that may act as anequivalent. (...) what I am looking for in you, being an experienced therapist, is: (...) can still I getaccess? Can I find access in a different way? (...) Do you believe that I could find an equivalent?

t: you are hinting at that parallelism, aren’t you?i: yes, which does not happen here spontaneously, because you don’t need itt: but it happens far from always, it is just a phenomenon that may or may not occuri: I am eager to look for it (...)t: for then you get a piece of it, at least(...)i: but that is not taking place right now in this conversation, to no extent, not to the slightest degree,

or would you think so?t: I don’t think so. But wouldn’t you be helped better, for your project, with people who are more

inclined to that parallelism?

(...)

The interviewer say that he believes something is escaping the conversation. The therapist doesn’t agreewith that. The therapist tries to convince of the impossibility to know the taste of milk without drinking it.This is something the interviewer should admit, the therapist says.

Now the interviewer attempts to make the interview self-referential. He does so by putting into questionthe distinction between ’house’ and ’scale-model of a house’, already mentioned in the interview. Thus,the interviewer is suggesting that a collapse of logical types might still be possible, in this case thedistinction between the therapeutic conversation (house) and the interview (scale-model). The effect issurprising:

t: it is impossible to make a house out of a scale-modeli: it certainly is

the therapist had expected a response other than consent:

t: I must have misunderstood, you knowi: I don’t believe in that distinction between the house and the scale-model.I think it all just depends

from how we are talking with each other, whether it turns out to be a house or a scale-model. Inour conversation it is a scale-model. The fact that we cannot but talk about a scale-model, that’sexactly where the house is eluding me. I don’t get a feeling of what is going on there in thetherapy, because the way we are talking about it prevents me to see it as different from a scale-model. That’s also what you are saying: "it is a different world". and we also agree: thisconversation [the interview] is not that one [the therapy]. I do think that sometimes it is possible toeliminate that very distinction between scale-model and house. For example, it happens in aparallel process, when such a thing occurs.

t: That’s right, then there is happening some part of it, some part of it is happening.i: now?t: theni: not now?t: I’m not sure. Right now I am suddenly thinking: if you have those feelings, uh, how did you put it,

that you don’t manage and that it is impossible, then it might be something of what is so difficult inthat therapy; that might be. Yeah, haha.

(...)

t: I am giving you the feeling ’it [the interview] is hopeless, I [interviewer] am running against awall and I cannot come closer’. Well, that same feeling is what I [therapist] are experiencing with

184

him; that it won’t happen, at least as he sees it, that he ever drops that inner conviction. That it isimpossible, you see, that’s something similar. Dropping that inner conviction that he is notacceptable, that is impossible for him. He won’t drop it. Hey, that’s fun! (...) I am giving you thatvery same feeling as he is giving me. That’s the parallelism.

i: (...) He cannot change, he will always run against a wall.t: yes, that is his inner conviction, and because of that I can be who I am, I can be patient and try to

give warmth, sympathy, concern. But that is all bounced off by that inner conviction. For he turnsit into: "yes, he may be nice to me, and it does have some value to me, but you will see, he willabandon me, for I am no good. So one of these days, when at a moment of inadvertence I anrevealing how I really am, then he [therapist] will say ’get out of here’".

i: and to what extent does that resemble what is happening now with us?t: that inner conviction, running against it. What you are doing with me, that’s why I can do with

him. To do my best in order to convince him that he is fine. For that’s what I am doing each day Isee him. By accepting him, appreciating him, etc. But I can waste my energy on it. But I justrefuse to believe that, so I just go on.

(...)

Now the therapist relates that in a therapy session, held a few weeks before the session to which theinterview pertained, something had happened that strongly resembled the part of the interview describedhere. The therapist, against all psychoanalytic rules, had expressed his own feelings of incapacity andimpotence. He had admitted that alone he could not achieve anything, if the patient did not feel capable ofdoing anything; and that the only remaining option for the patient was to commence, entirely the bottomup, some activity that was sufficiently simple, and hence feasible. The therapist had not imposed upon thepatient the high standards that had always been set by the parents.

t: (...) of course there are all kinds of rules that you obey, and still, at some moment you lash throughand you say ’well yes, but this or that’. It can happen. And I think that such moments did happen,and that they conveyed to him that I am just a normal human being, even one who is concernedabout him«520»

i: those are the moments of forgetting your role, is that correct?t: well, I don’t want to say that, it’s not my role. One can respond more or less strongly. And one can

respond in a way that is clearer, more emotional; that he experiences less transference thanotherwise. And that’s what conveys your involvement, your being affected. That is very importantfor such a person; that you don’t keep your distance, but instead that you show you can be moved

i: You also notice that it moves him in a way?t: Yes, I think that it has helped slowly to bring forth the changes I have been talking about.

In a debriefing conversation with this therapist, some time after the interview, the therapist remarks thatthe event discussed had not happened in the therapy session that had been our focus in the interview. Theevent was said to have caused a turn in the treatment’s course. The therapist says: "it is striking how suchan essential transference element from this treatment returned, in a sense, in your conversation with meabout this very patient, and I think it is fairly well possible that such things happen more often, when atherapist starts to relate about his work, so that the interviewer really gets more from what’s reallyhappening between both persons than when he is only making a report." Indeed, a less skeptical attitudethan before! According to the therapist, the subject matter of the ’crucial turn’, which had happened in thetherapy some weeks before, had become manifest here as a central theme by means of a parallel process.

520. cf.: "Whether he was biting me or I was biting him did not matter so much as the fact that for a little bit of time in the sessionwe had been in emotional contact instead of each of us living in separate worlds, and my irony minimized the fact that this contactbetween us was being lost" (Spillius, 1992, p. 72);

and: "Some degree of acting out by the analyst (...) is often inevitable in the early stages of becoming aware of what the patientis feeling (ibid., p. 63)

185

10.1.2.2. Elaboration

This example illustrates the interviewer’s difficulties to get ’inside’, and how they correspond to thetherapist’s maneuver of claiming a distinction between logical types. The therapist takes much effort toshow the interviewer that his goal is unattainable. The interview is not a therapy session, a scale-model isnot a house, listening to music is not making music, etc. Then the interviewer expresses how much hefeels his attempts to come closer are impeded by those logical types, which we may understand as amatter of naming the resistance. It also seems to work that way, for it helps the therapist to release thelogical types, for a moment, and to recognize a parallel process in the interview.

We may also understand this sequence in terms of an autonomous system that shows itself in the courseof the interview. In the system’s inside no distinction was made between logical levels, and theparticipants did not notice it. For the process of the conversation and its product (the subject matter of theconversation) coincided in expressions of the variety ’you cannot understand it’/’you cannot share theexperiences of the therapy’. Such expressions are more than mere messages concerning a factual situation:they are also constitutive of that situation, as they push the addressed person (the interviewer) into adistant position! They are performative language utterances. By articulating the interviewer’s incapacity tounderstand, the therapist defines the things spoken about to be such as to be beyond the interviewer’sunderstanding.

We see now, even though the therapist made verbal usage of the logical types, he in fact articulated aprocess in which logical types werenot distinguished: by continuously emphasizing the differencebetween ’this conversation here’ (interview) and ’that conversation there’ (therapy), the ’here’ obtained aspecial double role, unintendedly and unnoticedly. Not only was it the subject matter of the conversation,but also the process itself of the conversation in which this subject matter arose! The ’here’ became thelocus where the interviewer’s failure was mentioned by the interviewer, and both confirmed and provedby the therapist. Thus, unnoticedly the conversation got autonomous properties: the process and theoutcome of the conversation were no longer distinguished by the participants. Furthermore, it is to benoticed that the metaphors used by the therapist for underpinning his points (such as ’scale-model versushouse’, etc.) did lack this very double role.

Instead of speaking about ’naming the resistance’, we can more fully understand the interviewer’s acts interms of ’joining the system’, which enabled him to pass the system’s one-sided boundary. As has beensaid, the therapist created a distance to the therapy, which he sustained by saying that this distance was amatter of facts. By mentioning in the interview itself his own feelings of failure and inadequacy, theinterviewer demonstrated his acceptance of this fact and he confirmed this fact. Thus, the interviewersupported the conversation as a collapse of logical types, by no longer distinguishing the conversation’sprocess from its contents.

Thus, the interviewer’s distance to the therapy, as well as the interviewer’s concomitant feelings offailure, not only became an issue of discussion, but also a process feature of the interview, and hence anevent of immediate lived experience. It is this lived experience that, subsequently, the therapist is capableto recognize in terms of an experience he had himself (in a session some weeks before the one discussedin this interview)

In view of this non-distinction between process and contents (outcome) of the conversation, theinterviewer now made an essential step. He asked for the resemblances or parallels between the therapyand the interview. In this way he asked for the chances of identifying therapy and interview. Notice thatthis question was put only after the theme ’distance between therapy and interview’ had turned into alived experience, in which no longer a distinction was made between the process and the outcome of ’thisconversation here’.

Thus, the interview became the place where aninterfacewas created between the interview itself and thetherapy. By not distinguishing the interview’s process and contents, the interview became susceptible to’infections’ from its the subject matter, i.c. the therapy. Thus, properties of the subject matter of theconversation could become properties of the interview itself. As a result, the interviewer arrived in a

186

position of being locked into the therapeutic system, in a way similar to how Levenson (1972, p. 174)describes the analyst’s "ability to be trapped, immersed, and participating in the system and then to workhis way out".«521»

Furthermore, the occurrence of this ’infection’ could be recognized in terms of a parallel process, bylooking for similarities and parallels between the therapy and the interview. Accordingly, the therapist’sdiscovery of a resemblance between therapy and interview takes place as a formulation of the actualinterview situation. The therapist says: "I am giving you the feeling ’it [the interview] is hopeless, I[interviewer] am running against a wall and I cannot come closer’." The therapist recognizes the therapyin the interview, even in terms of his own conduct with respect to the interviewer and his responses. Onlythen it is discovered that the interviewer had already passed the system’s boundary for some time, andthat a parallel process was already going on. It is also only then that the smooth emergence of the parallelprocess has unnoticedly facilitated the passage from outside to inside.

The interviewer, by putting into focus the interview’s failure and his own incapacities, and by(nevertheless) seeking for a comparison between the interview and the therapy, helps the interview obtainself-referential qualities that eventually enable the therapist to find a resemblance to self-referentialmoments in the therapy: moments at which the therapist experienced his discontent and presented it in away comparable to how the interviewer feels now.

10.2. The parallel process and the unclarity of therapeutic performance

The psychoanalyst Devereux describes how in his field studies as an ethnographer he used his ownpersonal responses for the interpretation of data and events. His own defensive responses, anxieties andother behaviors and feelings during and after the field work were interpreted by him according topsychoanalytic principles, with the purpose of attaining a better insight in the phenomena studied.According to him, a social researcher’s position with respect to his object of investigation is comparableto a therapist’s position with respect to his patient:

"He allows a disturbance to be created within himself and then studies this disturbance even morecarefully than he studies the patient’s utterances. He understands his patientpsychoanalyticallyonlyinsofar as he understands the disturbances his patient sets up within him. He says: ’And this Iperceive’ only in respect to these reverberations ’at himself’ [...].

This finding is epistemologically crucial. The disturbance occurs ’within’ the observer and itis this disturbance which is then experienced as the real stimulus and treated as the relevantdatum."«522»(Devereux, 1967, pp. 301/2)

A researcher does not need to put aside his own emotional and other personal responses, as sources ofirrelevant distortions within a research process, but, instead, he may exploit them as features of the actualresearch process, and, hence, as possible sources of information on the research project’s subject matter. Ifhe does so, then even the imperfections of the research process can be used as extra sources of usefulinformation, and not, as usually, as sources of annoying distortions. This will be demonstrated in the casematerial to be presented below.

We will encounter a therapeutic situation where the therapist’s capacity to play and explore is lessened byhis own countertransferential feelings, generated by the pain he perceives in his patient, a pain which hecannotdistinguish from his own feelings. This situation, which is in fact one of projective identification(see chapter 11), affects the therapist’s behavior in the interview session. It is within this latter context

521. As quoted by Bromberg (1984, p. 40n). Bromberg (ibid.) comments: "I believe that this is really a special case of the capacityto be fully immersed without loss of the analytic field, and that the more highly developed the capacity in an analyst, the more fullyimmersed or "trapped" he can safely be for potentially longer periods of time, without the danger of being unable to "work his way out."

522. We recognize here the situation described in section 7.2, where an observer, looking for the first order distinctions of anindividual, ends up studying his own experiences (first order distinctions).

187

that the interviewer becomes confused, and a parallel process develops in the interview. The unraveling ofthis confusion leads to a better understanding of the dynamics of the therapeutic system studied.

First of all, in section 10.2.1 the central event from the therapy session discussed will be outlined. Next,in section 10.2.2 I will describe how the interviewer comes to make an incorrect assumption, that leads toa confusion in the interview. In section 10.2.3 a fragment from the interview is given, in which thisconfusion is presented. Section 10.2.4 describes how this confusion could be cleared up and how itprovided the desired insight in the dynamics of the therapeutic interaction.

10.2.1. The central event of the therapy session

The therapist has mentioned that the patient himself had proposed to terminate the sessions, but that hehad turned more doubtful when the therapist had not dissuaded this option. The episode from the therapysession with which the interview is mainly concerned, is presented by the therapist as follows:

-The therapist had remarked:"I notice, anyhow, that you are in doubt, so I think we had better talk about that in any case.Apparently it is not self-evident to you to stop, so there’s an hesitation somewhere. How does thatone come about? Shouldn’t we have a look at it?"

-Then the patient had replied:"these conversations have become very valuable to me".

-The therapist had remarked then:"perhaps it has something to do with our relationship, and perhaps it is painful for you to break offthe bond you are having with me".

-Upon that, the patient had rigorously objected against the therapist’s words.

For the therapist it had seemed as if the patient said something about the efforts it would take him tomention his feelings or attachment to the therapist (and hence, as if the patient mentioned a property oftheir actual interaction). It had become self-evident for the therapist that this patient was capable ofrecognizing in himself these efforts in the way the therapist himself was capable of recognizing them.(Afterwards the therapist admits that this self-evidence was not entirely justified: he could have predictedthat the patient would have opposed against his interpretation).

At several moments in the interview the therapist displays his conviction that his remark ("perhaps it hassomething to do with our relationship, and perhaps it is painful for you to break off the bond you arehaving with me") lacked a proper timing, came too quickly, because he had underestimated howfrightening it was for this patient to face this ’painfulness to discuss the relation’.

i: what is your part in that, you see, for that is what interests me: is there something in you thatmakes you act too quickly at such a moment?

t: I think so. I do think that I am underestimating how frightening this is. It is so clear for me,because of several indications,that he is going to say that, or that it plays a part; no, not that he isgoing to say it, but that it plays a part; but he has not yet made that step himself, I think.

i: and then you are underestimating how frightening it is for himt: yes, I suppose so

The therapist believes that the patient is going to say that which the therapist himself is perceiving to takeplace in their interaction. this turns out not to be the case. The therapist had understood the patient’swords ("these conversations have become very valuable to me") as pertaining to the interaction (inparticular on the painfulness for the patient to speak about the relationship). From the patient’s response,however, it becomes clear that he hadnot had in mind how painful it would be for him discuss therelationship. Therefore, we may call the ’painfulness to discuss the relationship’ aquasi-object for the

188

patient: he does not perceive it himself as an object.

Possibly it is much too painful for the patient to develop any opinion about this (assumed) painfulness forhimself to discuss the relationship; or possibly there is something else going on. In any case, the patienthimself is only oriented to the contents of his own words, not to the constructs the therapist making inaddition to them. The therapist is assuming that the patient can perceive something he can perceivehimself. For the patient it is: how valuable the conversations are for him; for the therapist it is: thepainfulness at that moment for the patient to discuss the relationship, as a feature of the interaction at thatmoment. The therapist is a critical observer of the patient, who himself is constructing a naive object(indicated a.o. by the statement "these conversations have become very valuable to me").

10.2.2. The presumed perceptibility of the therapist’s own countertransference

The first thing that interviewer noticed when listening to the therapist’s account of the event mentioned insection 10.2.1 was the therapist’s own emotional involvement. For instance, initially the therapist couldnot remember the patient’s statement (on the valuable conversations). The interviewer then proceeds byasking about the emotional impact of those statements for the therapist himself.

t: well ...., no I remember, yes, what happened, yes perhaps that I ... that’s how I feel it myself..., Ithink I find it myself so ... painful, that this man is having problems with it [viz. discussing therelationship, i.e. what the contacts with the therapist mean to him] and that I am trying all the timeto relieve it; and as a result I am continuously engaged in highlighting and not highlighting it; doyou imagine what I mean?

i: highlighting and not highlighting?t: Well, uh, as I feel it, it is so awfully vulnerable for this man«523», this subject matter [i.e.

the relationship], that I am really hesitant to touch it; he is also conveying something of the type"don’t excavate that"

i: and you don’t comply with that?t: Well, I don’t comply, but in a sense I do, I think; I see that it is so painful that I do give him this

little step [viz.: "perhaps it has something to do with our relationship, and perhaps it is painful foryou to break off the bond you are having with me"] when he just made his single statement ["theseconversations have become very valuable to me"].

i: yest: And then he denies, and then I moderate my idea or my hypothesis, like: well, perhaps I am still

wrong. And I also see how he is frightened as he utters his statementi: you feel the tensiont: yeah, I feel the tensioni: and you don’t want him to have itt: it is so awfully painful; I just can see that it is so painfuli: painful also for you, to see itt: yes, yes, awfully vulnerablei: and in fact, you know that your blankness.....

In this expression by the therapist the painfulness for the patient is put in the first place. This is an

523. Notice the therapist’s usage of the article ’it’ and the adjective ’vulnerable’ (Dutch: "het is zo ontzettend kwetsbaar voor dezeman") indicates an archaic ’middle voice’, as in classical Greek, situated between the active and the passive! Had the therapist said ’it isso painful for this man’, or an equivalent, then it would have been clear that the adjective (painful) pertains to the man. Now, on theother hand, the therapist refers to some impersonal ’it’ that has the property’vulnerable’. This ’it’, I think, is felt to be an object sharedwith the patient. It must be left to another occasion to investigate a relation between this middle voice and the common sense qualities(cf. Vico, section 3.3) of naive objects!

189

opportunity for the interviewer to ask about possible countertransferences«524» in the therapist. Butthe latter is quicker, so that the interviewer does not come to articulate his question aboutcountertransference:

t: Something is reverberatingi: what do you mean?t: well, some of this painful stuff also comes to me, and as a result I am loweringi: what are you lowering?t: well, not pressing too much, more ’let it be’i: uh, you accept his disapprovalt: yes, ’let it be’(...)i: you are saying: it is such a bad thing to look att: very vulnerable, yesi: so painful for him, as well as for you to see that perhaps .... you are making an assessment error as

to the degree to which he may cope with it, and that you don’t want to let this plod on (...)t: yeah, I have made my remark because of his statement, and most probably I have been much too

quick with it, I presumei: yes, apparently, for he wards it offt: so I don’t give up, even if I can see that it is painful, but in a way he cannot handlei: yes but you cannot know that in advancet: yes, but by now, afterwards, I am wondering about that, if that is truei: whether you could have predicted it?t: yes, whether I could not have much ...., well, that’s always hard to say in those matters, but I think

I could have predicted that he were not to respondi: in that way one might wonder: to which extent has it been for you ...(...)t: exactlyi: a need to say it

The therapist concluded that the painfulness of the issue ’relationship’ did not allow the patient to discussit. For the interviewer, now, the conversation seems to pertain to soreness in the therapist, whereas for thetherapist it looks like soreness in the patient, that he is willing to handle respectfully. Still, at thismoment, the soreness in the therapist seems to be something recognized and formulated by himself. (Italso seems as if the therapist himself, qua critical observer, can see that the painfulness for himself andthe assumed painfulness for the patient are mixing up, as well as that the therapist is losing the distinctionbetween his own pain and his patient’s.)

Therefore the interviewer is presuming - erroneously, as will appear later on - that the therapist is capableto observe his own feelings of countertransference with respect to the patient, in the same way as theinterviewer himself observes them. The interviewer is assuming that the therapist is also aware of howpainful it is for himself to see how painful it is for the patient to discuss their relationship. That thisassumption is not warranted becomes clear only after the interview. We meet here a quasi-object for thetherapist. In the interview itself this is not recognized, and confusion pops up.

524. I am using the term ’countertransference’ here after Paula Heimann’s (1950) interactional understanding: "... to cover all thefeelings which the analyst experiences towards his patient." (p. 81) Furthermore, the psychoanalyst is to "...sustainthe feelings whichare stirred in him, as opposed to discharging them (...), in order tosubordinatethem to the analytic task in which he functions as thepatient’s mirror reflection." (p. 82)

190

10.2.3. Confusion in the interviewer during the interview

This confusion looks as follows: the interviewer assumes that the therapist is capable of observing hisown feelings of countertransference in the way the interviewer observes them, namely that the therapistcan see that the painfulness for himself and the presumed painfulness for the patient mix up, and that thetherapist loses the distinction between his own pain and his patient’s. We may recognize this loss ofdistinction as evidence of projective identification (see chapter 11). The interviewer now states ahypothesis concerning these feelings of countertransference, that he wants to communicate to the therapist.In the course of discussing this hypothesis, however, many things go wrong. Initially, in the process ofelaborating this interview, it looked like a hardly elegant episode in a conversation which waspredominantly interesting. On second thoughts, however, it became evidentthat the confusion that hadoccurred in the interviewer was very useful as a source of information concerning the subject matter ofthe conversation(the therapeutic interaction).

The following goes wrong. The interviewer comes to use terms in an ambiguous way. The interviewer’shypothesis is about a certain ’preference for blankness’ in the therapist. But he sometimes means by theterm ’blankness’ an absence of articulation of emotions by the therapist, and sometimes a certainuntouchableness with the effect that the patient does not express emotions towards him. The hypothesisformulated is ambiguous, and it can mean:

- a preference in the therapist not to be faced with strong emotions of the patient towards him;but also:- a preference in the therapist not to express emotions towards the patient

The first meaning is the original one. The following fragments show how the second meaning is floatingthrough the interview, which leads to an unnoticed confusion of the two meanings:

(...)i: (...) you, personally, it feels good to you if it does not become too close, if you can remain blank.

My fantasy: it feels good to you, it fits you, and it is a pleasurable role for you, and you canoperate well with is, you can do things well with it [thus far the therapist is humming withconsent]. There, it colors your relationship, this preference of yours, in that your relationship is onein which you, uh, don’t get confronted with strong emotions. He won’t develop strong emotionswith respect to you, because you are too smooth, aren’t you? [uh uh], too smooth in terms ofRogers, in the way of reflecting them, nothing sticks to the mirror. And that’s where some of youremotions have gone to. This was the idea I got at a certain moment.

[...]you also like it, it means something for your relationship, namely that you don’t get confrontedwith certain strong emotions in this contact

At this place the interviewer’s formulation of his hypothesis is still rather flawless. A little further on hewill start making errors; he will then suggest that it would feel good to the therapist not to expressemotions towards the patient, whereas, now, it is still his preference not to be confronted with thepatient’s emotions. Pivotal for the confusion is the concept ’blankness’, that obtains a variety of meanings.

t: uh, but that is not what I am after, because I think those emotions are there all the time, butbecause they are so unidirectional, from his side, it is so awfully threatening. So I try to bring theminto the open by saying: "I have the idea that perhaps you have some problems with terminating thesessions, that possibly there something else bothering you", and then he rejects it

i: yes, clearlyt: but what you are suggesting is that due to my way of being blank he is reinforced also to be blank?i: that, to some degree, it could help you prevent him developing strong feelings, with or without

expressing them to you; authority conflicts, to name somet: yes, it could, but only if I would systematically deny and avoid them

191

i: which you don’t?t: no, as far as I can feel [sounds slightly irritable](...)i: and I am saying: look, that reflecting job of yours, that is up to your taste, you are good at it; and

to what extent does it prevent strong feelings to be expressed? Your preference for it, not the jobitself of being reflective, but your personal preference for it, that ’s what I mean

t: personal preference for....?i: smootht: being smooth, being blank, not clearly expressing my feelings for the other?i: not that you don’t express your feelings, but that not showing them feels good to you.

Here the interviewer gets confused: the original aim was to find out whether the therapist’s behaviors hadthe effect of discouraging the patient to develop strong emotions for the therapist; now, however, theinterviewer is suggesting that it feels good to the therapist not to express his own emotions. That is adifferent matter. The therapist replies defensively: there is no question that the therapist expresses hisfeelings to a patient. Here the therapist is absolutely right. The interviewer finds himself bound torecognize that.

10.2.4. The interview’s confusion unraveled: the second interview

Now this confusion during the interview can be used as a source of information about the original subjectmatter, the therapeutic process. Our focus is no longer what the therapist said in the interview about thetherapy, but to understand why the interviewer got confused and what that reveals about the therapeuticprocess.

A first access to the confusion in the interviewer was already discussed in the interview itself, after thefragments presented here. According to the therapist, the interviewer might have got an idea similar to thepatient’s idea«525»: he might erroneously believe the therapist to have a secret preference forwarding off and discouraging the patient’s strong emotions for him (therapist); the patient too might havegot such an incorrect idea. The interviewer had looked through the patient’s eyes, so to say. Accordingly,interviewer and patient would be mistaken in a similar way. In brief, a parallel process with theinterviewer in the patient’s place. This interpretation does not yet cover the ambiguity of the ’preferencefor blankness’ hypothesis, and for that reason it is not yet satisfactory. In the subsequent episodes of theinterview, not given here, this idea is being considered by the interviewer and the therapist.

It was only during the elaboration of the tape recordings of the interview that the ambiguity of the term’blank’ was noticed, as described in the previous section. One is free to consider this as a kind of slovenlywork from the part of the interviewer, but it is more interesting to see whether this disturbance of theresearch process may illuminate the intended object of research. For that purpose a second interview hasbeen arranged, in which a elaborated transcript of the first interview was gone through together. Theinterviewer’s most important question for the therapist was to find an interpretation for the ambiguitymentioned.

The following has jointly been found in this second interview:

a) the painfulness for the patient to discuss the relationshipandb) the painfulness for the therapist to face the previous item a)

mix up for the therapist in the therapy session discussed (cf. section 10.2.2). Furthermore, the therapist’sway of decreasing this painfulness for himself is also the way he believes to be at his patient’s service:

525. This is the type of identification that Racker (1957, p. 311) called ’complementary’, in opposition to ’concordant identification’,where the therapist (or interviewer) would take the position of the patient (or therapist).

192

"refute the painful by bringing it up".

Thus, the therapist identifies with the patient; he does not distinguish painfulness for the patient formpainfulness for himself,nor does the interviewer! What matters is not whether the therapist had apreference, in the therapy session discussed, for not expressing emotions himself, or for not enduring thepatient’s strong emotions. What matters is that the therapist could not distinguish between these two. Thiswas a part of his countertransference that he could not perceive himself, and that did not occur to theinterviewer either.

The same problem (’is the painfulness in the patient or in the therapist’) was also present in the interview,with respect to the interviewer’s hypothesis. The discussion of the therapist’s countertransference entailedconfusion in the interviewer and an ambiguous hypothesis. In this way the confusion that popped up in theinterviewer was a disturbance of the interview, but also one that conveyed information about the subjectmatter of the research. Moreover, by investigating the confusion that had slipped into this hypothesis, astarting point was found to clarify the identification that had happened in the therapy. This does notdemonstrate, of course, that the ambiguity of the interviewer’s hypothesis was caused by properties of thetherapy session discussed, but we do know that the analysis of this ambiguity de facto led to the discoveryof the identification that had happened in the therapy session.

Thus, we find that the interviewer’s concern for the thinking errors he made during the interview, washelpful for the recognition of features of the therapeutic process that remained unclear before. Thetherapist was capable to make the essential connection between the confusion that occurred to theinterviewer on the one hand, and on the other hand the absence, in the therapist, of a distinction betweenthe two items a) and b) mentioned above.

It is here that we find the explorative qualities of the therapeutic process studied. Exploration is in thetherapist’s capacity tocontain«526»the patient’s sense of pain or fear about the impendingtermination of the treatment; and it is only to the extent that the therapist doesnot provide thiscontainment (cf. his technical mistake of too early giving his interpretation), that exploration in thistherapy appears as suboptimal.

We found in our case example of this section 10.2 that the researcher became involved in the therapist’stask of ’containing’ the patient’s perceived pain. Furthermore, in section 10.1.2 we noticed that theresearcher was introduced, during the interview session, to the therapist’s burden of dealing with thepatient’s negative self-definitions. The same happened to the observers of the family therapy session, asdescribed in section 10.1.1. Bromberg (1982) regards ’projective identification’ as the explanatorymechanism behind the parallel process.

We find that the process of observation of a therapeutic process may involve the observer in a way that isuseful to, or in support of, or even a substitute for the occurrence of exploration processes in the therapist.That is to say, the therapist may use some external person, such as an observer, in order toborrow fromhim a point of view, a perspective from which his (therapist’s) job of exploring the therapeutic processescan be performed more easily!

We arrive here, in fact, at a formulation quite close to that of Bion’s (1962a) theory of thinking. As weshall see in the next chapter, according to Bion projective identification was to be considered a mode ofcommunication alternative to thinking: when thinking is impaired for some reason, it is sometimespossible to have another person (’the container’) do the thinking; this, in fact, is not too far away fromborrowing the supervisor/observer’s point of view and have him do the exploring! Thus, when theobserver becomes involved in the therapist’s own job of exploring, when the observer/interviewer activelyseeks this role and also gets it, then part of the therapist’s thinking is taken over by the observer. This isthe more plausible to the extent the therapist’s own thinking and exploring activities were already chargedwith the task of digesting those materials the patient is imposing upon him.

526. In section 11.4 I will discuss Bion’s usage of the term ’containment’.

193

It is getting time to turn to the final chapter, and face the phenomenon of projective identification. Wehave found in this chapter that a therapist, an observer, an investigator, all can get involved in the closureof an autonomous therapeutic system, be it a family or an individual. When that happens, there is adecrease of one’s capacities to think critically. This may be a position one wishes to escape from (as maythe therapist who finds himself to be locked in); or it may be a position one seeks in order to study it (asmay the investigator or supervisor). The lock seems to consist of a loss of one’s capacity to distinguishlogical types, so that a fully self-referential situation can occur, in which speaking or thinking about thesystem does not bring one into a position external to it. It is this loss of distinguishing logical types in thetherapist and in the observer, that is a major effect of the occurrence of projective identification. Let usturn to this now.

11Some notes on projective identificationThe present chapter attempts to elucidate the concept of projective identification by relating it to some ofthe concepts I developed above in this book, especially the notions of incorporation and one-sidedboundary. I will give some special attention to the status of projective identification: is it an unconsciousphantasy, as Melanie Klein maintained, or is it an event in a ’different’, transcendental reality, such asJung’s collective unconscious?

I will present the event of projective identification from the perspective of ’not understanding’ (section11.1). This is a basic feature of the negative way, and it is the experience of not understanding, no longerunderstanding or getting confused, that is an often recurring event both in psychotherapeutic sessions andin research attempts to clarify them. I will then present in section 11.2 the regular conception ofprojective identification, and in section 11.3 I will make some comments on one of this conception’s basicconstituents, i.e. the idea of ’unconscious phantasy’. They will turn out to be instances of what I calledquasi-objects. In section 11.4 I will discuss Bion’s model of ’container’ and ’contained’, an extension ofKlein’s concept of projective identification, and I will focus on two of its major implications, oneconcerning the immediacy of the contained material, as if no boundary existed between the two personsinvolved (mother-infant, therapist-patient), the other concerning the ontological status of projectiveidentification. I will pay attention to these two implications, and discuss in the two following sections ofthis chapter (11.5 and 11.6) their relation to my earlier discussed concepts of ’incorporation’ (section 4.2)and ’one-sided boundary’ (section 7.3) respectively. Finally, in section 11.7 I will make some remarks onthe idea that projective identification can be understood as the usage of a model coming to a limit.

11.1. Not thinking as an empirical phenomenon

In chapter 10 I gave several examples of the problems of observing and understanding psychotherapeuticevents, as encountered from the point of view of an external observer, an investigator (interviewer) or asupervisor. In this chapter I will pay attention to a property of psychotherapeutic processes that, I think, isin particular responsible for the difficulties researchers have encountered in empirically disclosing thefacts of therapeutic encounters, in particular those ’facts’ that were considered to be effective quatreatment interventions. For if not the patient, then at least his medical insurance is interested in a bestresult at lowest cost. It has therefore become a major research effort to make psychotherapy moreeffective and accordingly to look for the effective constituent factors. This, in fact, has favored the rise ofan experiment oriented therapy culture, mostly of a behaviorist identity (cf. footnote 467, page 161), butnot necessarily (e.g. Miller, Luborsky, Barber, Docherty, 1994).

Unfortunately, the empirical evidence based orientation of scientists has not lead to a better understandingof those phenomena that, due to the experimental method itself, came beyond the scope of theinvestigations«527». This, however, is not to say that these phenomena could not be empiricallyobserved at all, set aside that thinking of them would be a kind of metaphysical speculation.

Can we devise a scientific instrument that studies them, even when applied by scientists who do not liketheir ’good old’ research methods to have blind spots? I think the answer is positive, and should first ofall be sought in the legacy of William James’ idea of ’radical empiricism’ (cf. section 7.4.2). That is tosay, "any kind of relation experienced must be accounted as ’real’ as any thing else in the system". Toapply this to our subject matter, we need to take our experiences seriously, even if we cannot easily find

527. In analogy to the ’virtual focus’ of a concave lens, a property that can be defined only in terms of what the lens doesnot do,we may call these phenomena virtual artifacts, as they were ignored due to properties of the method itself, in particular the property ofcreating a detached position from which to construct critical objects!

194

195

the objective events to which we believe they should correspond! The type of relation that is of particularinterest to us here, is the transition from a situation that can more easily be expressed in words, to onethat is less easy to verbalize. It is precisely this transition that can be experienced, just as thedisappearance of a black dot can be experienced, in our kitchen table’s attempts to become aware of ourretina’s blind spot«528»! These transitions toward a condition of no longer perceiving,understanding, controlling and even thinking should be considered to really take place, because they canbe experienced as events that happen to the observer (therapist).

Projective identification is a concept that is in particular related to this transition, and among its numerousdescriptions there are many experience based accounts by psychotherapists, predominantly psychoanalysts,who describe how they noticed a decrease of their capacity to think.

E.g.:"... she launched into a description of a lot of inconveniences and minor grievances, mainly atwork. She said it was a peculiar twist that analysis stirs things up in such a way that all theseminor complaints turn into an attack here.

I said she wanted me to regard them as minor complaints, but ...No, she said, only to know the difference between major and minor. (There was utter

contempt in her voice.) There was a short silence, then she said, ’I don’t know whether it’s you orme, but in the past ten days it seems to me you just totally and utterly keep missing the point."(Her tone was exceedingly scathing.) ’Yesterday you apparently didn’t notice it was so painful forme to admit that I find analysis and everything you do so terribly uninteresting. I can’t stand it.’

I waited a bit, then started to speak, but she broke in, ’Don’t talk! [almost screaming]You’re just going to repeat what I said, or you’re going to alter it. You don’t take things in, youdon’t listen to what I say, or you listen and you just want to hear it the way you want it to be andyou distort it.’ (There could hardly be a better description of what she did to my interpretations, butinterpreting projection directly is not usually helpful, especially when a patient is in a flight ofparanoia.)

(I was finding it hard to think, and I knew that my own self-doubt and feeling that I was abad analyst were getting powerfully stirred up by her accusations. But I managed one smallthought, which was that she must be feeling inadequate too ... (...) It felt to me as if I was like adamaged animal making her feel guilty, and she wanted to stamp me out.)" (Spillius, 1992, p. 67)

And Betty Joseph writes:"Frequently, when I interpreted more firmly, he would respond very quickly and argumentatively,as if there were a minor explosion which seemed destined, not only to expel from his mind what Imight be going to say, but enter my mind and break up my thinking at that moment." (Joseph,1988, p. 142)

Another example, also from Joseph (1988) has already been quoted above (section 10.1.1.2, p. 178).Notice that it usually takes considerable courage to make such confessions in writing, as they are tooeasily interpreted as signs of incompetence. And, of course, the reverse is not true: not thinking, notunderstanding etc. is by no means a quality warrant for an aptly handled projectiveidentification«529»!

Is projective identification, courageously described or not, an important concept? I think it is, as itpertains to a major blind spot of regular research methods. In fact, projective identification can beunderstood as amode of functioning(by the infant or the patient)that prevents the occurrence of the self-

528. Even though the dot’s absence itself from our perceptual field cannot be noticed. Cf. also footnote 327, page 118.

529. cf.: "I have often noticed that those whose interest is primarily clinical are often the least logical, in that they seem tocontradict themselves a lot. This is because they leave many possibilities open, but it would be ridiculous to suppose that if all goodclinicians are illogical then all illogical analysts are good clinicians. Yet this seems to be the unconscious assumption of many ananalyst, for, in their enthusiasm for the ’proper analytic attitude’, they seem to have developed a phobia of reason and logic." (Skelton,1995, p. 390)

196

referential exploration(in the mother or the therapist) that was typical of naive perception (cf. chapter 9);or, alternatively, as amode of functioning that hinders or obstructs such an already establishedexploration process. Accordingly, it is of interest to explore the very process of losing sight, decreasingunderstanding, as it occurs in therapists, and as it also may occur in those involved in empiricalinvestigations. In the previous chapter I have presented some of my own field work experiences withthese phenomena.

11.2. The idea of ’projective identification’

11.2.1. Klein’s original concept

After Melanie Klein introduced in her seminal 1946 paper the concept of ’projective identification’, therehas been an increasing degree of recognition of this concept, both in contemporary psychoanalysis and inother directions of psychotherapy«530». This has led to much confusion as well«531».Eminent introductions on projective identification can be found a.o. in Hinshelwood (1991), Grotstein(1981, 1994a,b, 1998b), Anderson (1992), Spillius (1988a,b), Kulish (1986), Ogden (1982, 1989), Segal(1975) and the Symingtons (1986).

The original concept was developed within a context of psychoanalytic work with very young children.During her work since the twenties of this century Klein elaborated her own ideas on the early emotionaldevelopment of children, to the extent of departing from and disagreeing with Freud.

In the words of Bion:"Melanie Klein has described an aspect of projective identification concerned with the modificationof infantile fears; the infant projects a part of its psyche, namely its bad feelings, into a goodbreast. Thence in due course they are removed and re-introjected. During their sojourn in the goodbreast they are felt to have been modified in such a way that the object that is re-introjected hasbecome tolerable to the infant’s psyche." (1962b, p. 90)

Apparently this concept pertains to a move of intolerable feelings from the infant’s psyche to themother’s, where they are ’digested’ (a term often used by Bion). Only then can they be returned to thechild. The concept of ’good breast’ is a prominent one in Klein’s thinking. It refers to the good mother(but equally to other ’good’ objects), as perceived and experienced by the child through its immaturecognitive and emotional apparatus.

Projective identification, in Klein’s own words, was a"... ’phantasy remote from consciousness’ that entails a belief in certain aspects of the self beinglocated elsewhere, with a consequent depletion and weakened sense of self and identity, to theextent of depersonalization; profound feelings of being lost or a sense of imprisonment may result."(Hinshelwood, 1991, p. 179)

Such a phantasy«532», therefore, was believed to be about the evacuation of parts of the self,especially the self of neonates and infants, but in its pathological variants also parts of the self ofpsychotic patients.

530. cf. Mahrer (1983) who speaks about a ’blending or sharing’ between therapist and patient (e.g. p. 247)

531. cf. Kulish, 1986

532. Notice that the term ’phantasy’ is differentiated by the Kleinians from ’fantasy’. As Apprey (1986, p. 112) puts it: "Kleinians(...) restrict "fantasy", as Freud did, to conscious fantasy as in fiction, daydreams, and the like. In contrast, they use "phantasy" whenreferring to dynamic unconscious phantasies, which are psychic representatives of instinctual drives."

Cf. also: "Phantasy in the Kleinian view is primitive, dynamic, and constantly active, coloring external reality and constantlyinterplaying with it." (Segal, 1981, p. 5, as quoted by Apprey, 1986, p. 112)

197

Afterwards the term ’projective identification’ was applied more and more to a variety of interpersonalinteractions, to the degree of losing its specific denotations«533». In general, the British stayedmore with the original definition:

"It is still generally accepted, at least by British Kleinian analysts, that projective identification is aphantasy not a concrete act, but it is now accepted that patients can behave in ways that get theanalyst to feel the feelings that the patient, for one reason or another, cannot contain within himselfor cannot express in any other way except by getting the analyst to have the experience too"(Spillius, 1992, p. 62)

11.2.2. The spectator’s engagement: a benign instance of projective identification

Klein’s interpretations of the unconscious processes in very young children have to a large extent beenaccomplished through their observation in play. It is of interest, in this respect, to find a connection toGadamer’s ideas on the observation of theatre play. Here, from an unexpected side, we may recognize adescription of the phenomenon of projective identification; in particular, it is the idea of identificationbetween actor and spectator that is of our interest, as it enables the latter to receive the focus of the play,as a ’Sinnganzes’ (meaningful whole):

"In general, games, however much they are in essence representations and however much theplayers represent themselves in them, are not represented for anyone, ie they are not aimed at anaudience. Children play for themselves, even when they represent. (...) A procession as part of areligious rite is more than a demonstration, since its real meaning is to embrace the whole religiouscommunity. And yet the religious act is a genuine representation for the community, and equally atheatrical drama is a playful act that, of its nature, calls for an audience. The representation of agod in a religious rite, the representation of a myth in a play, are play not only in the sense that theparticipating players are wholly absorbed in the representative play and find in it their heightenedself-representation, but also in that the players represent a meaningful whole for an audience. Thusit is not really the absence of a fourth wall that turned the play into a show. Rather, opennesstowards the spectator is part of the closedness of the play. The audience only completes what theplay as such is.(...)

Even a theatrical drama remains a game, ie it has the structure of a game, which is that of aclosed world. But the religious or profane drama, however much it represents a world that iswholly closed within itself, is as if open toward the side of the spectator, in whom it achieves itswhole significance. The players play their roles as in any game, and thus the play is represented,but the play itself is the whole, comprising players and spectators.(...)

[The players’] mode of participation in the game is no longer determined by the fact thatthey are completely absorbed in it, but by their playing their role in relation and regard to thewhole of the play, in which not they, but the audience is to become absorbed. When a play activitybecomes a play in the theatre a total switch takes place.It puts the spectator in the place of theplayer. He - and not the player - is the person for and in whom the play takes place. (...)Basicallythe difference between the player and the spectator is removed here.The requirement that the playitself be intended in its meaningfulness is the same for both." (Gadamer TM (1986a), pp. 97-99;italics added)«534»

533. Hinshelwood (1991, pp. 200-1), not entirely without a wink, so it seems, selects the following fragment from Spillius (1983,p. 321) to illustrate this: "... the concept is now used by non-Kleinians, and papers are even being written about it in the United States.In the course of such general popularity the concept has been widened and is sometimes used loosely."

534. "Im allgemeinen werden Spiele, so sehr sie ihrem Wesen nach Darstellungen sind und so sehr sich in ihnen die Spielendendarstellen, nicht für jemanden dargestellt, d.h. die Zuschauer sind nicht gemeint. Kinder spielen für sich, auch wenn sie darstellen. (...)Erst recht ist etwa die Prozession, die ja ein Teil einer Kulthandlung ist, mehr als eine Schaustellung, da sie ihrem eigenen Sinnen nachdie ganze Kultgemeinde umfaßt. Und doch ist der kultische Akt wirkliche Darstellung für die Gemeinde, und ebenso ist das Schauspielein Spielvorgang, der wesenhaft nach dem Zuschauer verlangt. Die Darstellung des Gottes im Kult, die Darstellung des Mythos im Spielsind also nicht nur in der Weise Spiele, daß die teilnehmenden Spieler im darstellenden Spiel sozusagen aufgehen und darin ihregesteigerte Selbstdarstellung finden, sondern sie gehen von sich aus dahin über, daß die Spielenden für den Zuschauer ein Sinnganzes

198

We notice here a transition of focus from actor to spectator. This is an instance of projective identificationof the variety that Young (1992) calls ’benign’. The notion of focus is important here. It is precisely sucha focus that the infant is missing, and that the mother is providing by offering a place to ’digest’ theinfant’s experiences.

Accordingly, the good mother,"the gratifying breast, taken in under the dominance of the sucking libido, is felt to be complete.This first internal good object acts as afocal point in the ego. It counteracts the processes ofsplitting and dispersal, makes for cohesiveness and integration, and is instrumental in building upthe ego." (Klein, 1952, p. 297; italics added)

11.2.3. Interpersonal conceptualizations

A major theoretical and clinical extension to Klein’s work has been made by Bion (1959, 1962b), himselfa Kleinian, when he distinguished two types of projective identification, a malicious one for the sake of’evacuation’, and a healthy one for the sake of ’communication’. "The difference depends on the degreeof violence in the execution of the mechanism" (Hinshelwood, 1991, p. 184). Subsequently a great numberof differentiations have been proposed with regard to the various dimensions of projective identification(e.g. Grinberg, 1979, p. 228). Hinshelwood points at the various interpersonal interpretations of theconcept of projective identification, as by Ogden (1982), and at the criticism that such interpretationsgenerate in circles of more orthodox object relation theorists.

Several grounds can be found for the existence of the opposition between the ’orthodox’ kleinians, whostuck to understanding projective identification in terms of unconscious phantasies, and others moreinterested in applying the term to interpersonal behavior patterns. Though the parties agree that projectiveidentification happens within an interpersonal context, the Kleinians prefer to see it as a phenomenon thattakes place at the level of "the use to which the analyst is put in being unwittingly drawn into thepatient’s phantasy world." (Hinshelwood, 1991, p. 200). In other words, it is a phenomenon that befallsthe therapist, in which he is made witness of some experience of his patient, in a way that he does notforesee. According to Hinshelwood (1991, p. 200) the controversy is about incompatible approaches: onemay approach the phenomenon by relating it to theory, i.e. by defining it, versus by indicating it inclinical material. Would an opposition between ’descriptive’ versus ’theoretical’ approaches be at the rootof the problem? We are reminded of our discussion of psychiatric diagnostic classification, where thissame opposition was also believed to be the crux (cf. footnote 412 on page 142).

Greenberg & Mitchell (1983)«535»maintain that the concept of projective identification has beenused for "integrating intrapersonal and interpersonal spheres in psychoanalysis"«536», a task too

darstellen. Es ist also gar nicht wirklich das Fehlen einer vierten Wand, das das Spiel zur Schaustellung veränderte. Das Offensein zumZuschauer hin macht vielmehr die Geschlossenheit des Spieles mit aus. Der Zuschauer vollzieht nur, was das Spiel als solches ist.(...)

Auch das Schauspiel bleibt Spiel, d.h. es hat die Struktur des Spiels, eine in sich geschlossene Welt zu sein. Aber das kultischeoder profane Schauspiel, so sehr es eine ganz in sich geschlossene Welt ist, die es darstellt, ist wie offen nach der Seite des Zuschauers.In ihm erst gewinnt es seine ganze Bedeutung. Die Spieler spielen ihre Rollen wie in jedem Spiel, und so kommt das Spiel zurDarstellung, aber das Spiel selbst ist das Ganze aus Spielern und Zuschauern.(...)

[Die] Art der Teilhabe [der Spieler] an dem Spiel ist nun nicht mehr dadurch bestimmt, daß sie ganz in ihm aufgehen, sonderndadurch, daß sie ihre Rolle in bezug und im Blick auf das Ganze des Schauspiels spielen, in dem nicht sie, sondern die Zuschaueraufgehen sollen. Es ist eine totale Wendung, die dem Spiel als Spiel geschieht, wenn es Schauspiel wird.Sie bringt den Zuschauer andie Stelle des Spielers.Er ist es - und nicht der Spieler -, für den und in dem das Spiel spielt. (...)Im Grunde hebt sich hier dieUnterscheidung von Spieler und Zuschauer auf.Die Forderung, das Spiel selbst in seinem Sinngehalt zu meinen, ist für beide diegleiche." (WM, pp. 104-5; italics added)

535. I am following here Kulish’ (1986) description of their ideas.

536. Kulish (1986, p. 102)

199

large for a single concept. The basic opposition was between"... Freud’s original drive/structure model, in which basic components of the mind derive frominstinctual drives, and the relational/structural models, in which basic mental structures evolve fromthe individual’s relations with other people. Klein stands as a transitional figure between the twomodels. Emphasizing her allegiance to the drive/structure model, she nevertheless pioneered thedescription of internal object relations." (Kulish, 1986, pp. 101-2)

11.3. The significance of the concept of ’unconscious phantasy’

11.3.1. ’Unconscious phantasy’ as a theoretical construct

There is a very delicate issue in the history of kleinian thought, viz. the opposition between Melanie Kleinand Anna Freud, an opposition that even took the shape of personal feuds and a schism between Britishpsychoanalysts during the 40’s of this century and even afterwards. The stake of this opposition was theintellectual legacy of Freud. Grosskurth (1986, p. 470) quotes from a letter of 1941 by Klein to ErnestJones: "It is tragic that his daughter, who thinks that she must defend him against me, does not realisethat I am serving him better than she."«537». According to Klein, neonates and infants should beconsidered as actively engaging into contact with their environment. They are to be seen as relating to’objects’«538»as much as do older (oedipal stage) children. Therefore, in defiance of Freud’s ownideas on pre-oedipal children, she elaborated methods of interpreting their unconscious processes, parexcellence, by means of observing their play.

On the other side, Anna Freud was faithful to her father’s ideas, and criticized Klein accordingly. AsIngleby puts it:

"In Anna Freud’s theory - derived from her father’s - the newborn is immersed in a state of’primary narcissism’. As long as a sense of reality has yet to be developed, there can be noquestion of social relations. [According to Anna Freud,] object relations are a function of the ego:they arise together with it against the end of the first year. Around that time the child becomesattached to itsreal parents, without the mediation of ’unconscious phantasies’ or ’internal objects’.As long as this attachment persists, it is hard for a therapist to establish an affective relation to theinfant: if the therapist succeeds, this necessarily subverts the infant’s relation to the parents - whichis not usually regarded as desirable. According to Melanie Klein, on the other hand, the infant’sfirst and deepest relation is not with the parents, but with the magical figures«539» inhabitingits ’unconscious phantasy’". (Ingleby«540», 1992, pp. 64-5).

Thus, Anna Freud criticized Klein in terms of her interference with the child’s parental binds, and Kleindefended herself in terms of the prominent position of the child’s unconscious phantasies concerningmagical figures! The unconscious phantasies, therefore, were not only part of Klein’s theoretical andpractical equipment, but also constituted a tool for legitimating her deviant position. It was by postulatingthe prominent position of the unconscious phantasies that Klein could account for her work on therelational aspects of infantile functioning, both theoretically and professionally.

537. cf. Ingleby (1992, p. 57)

538. The term ’persons’ is avoided, as according to Klein the young infant not yet distinguish between inanimate things andpersons, but is primarily oriented to the ’objects’ of its unconscious phantasies. (cf. Ingleby, 1992, p. 65)

539. a paradigm is the ’good breast’, an ever returning source of satisfaction as well as frustration for the infant

540. Professor Ingleby’s corrections to my translation of his originally Dutch text are gratefully acknowledged here.

200

11.3.2. ’Unconscious phantasy’ as a quasi-object: the ’as-if-proviso’

The unconscious phantasies were believed to exist, not as ’internal objects’, entities purported to beinsidethe mind, but rather as mental functions:

"As phantasies derive directly from instincts on the borderline between the somatic and psychicalactivity, these original phantasies are experienced as somatic as well as mental phenomena. (...)

For example, an infant going to sleep, contentedly making sucking noises and movementswith his mouth or sucking his own fingers, phantasies that he is actually sucking or incorporatingthe breast and goes to sleep with a phantasy of having the milk-giving breast actually insidehimself." (Segal, 1975, p. 13)

The concept of ’unconscious phantasy’ was useful to emphasize the as-if quality of unconsciousexperiences«541». This had as a major epistemological implication that by using this concept, itpermitted to make statements (usually called ’interpretations’) about other persons’ unconsciousexperiences, in terms of a proviso: the interpretations produced are quasi-objects«542»! The personspoken about is not supposed to know what is said about him. This is an epistemological trick, that hasoften been abused by therapists who were unwilling to test or modify their hypotheses about theirpatients’ experiences. Critical colleagues could be kept at a distance if the patient’s unconsciousphantasies by definition could not be measured empirically. The trick could even be applied, andsuccessfully, to patients themselves, in ploys of the variety "of course you are not doing this intentionally;it’s all unconscious".

But there is also a positive side of this ’trick’, which, I think, was predominant in Klein’s thinking. Forthis concept also enables a therapist to be his own ’critical colleague’, and stay aware of theincompatibility between the infant’s pre-linguistic experiences and the therapist’s own mature cognitiveand emotional modes of perceiving the world, i.e. his critical discourse! Hence, the concept ofunconscious phantasy entails the critical recognition that an adequate«543»description of theinfant’s experiences is not possible in terms of naive objects (such as ’good breast’). These descriptionsalways keep a proviso of ’as-if’-ness. I will call this "Klein’s as-if proviso", a term she did not use, butwhich I find useful here. Unlike the issues we discussed above in chapter 8 concerning the opposingcritical discourses in the field of psychiatric diagnosis and concerning the intermediating opportunities ofnaive discourse there, we find here, in the concept of unconscious phantasy as a purported vehicle forprojective identification, a recognitionthat a description in terms of naive objects does not cover theinfant’s archaic experiences, especially not those that the infant itself cannot keep for itself and has toevacuate (project) into the ’breast’.

For that reason, there is, implicit in the idea of projective identification as an unconscious phantasy, therecognition of the impossibility to adequately describe the infant’s experiences! This is a moment ofnegative understanding in Kleinian thinking, and if a therapist is using the concept of unconsciousphantasy in this way, then it serves as a support to his sincerity. The negative approach, therefore, doesnot mean simply to give up the attempt to understand, but to underscore the unconscious phantasycharacter, and hence to recognize de facto the as-if proviso of the descriptions made! The infant’sexperiencesare such that wecannotadequately describe them, and to call them ’unconscious phantasy’ isa way to recognize that!

541. e.g.: "Freud, in his description of the super-ego, is not implying that there is a little man actually contained in our unconscious,but that this is one of the unconscious phantasies which we have about the contents of our body and our psyche." (Segal, 1975, p.11)Accordingly, Segal maintains that also Klein used the concept of unconscious phantasy for the sake of avoiding misunderstandings.

542. See my definition of this term in section 2.2.3.

543. See also footnote 75 on page 39 concerning Schutz’ postulate of adequacy

201

11.4. Some implications of Bion’s ’container-contained’ model

With respect to the diverging interpretations of projective identification, we find now a peculiarcomplication: those theorists, most prominently Bion (1962b), who gave an interpersonal turn to Klein’s’unconscious phantasy’ based concept, actually seem to have taken her relational ideas to a logical close.For instance, Bion, in a passage immediately following the one quoted at the top of this section, created’for use as a model’ a basic change of emphasis:

"From the above theory I shall abstractfor use as a modelthe idea of a container into which anobject is projected and the object that can be projected into the container: the latter I shalldesignate by the term contained. (...)

Container and contained are susceptible ofconjunction and permeation by emotion. Thusconjoined or permeated or both they change in a manner usually described as growth." (Bion,1962b, p. 90; italics added)

Notice that Bion does not elaborate the ontological status of this working model. He simply puts ’as amodel’ the idea of conjunction of the two parts, container and contained, and in thisconjunction«544»the two parts can be permeated by emotion.

Accordingly, Bion comments on a case description in terms of ’splitting off his fears’ and ’putting theminto me’, without worrying too much onhow this could have been possible at all:

"When the patient strove to rid himself of fears of death which were felt to be too powerful for hispersonality to contain he split off his fears and put them into me,the [patient’s] idea apparentlybeing that if they were allowed to repose there long enough they would undergo modification bymy psyche and could then be safely reintrojected." (Bion, 1959, p. 103; italics added)

Here I put in italics Bion’s nearly relaxed reference to something that for Melanie Klein would have beena cornerstone of the argument. However, by not emphasizing the unconscious phantasy aspects, Bion, andafter him other theorists, also left Klein’s foothold. These theorists no longer specified that world ofmagical objects as the ground for projective identifications. But did they rather believe that the world of’real’ interpersonal behavior would do as well? I think it is here that a major source of confusion onprojective identification is to be found.

For the character of Bion’s container-contained idea, qua model, turned out not to be as prominent as itwas in his initial (1962b) presentation. In fact, Bion’s model quickly became his substantial theoreticaladaptation to Klein’s ideas on projective identification. It amounted to an interpersonal interpretation (cf.Grotstein, 1998b), one that takes projective identification as an event between two persons. But thoughBion did not reject the notion of unconscious phantasy as such, his ’model’ in fact begged the question ofthe ontological status of projective identification.

Now one implication of Bion’s container-contained ’model’ of projective identification, was that itpermitted the receiver of the evacuated experiences to formulate them asexperiences of himself, due tothe assumed conjunction and the permeation of emotions. For instance, Grotstein«545»relatedBion’s recommendation: "don’t listen to the patient; listen to yourself listening to the patient!". Thecontainer is believed to have immediate access to his own experiences, that is, he is believed to becapable of dealing with the patient’s experiences as his own. These experiences Bion calls ’the contained’.It is considered of less importance to whom they originally belong, than is their quality, especially theirpainfulness.

For instance, Bion (1992, p. 43) writes: "Anxiety in the analyst is a sign that the analyst is refusing to

544. cf. Jung’s ideas on conjunction (e.g. footnote 552 on page 203)

545. in his Turin lecture presenting his 1998b paper

202

’dream’ the patient’s material: (not dream) = resist = not (introject)."«546»There is an interestingresemblance to a finding by Thorpe (1982) on the methods of South-African traditional Xhosahealers«547». Also in Kulish (1986, p. 99) we find a description of the author’s utter confusion,when she recognized her patient’s dream content to be coinciding with that of a book she had beenreading recently! Interestingly, with respect to such dreamwork, interpersonal boundaries do not seem tobe the kind obstacle we would expect them to be.

Now if in Bion’s model the interpersonal boundaries are not taken into consideration, and if thetherapist’s access to the ’contained’ material is immediate, then this means that Bion’s model deprived thetherapist’s descriptions from their as-if proviso! This is the second implication of Bion’s model. I thinkthat it is this implication that eventually led to a search for a mechanism responsible for the processes ofprojective identification, a search for ontological explanations.

Thus, Bion’s model of container-contained was initially helpful inbeggingthe ontological questions onhow projective identification could be possible at all. As Ogden puts it:

"The relationship of container and contained is non-linear and must not be reduced to a linear,sequential schematization of the following sort: an aspect of the projector in phantasy and throughactual interpersonal interaction is induced in the Other; after being altered in the process of beingexperienced by a ’personality powerful enough to contain them’, these ’metabolized’ aspects of selfare made available to the projector who by means of identification becomes more fully able toexperience his thoughts and feelings as his own. Such a conception of projective identificationobscures the question of the nature of the interplay of subjectivities involved in projectiveidentification by treating the projector and recipient as distinct psychological entities. It is here thatthe dialectical nature of Bion’s concept of the container and the contained affords the possibility ofconceptually moving beyond the mechanical nature of the linear understanding of projectiveidentification just described." (Ogden, 1992b, p. 618)

Notice that Ogden is specifying here how projective identification doesnot work. Eventually Bion alsofound himself in need of ontological answers and developed a concept named "O" (cf. Grotstein, 1998a,b),by means of which individual experience was considered to be of a transcendental nature, comparable tothe platonic Idea and to the kantian thing in itself«548». It is to be left as an interesting question,however, whether the kantian thing in itself was really what Bion was after. As Grotstein relates:

"’O’ is a dark spot that must be illuminated by blindness," [Bion] stated«549». Bion liberallytranslated a letter by Freud to Lou Andreas Salomé«550»as "The analyst must cast a beamof intense darkness into the interior of the patient’s associationsso that some objectthat hashitherto been obscured in the lightcan now glow in that darkness". (Bion, personalcommunication)". (Grotstein, 1998b, p. 13; italics added)

546. I owe this reference to Mike Eigen (contribution to internet discussion list "Bion97", of may 15, 1998; to be found at http://pages.inrete.it/bion97/1339.html).

547. in a reply (may 18, 1998) to the discussion item mentioned in footnote 546: "The healers in the Xhosa church (GrahamstownSouth Africa) that I interviewed spoke clearly about how they dreamed their patient’s material, fought the demons in their dreams andthus cured the patient." (to be found at http://pages.inrete.it/bion97/1347.html)

548. Bion especially used the kantian concept, but I believe that this usage of the term may lead to misunderstandings. As Iunderstand Bion, it was primarily the infant’s experience qua raw and undigested experience, intolerable and unmanageable, that he triedto understand as something inexpressible. For each attempt to express it, is an act of ’digestion’. To call such an undigested orundigestible a ’thing in itself’, however, would put it at the transcendental plane where Kant put the things in themselves that we canonly think of as the sources of our sensory experiences. Rather, as I understand it, Bion’s idea of undigestible elements pertained tobodily, organic, experiences, that in some way violated the integrity of the (infant) organism. However inexpressible and undigestible, Iwould understand them not as ideal forms, of a transcendental world, but as pertaining to the infant’s world of sensory (painful or other)impressions.

549. Grotstein refers here to Bion’s Attention and Interpretation (1970, p. 88).

550. Grotstein refers here to Freud, S. & Andreas-Salomé, L., (1966), Letter dated "25.5.16", in: Letters. Ed. Ernest Pfeiffer. Trans.William and Elaine Robson-Scott. London: Hogarth Press, 1972, p. 45.

203

Rather than a kantian transcendental thing, we get here the same image as in Merleau-Ponty’s line: "In thedark night of thought dwells a glimmering of Being.«551»"

Also Ogden cannot entirely avoid ontology, in his treatment of projective identification. He presents hisidea of ’analytic third’ as the ’interpenetration of subjectivities’ (1992, p. 618), and he uses a conceptualframework of hegelian dialectics to formulate his ideas on how the subjectivities of the therapist and thepatient together may give rise to a third subject, the ’analytic third’ (1994a). We find here strongreminiscences to Jung’s (1946) treatment of the mysticalconjunctiobetween two persons, though Jungdoes also rely on the idea of unconscious phantasy«552» in describing his concept of ’induction’,which in fact is equivalent to ’projective identification’ (cf. Gordon, 1965; Schwartz-Salant, 1988). In fact,Jung’s elaboration in terms of the ’shared unconsciousness’ amounts to a transcendental argument: theshared unconscious is different from our common reality.

In general, it seems that the use of projective identification without the as-if proviso of the ’unconsciousphantasy’ conception has produced a great many ontological considerations that have not always beenhelpful tools for adequately describing the infant’s (or patients’ in general) experiences as conveyedthrough projective identification. This, I think, is the truth in Hinshelwood’s«553»remarks on theopposition between indicating projective identification in clinical material, versus defining it theoretically.The trap, I think, is that projective identification becomes defined in theoretical terms, perhaps even assome transcendental event (Jung), without too much clarity about how the infant’s (or patient’s)experiences relate to this domain of transcendental reality.

To resume, I discussed two implications of Bion’s model, a) the patient’s (or infant’s) evacuatedexperiences are understood as the containing person’s own experiences; these are the elements to bedigested, regardless of their origination; and b) if the as-if proviso is left, then the question remains whatactually projective identification can be. It is with respect to these two implications that Merleau-Ponty’sconcept of ’flesh’, which I introduced in section 3.1, is helpful.

As I remarked in section 7.4.2, the concept of ’flesh’ entails a convergence of knowing and being: theevent of experience is the ontological thing that is the case, and hence, the ontological status ofimmediately knowing the patient’s experiences, as in projective identification, can be understood as amoment of articulation of flesh; not in a transcendental reality, nor in unconscious phantasy, but in areality that is constituted by our experiences. That is: there is a vanishing boundary between the knowingact and the known, ’seer’ and ’seen’, sentient and sensible, touching and touched«554». This meanshere: the therapist’s experience (’the touching’) and the patient’s experience (’the touched’) cometogether,not as two entities that enter into a conjunctio mystica or the like, but as a subject and an objectthat meet at the point where they are both unreflected experience.

It is this concept of ’flesh’, therefore, that offers an explanation both for the immediacy of the therapist’s

551. Merleau-Ponty (1964b, p. 15) I found these lines in Madison (1981, p. 196), who comments them with: "Consciousness cannotthink Being except by thinking its own blindness, what it cannot see because it is that which makes it see." (ibid.)

552. "An unconscious tie is established and now, in the patient’s fantasies, it assumes al the forms and dimensions so profuselydescribed in the literature. The patient, by bringing an activated unconscious content to bear upon the doctor, constellates thecorresponding unconscious material in him, owing to the inductive effect which always emanates from projections in greater or lesserdegree. Doctor and patient thus find themselves in a relationship founded on mutual unconsciousness." (Jung, 1954, p. 176)["Es ist eine unbewußte Verbindung eingetreten, welche in der Phantasie des Patienten nun alle jene Formen und Dimensionen annimmt,von denen die Fachliteratur reichlich Kunde gibt. Damit, daß der Patient einen aktivierten Inhalt des Unbewußten an den Arztheranbringt, wird durch Induktionswirkung, die stets von Projektionen in mehr oder minderem Maße ausgeht, auch bei diesem dasentsprechende unbewußte Material konstelliert. Damit befinden sich Arzt und Patient in einer auf gemeinsamer Unbewußtheidberuhenden Beziehung." (1971, p. 187)]

Notice that ’shared unconsciousness’ would be a better English translation for ’gemeinsame Unbewußtheid’, than ’mutualunconsciousness’.

553. 1991 (p. 200)

554. see my quotation from Madison, 1981, p. 197, above in section 7.4.1

204

experiences and for the ontological status of projective identification as a relation more primary thanbetween subject and object.

In the sections that now follow I will elaborate these two implications. In section 11.5 I will relate thefirst implication, that of the immediacy of experience, to the concept of ’incorporation’, which we alreadyencountered in section 4.2. Next, in section 11.6 I will relate the second implication, that concerning theontological status of projective identification, to the concept of ’one-sided boundary’, which I introducedin section 7.3.2.

11.5. Projective identification as incorporation

According to Bion, it was the containment by the mother of unbearable experiences in the infant, thatwould lead to the infant’s capacity to endure those experiences itself. Conversely, the infant can be said toseize the opportunity of having the mother digest these experiences. This, according to Bion, was to beunderstood as the onset ofthinking in the infant (Bion, 1962a). The mother’s major instrument ofcontainment was what Bion called her ’reverie’«555».

Accordingly, if the child obtained less than the critical minimum of containment, it would not learn tocoordinate its unbearable experiences itself, and recognize these in terms of the good mother’s absence.Projective identification would become a malicious component of the person’s later modes of functioning,where other persons are abused for the containment of experiences that otherwise, by ’digesting’ them,would have become thoughts«556».

Hence, we notice here a convergence with Maturana’s notion of second order distinctions and Bion’s ideaof the ’tolerance’ that an infant should develop for painful experiences. It is this tolerance that we cannow understand in terms ofa capacity to make second order distinctions without the supporting presenceof another individual to coordinate actions with!«557»To speak in terms of GiambattistaVico«558», the infant cannot yet find a ’selected fact’, by means of which other experiences can bearranged«559». And it is this capacity in the mother for which the infant needs his painfulexperiences to be contained by the mother.

555. This concept, introduced by Bion can be considered a major extension of Freud’s ideas on the analyst’s attitude. For it is theattitude of reverie that enables the mother to be maximally sensitive to her child’s emotions, and this concept serves as a model for theattitude that enables the therapist to receive the patient’s undigestible experiences.

Bion writes: "[R]everie is that state of mind which is open to the reception of any "objects" from the loved object and istherefore capable of reception of the infant’s projective identifications whether they are felt by the infant to be good or bad." (Bion,1962b, p. 36)

556. Spillius gives a clear description of the gist of Bion’s ideas on projective identification and the onset of thinking. This happenswhen "no breast [is] available for satisfaction. What happens next depends on the infant’s capacity to stand frustration. Klein hadpointed out that in earliest experience an absent, frustrating object is felt to be a bad object. Bion took this idea further. If the infant’scapacity for enduring frustration is great, the ’no-breast’ perception/experience is transformed into a thought, which helps to endure thefrustration and makes it possible to use the ’no-breast’ thought for thinking, that is, to make contact with, and stand, his persecution andthen split it off when the external breast arrives again. Gradually this capacity evolves into an ability to imagine that the bad feeling ofbeing frustrated is actually occurring because there is a good object which is absent but which may return. If however, capacity forfrustration is low, the ’no-breast’ experience does not develop into the thought of a ’good breast absent’; it exists as a ’bad breastpresent’; it is felt to be a bad concrete object which must be got rid of by evacuation, that is, by omnipotent projection. If this processbecomes entrenched, true symbols and thinking cannot develop." (Spillius, 1988c, pp. 154-5)

557. Notice that Maturana does not elaborate on the conditions for making second order distinctions by a single individual. Cf.Maturana’s dictum: "Everything said is said by an observer to another observer that could be him or herself" (1987a, p. 332).

558. cf. section 3.3

559. It is interesting that also Bion used the term ’selected fact’, be it in the sense of: "... a particular fact that suddenly occurs tothe analyst which makes sense of the disparate elements previously noted. What before may have been a jumble of fragmented materialnow becomes unexpectedly coherent and understandable; meaning suddenly dawns." (Symington & Symington, 1996, p. 92).

205

The mother’s reverie is a performance that we should understand as an instance of incorporation. Thoughthis term is not used very often in the literature«560», many formulations hint at it«561». Insection 4.2 I presented the idea of incorporation in terms of a ’combination’ of first and second orderdistinctions, that is used as second order distinctions in a ’new’ situation, to distinguish ’new’ first orderdistinctions. I repeat it here for the reader’s convenience:

old situation :first order distinctions:

experiences, to be used as tokens for the concrete objects

second order distinctions:coordinations of these experiences, making them into tokensfor the concrete objects

new situation :first order distinctions:

newly occurring experiences due to the instrumental usage ofthe concrete objects

second order distinctions:a combination of the first and second order distinctions ofthe ’old situation’, using the first order distinctions ofthe ’new situation’ as tokens for the newly emerging gestalt(consisting of the concrete objects of the ’old situation’ asits elements)

It is with respect to this idea of an ’old’ and a ’new’ situation, that we may now understand projectiveidentification as a situation in which the infant’s first order distinctions become the mother’s ’new’ firstorder distinctions, which, by ’reverie’, are ’digested’ and coordinated by her into new naive objects:

new situation :mother’s first order distinctions:

infant’s experiences

mother’s second order distinctions:a combination of mother’s original first and second orderdistinctions, using the infant’s experiences as tokens forthe newly emerging gestalt

For instance, when the baby is in pain, the mother looks for modes of carrying or rocking it, with thepurpose of decreasing the infant’s discomfort. The latter’s sounds and gestures are taken by the mother asher own immediate sensory experiences, through which she experiences the infant’s discomfort as herown.

560. An exception is Grinberg (1962, p. 439), who also uses the term ’incorporate’ to denote the reception by the therapist of theprojective identification

561. cf.: "The nature of the functions which excite the patient’s curiosity he explores by projective identification. His own feelings,too powerful to be contained within his personality, are amongst these functions. Projective identification makes it possible for him toinvestigate his own feelings in a personality powerful enough to contain them. (Bion, 1959, p. 106)

Cf. also Arlow (1993), as cited by Sandler (1995, p. 67:"Terwijl de analyticus passief naar de patiënt luistert, ervaart hij wat de patiënt verbaal voortbrengt als zijn directe zintuiglijkewaarneming. Dit neemt zijn aandacht vrijwel volledig in beslag. Op dat moment bestaan de voorstellingen van de patiënt uit datgene wathet bewustzijn van de analyticus gepresenteerd krijgt, met als gevolg dat hij zich gaat identificeren met de analysant [sic]."

[As the analyst passively listens to the patient, he experiences what the patient’s account generates as his immediate sensoryperception. This absorbs nearly his complete attention. At that moment the patient’s representations consist of that which is presented tothe analyst’s consciousness, and as a result he will identify with the analysand." (my translation)]

206

But this is not confined to the care of infants. For instance, a 5 year old visits with his mother a skyscraper. As the child looks from the balcony downward, without any danger of falling, and looks at thetraffic some 150 meters below, the mother, when noticing what her child is doing, feels her own stomachas if she were looking downward herself. As she sees him looking downward, she is no longer interestedin the difference between her own physical location and his. She does no longer distinguish between theirtwo positions and responds accordingly.

This is comparable to what happens when a blind man uses his stick: then the tip of his stick becomes(new situation) the locus of his first order distinctions, and his former first order distinctions, thesensations of the stick in his hand, become second order distinctions that coordinate and make the firstorder distinctions into tokens for the objects he is perceiving (cf. section 4.2).

Thus, we may understand projective identification as an operation inverse to incorporation. By thisoperation, the patient comes to dispose of part of the therapist’s capacities to perceive and think. We maycall this an instrumental usage, especially in its virulent«562»manifestations. If anything, thetherapist has become here a tool, an instrument that releases the patient from bearing his own painfulexperiences, not unlike the way a butler releases his employer from many practical affairs in the house.

In the course of emotional development, the infant is supposed to introject the mother’s capacity tointegrate and coordinate its discomfort, and in the mean time it is the mother who takes over the infant’simmature coordination of its own experiences.

Bick formulates the same double role in terms of the infant’s integration through introjection of themother’s capacity of integration:

"The thesis is that in its most primitive form the parts of the personality are felt to have no bindingforce amongst themselves and must therefore be held together in a way that is experienced by thempassively, by the skin functioning as a boundary. But this internal function of containing the partsof the self is dependent initially on the introjection of an external object, experienced as capable offulfilling this function. Later, identification with this function of the object supersedes theunintegrated state ..." (Bick, 1988, p. 187)«563»

11.6. Projective identification and the one-sided boundary

I will now turn to the second implication of Bion’s model, which pertained to the ontological status ofprojective identification. We can now understand the idea of the mother’s reverie in terms of herapproximation of the interface between the infant and herself. Her being ’open to the reception of any"objects"’ is her approximation of this interface, her dreaming away is her passage of a one-sidedboundary.

The idea presented here is that Bion’s interpretation of projective identification emphasizes the absence ofdistinction between the ’sender’ and the ’receiver’ of the projective identification; or better: it emphasizesthat the two parties involved may temporarily and partially lose their individuality, not merely theirsenseof being distinct, and become identified. Bion’s idea of containment as a conjunction in fact highlightsthis moment of identification. It ignores the aspect of phantasy, and instead, attempts to express thecontents of this phantasy, without bringing it home as a phantasy. We might call this a forgetfulness, onethat requires the passage of a one-sided boundary, which cannot be described adequately from the positionof an observer. The recognition of this inadequacy, indeed, was what I called (section 11.3.2) Klein’s as-ifproviso.

We have, therefore, a peculiar situation with respect to the interactional conception of projective

562. cf. Young (1992)

563. cf. also my quotation above from Klein, 1952, p. 297) on page 198 of section 11.2.2

207

identification: the critical description of projective identification as an interactional event (qua "linearsequential schematization", as Ogden called it) is different from the experience of being involved in aprojective identification; the former utilizes a particular distinction that is missing in the experience itself.It is here that we find a one-sided boundary. For any attempt to describe projective identification incritical terms has the feature of the position called ’outside’: it describes and makes a distinction betweentwo parts (infant and mother; patient and therapist), which are declared to become identified, at least inphantasy, due to a particular kind of ’interaction’ or a particular ’exchange’.

From the position called ’inside’, on the other hand, there is precisely the lack of any awareness ofprojective identification that makes it possible to occur. Furthermore, it is by pointing at the occurrence ofprojective identification (e.g. as a phantasy in the patient) that an analyst can stay in (or return to) the’outside’ position and thus maintain (or regain) the distinction between himself and his patient.

The containment relation, on the other hand, is not defined in terms of the asymmetry between ’inside’and ’outside’, but, instead, in terms of the containing person (mother, therapist) as immediately digestingthe other person’s (infant, patient) undigestible experiences. Here the two parties have the sense of beingindistinct. Is this indistinctness a matter of ’phantasy’, i.e. unreal? Winnicott gives a helpful suggestion,that reminds us of James’ ’psychologist’s fallacy’:

"The infant perceives the breast only in so far as a breast could be created just there and then.There is no interchange between the mother and the infant. Psychologically the infant takes from abreast that is part of the infant, and the mother gives milk to an infant that is part of herself. Inpsychology, the idea of interchange is based on an illusion in the psychologist." (Winnicott, 1973,p. 14)

At this place I like to push my point a little further. If those involved in a projective identification(partially and temporarily) share a sense of identity, as a primary way of allowing one (the baby, thepatient) to make use of the digestive capacities in the other (the mother, the therapist) then there issomething going on that does not fit with traditional notions of each person’s individuality, especially hisinner life, being accessible only to himself«564».

To put it differently, if we assume a reality pre-existent to our distinctions, as in regular objectivisttheories, it becomes essential to distinguish between physical reality versus patient’s and therapist’sphantasies! For in such pre-existent reality it should be possible, in analogy to Quine’s (1969; cf. footnote61, page 34) arguments, to ’naturalize’ the facts of projective identification, and to look for empiricalcorrelates of their occurrence! The easiest to be observed, then, would be the individual distinctness of thetwo interacting persons.

But when speaking about the condition of containment, we are pointing at the inner side of a one-sidedboundary, there where we cannot reach with our critical understanding! It is here that we run against ourcommon assumption of reality as pre-existent to our experiences: for it is at this level of functioning thatwe find a limit to the usual opposition between objective fact and subjective experience. We are speaking,so to say, about a’quantum of distinction’: the minimal act of distinction, in which no differentiation canbe made between the event of experience and the thing experienced; it is here that distinctions are thearticulations, not the subjective representations, of the flesh. In this way is the locus of projectiveidentification between two persons a ’true negative’. It consists of the absence of those distinctions thatwe, external observers, need in order to think. This ’true negative’ exists as the other side of a one-sidedboundary, there where identity is more fully present than a statement of identity«565»could declareit to be.

Hence, we find here an instance of the anti-solipsist interpretation I gave of Maturana in section 2.3.4. Forit is in the occurrence of a containment relation that a true identification can occur between those

564. see also section 9.3 on ’reasoning by analogy’

565. cf. Heidegger’s "Satz der Identität", see section 3.4, page 50

208

involved, one that can exist only for them, not for an external observer. Accordingly, such aninaccessibility of these interpersonal events is not compatible with a belief in their empirical investigationas positive facts, i.e. facts open to empirical methods of research.

Conversely, if the events studied are of a level sufficiently above that of the ’quantum of distinction’, thenwe can investigate them critically by means of empirical methods. Therefore, the empirical inaccessibilityof the containment relation does not compel us to embrace a solipsist philosophy, or one of ’radicalrelativism’«566». It cannot withhold people, whenever naive objects appear as conflicting, to lookfor critical objects; that is: to look for aspects (e.g. empirical findings) about which agreement can begained. But when no naive objects are constructed, for instance, as in projective identification, when thebehaviors observed are the processes of the construction of second order coordinations, especially at theirmoments of faltering and imperfection, then an objectivist and critical attitude will be of no avail to anobserver.

11.7. Beyond the modeling relation

Incorporation is a transition from having one’s own experiences to using them for the purpose of havingother experiences. Accordingly, the therapist can utilize his own experiences in order to accept those ofhis patient as his own. But in order to do so, he has to forget the things he has been thinking of andperceiving thus far. The act of incorporation, therefore, amounts to the passage of a one-sided boundary.

Greenson, usually not known for his discussions of projective identification, describes his work in termsof a ’working model of the patient’ which the therapist is to build as a "counterpart or replica" (1960,p. 421) of the patient, and from which he is able to predict the patient’s associations and experiences.However, Greenson also uses phrases as "I permitted part of myself to become the patient" (p. 421) and"to experience the feelings of another person" (p. 418). My point is that, when the therapist permits partof himself to become the patient (or, to permit the patient to occupy part of himself) he isnot in theposition to critically use a model as a ’counterpart or replica’. Rather, the very concepts of model, replica,representation etc. do no longer apply to his understanding of the patient«567».

If the therapist’s understanding of the patient thus far can be called a model, now it disappears as such. Itis not incorporated as in classical science’s use of models (cf. section 6.1.2), but the patient is no longerperceived as a separate individual, and the therapist’s capacity to perceive him as such is affected. Instead,to the extent the therapist loses sight of the patient, he may start to perceive and think on behalf of thepatient, i.e. as a substitute for the patient’s missing perception and thought. The therapist becomes (partof) the patient by (partially) incorporating the patient himself,not a model of the patient. Accordingly, themodeling relation (see section 6.1.2), far from being trivial as in classical science, has become the veryscene of the drama: it is this relation that is undermined and eventually reduced to zero.

First, the therapist has an understanding of the patient and is capable of critically observing him,reflecting on his behaviors and on his words, as well as reflecting on his own possible replies, etc. Then,when projective identification starts, the therapist loses his critical distance, and with it, his model of thepatient. The naive objects the therapist now comes to perceive, if any, are those that the patient himselfcould not coordinate (digest), and that he is letting now to the therapist.

That is: eventually the therapist’s model of the patient and the patient himself converge; the therapist hasno longer his own model of the patient at his disposal; rather, (part of) the patient has become for thetherapist what his model was before. The therapist can discover this development only by noticing the

566. For instance, Carveth (1995) gives some interesting arguments against a ’radical relativism’: "If classical Freudian theory isvalidated as true by members of the New York Psychoanalytic Institute, does that make it true? I have no doubt that the nonexistence ofthe Holocaust is psychologically real to those whose experience is validated by the revisionist historians. Does that mean that theHolocaust did not occur?" (p. 19).

567. Notice that also Freud eventually preferred to put the idea of evenly-suspended attention ’in a formula’, cf. section 9.3.

209

decrease of his own capacities to critically think and perceive, and instead the increase of particular naiveobjects (e.g. extremely ’bad’ objects, corresponding to negative self-experiences in the therapist). It is herethat empirical findings are to be found that are as radical as William James (cf. section 7.4.2) proposedthem to be. It is also here that the therapist can make critical usage of naive discourse and even of hispre-reflective experiences.

These are findings that concern the ways perceptual and cognitive performance is affected by the object ofattention. It is not a domain of empirical phenomena of which psychotherapists are holding the exclusiverights. Researchers too may enrich their arsenal of empirical observation techniques with them.

12ConclusionWe have started this investigation from the idea that in psychotherapy some things may go on that arehard to observe empirically, because of some inherent property of those things, not because of a defect ofour measurement instruments. In this book I have specified this property in terms of a certain self-referentiality that may enter into psychotherapeutic sessions, and that may effectively impede an externalobserver to get the gist of what is taking place.

First, the self-reference of psychotherapeutic encounters is incompatible with the conduct of regularempirical research, because the performance of the therapist cannot be caught in procedural terms. Thereis no satisfying way to formalize the therapeutic task, as for instance in a protocol. Put otherwise, if sucha protocol is constructed, it is necessarily incomplete with respect to those tasks that are self-referential.This holds especially for the task of exploration. Accordingly, a researcher interested in ’what thetherapist is doing’ cannot succeed without leaving the format of a therapeutic procedure.

Second, the patient’s and the therapist’s primary experiences cannot be perceived without making theminto tokens for perceivable objects that they arenot themselves; that is: in their raw, unbound state theyhave an unmistakable impact upon the persons involved, but they cannot appear as objects. If theresearcher is willing to study their impact upon himself, then he has to radically use his own experiences,in a self-referential act of exploring them, not in a study of something external to him.

In order to learn how his concepts are inadequate, and how his own process of (mis)understandingdevelops as a function of the object of his study, he has to take these failing descriptions asmanifestations of something that he cannot (or not yet) grasp. It is here that those who are studied may behelpful in finding resemblances and comparisons between the process of investigation and its object.Leading questions may be, for instance, "how does my understanding of your session (your work/youractivities) miss the point? what am I overlooking? what makes me overlook it? how does it resemble youryour own failures to understand? how does it resemble the patient’s difficulties to change?". Thesequestions do not warrant positive solutions to the measurement problems of psychotherapy research, but atleast they prevent some unwarranted simplifications.

Before disqualifying this approach as ’subjective’, we should realize that a more objective counterpartdoes not exist. Merleau-Ponty is given the final word:

"The sensible is that: this possibility to be evident in silence, to be understood implicitly, and thealleged positivity of the sensible world (...) precisely proves to be anungraspable, the only thingfinally that is seen in the full sense is the totality wherein the sensibles are cut out." (1968a,p. 214)«568»

568. "Le sensible, c’est cela: cette possibilité d’être évident en silence, d’être sous-entendu, et la prétendu positivité du mondesensible (...) s’avère justement comme uninsaississable, seule se voit finalement au sens plein la totalité où sont découpés les sensibles."(1964a, p. 267-8)

210

ReferencesAbelson, R.P., "Computer simulation of ’hot cognitions’", in: S. Tomkins, S. Messick (eds.),Computer

simulation of personality. New York: Wiley, 1963.American Psychiatric Association,Diagnostic and statistic manual of mental disorders (DSM-III-R)/

(DSM-IV). Washington DC: Am. Psychiatr. Association, 1987/1994.Anderson, R. (ed.),Clinical lectures on Klein and Bion. London/New York: Tavistock/Routledge, 1992.Anzieu, D., "Au fond du soi, le toucher",Revue française de Psychanalyse, (1984)48, 6, 1385-1398.Apprey, M., "A prefatory note on motives and projective identification",Int. J. of Pyschoanalytic

Psychotherapy, (1986)11, 111-116.Arieti, S., "Creativity and its cultivation: relation to psychopathology and mental health.", in: D.A.

Hamburg, H.K. Brodie (eds.),American Handbook of psychiatry, second edition. Vol. 6: Newpsychiatric frontiers. New York: Basic Books, 1975.

Arieti, S., Creativity. The magic synthesis. New York: Basic Books, 1976.Aristotle, The Organon. The Categories. London/Cambridge, Mass.: Heinemann/Harvard UP, 1955.Arlow, J.A., "The supervisory situation",J. Amer. Psychoanalytic Assn., (1963)XI , 576-594.Arlow, J.A., "Discussion on Baranger’s paper on ’The mind of the analyst: from listening to

interpretation’",Int. J. of Psychoanalysis, 74, 1174-1154.Bakker, R.,De geschiedenis van het fenomenologisch denken. Utrecht: ’t Spectrum, 1966.Baranes, J.-J., "Discussion de la communication d’Edmond Gilliéron: Pourqoi le groupe?",Revue de

psychothérapie psychanalytique de groupe, (1991)14, 21-26.Bateson, G., D.D. Jackson, J. Haley, J.H. Weakland (1956), "Toward a theory of schizophrenia", in: G.

Bateson,Steps to an ecology of mind. New York: Ballantine Books, 1972.Bateson, G.,Mind and nature. A necessary unity. New York: Dutton, 1979.Bayer, R., R.L. Spitzer, "Neurosis, psychodynamics, and DSM-III. A history of the controversy",Archives

of General Psychiatry, (1985)42, 2,187-196.Berg, J.H. van den,Metabletica. Nijkerk: Callenbach, 1956.Bergin, T.G., M.H. Fisch, "Preface", in: G.B. Vico,The New Science. Unabridged translation of the Third

Edition (1744) with the addition of "Practic of the New Science". Translated by T.G. Bergin &M.H. Fisch. Ithaca/London: Cornell UP, 1986.

Berkeley, G., (1734) "Introduction to A Treatise Concerning the Principles of Human Knowledge"[excerpt], in: R.J. van Iten (ed.),The problem of universals. New York: Appleton-Century-Crofts,1970.

Bernstein, R.J.,Beyond objectivism and relativism: science, hermeneutics, and praxis. Philadelphia:Philadelphia U.P., 1983.

Bick, E., "The experience of the skin in early object-relations", in: E. Bott Spillius (ed.),Melanie Kleintoday. Developments in theory and practice. Volume 1: mainly theory. London/New York:Routledge, 1988.

Bion, W.R., (1959) "Attacks on linking", in: W.R. Bion,Second thoughts. New York: Aronson, 1967.Bion, W.R.,Experiences in groups, and other papers. London: Tavistock, 1961.Bion, W.R., (1962a) "A theory of thinking", in: W.R. Bion,Second thoughts. New York: Aronson, 1967.Bion, W.R.,Learning from experience. New York: Basic Books, 1962(b).Bion, W.R.,Cogitations. (Edited by Francesca Bion) London: Karnac, 1992.Blashfield, R.K., "Feighner et al., invisible colleges and the Matthew effect",Schizophrenia Bulletin,

(1982)8, 1,1-6.Bohr, N., "The quantum postulate and the recent development of atomic theory", in:Nature (Supplement),

(1928)121, 580-590.Bohr, N., "Light and life" (1932), in: N. Bohr,Atomic physics and human knowledge. New York: Wiley,

1958.Bohr, N., "The quantum of action and the description of nature", in: N. Bohr,Atomic theory and the

description of nature. Cambridge: Cambridge U.P., 1934.Bohr, N., "Discussion with Einstein on epistemological problems in atomic physics", in P. Schilpp (ed.),

Albert Einstein: philosopher-scientist. Evanston, Ill.: The library of living philosophers, 1949.

212

Bohr, N., Atomic physics and human knowledge. New York: Wiley, 1958.Bohr, N., Essays 1958-1962 on atomic physics and human knowledge. New York: Random House, 1963.Bohr, N., Collected worksvol. 6. (E. Rüdinger, general editor; J. Kalckar, editor). Amsterdam: North-

Holland, 1985.Bromberg, P.M., "The supervisory process and parallel process in psychoanalysis",Contemporary

Psychoanalysis, (1982)18, 92-111.Bromberg, P.M., "The third ear", in: L. Caligor, Ph.M. Bromberg, J.D. Meltzer,Clinical perspectives on

the supervision of psychoanalysis and psychotherapy. New York/London: Plenum, 1984.Brown, G. Spencer,Laws of form. London: Allen & Unwin, 1969.Buber, M., (1922)Ich und Du. Heidelberg: Schneider, 1983. Translated by W. Kaufmann:I and Thou.

New York: Charles Scribner’s Sons, 1970.Buber, M.,Das Problem des Menschen. Heidelberg: L. Schneider, 1948.Buber, M., "Vorwort", in: H. Trüb,Heilung aus der Begegnung. Ein Auseinandersetzung met der

Psychologie C.G. Jungs. Stuttgart: Klett, 1952. Translated by M. Friedman: "Healing throughmeeting", in: M. Buber,Pointing the way. Collected essays. London: Routledge & Kegan Paul,1957.

Butts, R.E., "Some tactics in Galileo’s propaganda for the mathematization of scientific experience", in:R.E. Butts, J.C. Pitt (eds.),New perspectives on Galileo. Dordrecht/Boston: Reidel, 1978.

Caligor, L., P.M. Bromberg, J.D. Meltzer,Clinical perspectives on the supervision of psychoanalysis andpsychotherapy. New York/London: Plenum, 1984.

Caligor, L., "Parallel and reciprocal processes in psychoanalytic supervision", in: L. Caligor, Ph.M.Bromberg, J.D. Meltzer,Clinical perspectives on the supervision of psychoanalysis andpsychotherapy. New York/London: Plenum, 1984.

Campbell, D.T., D.W. Fiske, "Convergent and discriminant validation by the multitrait-multimethodmatrix", Psychological Bulletin, (1959)56, 81-105.

Carveth, D.L., "Self psychology and the intersubjective perspective: a dialectial critique",Progress in SelfPsychology, (1995)11, 3-30.

Ciompi, L., Affektlogik. Ueber die Struktur der Psyche und ihre Entwicklung. Ein Beitrag zurSchizophrenieforschung. Stuttgart: Klett-Cotta, 1982.

Copleston, F.,A history of philosophy, vol. 1 part II. New York: Image Books, 1962.Cusanus, Nicolaus (1514),De Docta Ignorantia. Hamburg: Meiner, 1970.Dell, P.F., "Pathology: the original sin", Lecture presented to the First International Conference on

Epistemology, Psychotherapy, and Psychopathology, Houston TX, september 1982. Unpublishedmanuscript, 1983.

Denzin, N.K.,On understanding emotion. San Francisco: Jossey-Bass, 1984.Descartes, R.,Discours de la méthode(1637). Page number references in rectangular brackets [AT]

always refer to the edition of C. Adam, P. Tannéry,Oeuvres de Descartes, Paris, 1897-1913.French quotations have been derived from R. Descartes, Oeuvres philosophiques, Tome I (1618-1637. Édition de F. Alquié. Paris: Garnier, 1963. English quotations have been derived from E.S.Haldane, G.R.T. Ross,The philosophical works of Descartes, Vol. I. Cambridge: Cambridge U.P.,1973.

Descartes, R., Letter to Morin, 13 July 1638. C. Adam, P. Tannéry,Oeuvres de Descartes, Paris, 1897-1913, Vol. II. Paris: Vrin, 1975.

Descartes, R.,Meditationes de prima philosophia(1641). Page number references in rectangular brackets[AT] always refer to the edition of C. Adam, P. Tannéry,Oeuvres de Descartes, Paris, 1897-1913.English quotations have been derived from E.S. Haldane, G.R.T. Ross,The philosophical works ofDescartes, Vol. I. Cambridge: Cambridge U.P., 1973.

Descartes, R.,Geometry/Géometrie(1637) (translated from the French and Latin by David Eugene Smithand Marcia L. Latham). New York: Dover Publications, 1954.

Devereux, G.,From anxiety to method in the behavioral sciences. Den Haag: Mouton, 1967.Dies, R.R., "Leadership in short-term group therapy: manipulation or facilitation?",Int. J. Group

Psychotherapy, (1985)35, 3, 435-455.Doehrmann, M.J.G., "Parallel processes in supervision and psychotherapy",Bulletin of the Menninger

Clinic, (1973)40, 1, 3-104.Dreyfus, H.L., S.E. Dreyfus,Mind over machine. The power of human intuition and expertise in the era of

213

the computer. New York: Free Press, 1986.Dummett, M., (1963) "The philosophical significance of Gödel’s theorem", in: M. Dummett,Truth and

other enigma’s. London: Duckworth, 1978.Egan, G.,The skilled helper. A model for systematic helping and interpersonal relating. Monterey,

California: Brooks/Cole, 1975.Ekstein, R., R.S. Wallerstein,The teaching and learning of psychotherapy. New York: Int. Univ. Press,

1958.Emmelkamp, P.M.G., "Behavior therapy with adults", in: S.L. Garfield, A.E. Bergin (eds.),Handbook of

psychotherapy and behavior change (Third Edition). New York: Wiley, 1986.Erickson, M.H., E.L. Rossi, S.I. Rossi,Hypnotic realities. The induction of clinical hypnosis and forms of

indirect suggestion. New York: Irvington Publishers, 1976.Erikson, E.H.,Toys and reasons: stages in the ritualization of experience. London: Boyars, 1978.Erwin, E., Pholosophy and psychotherapy. Razing the troubles of the brain. London etc.: Sage, 1997.Eysenck, H.J.,The scientific study of personality. London: Routledge & Kegan Paul, 1952.Eysenck, H.J., "A critique of contemporary classification and diagnosis", in: T. Millon, G.L. Klerman

(eds.),Contemporary directions in psychopathology. Toward the DSM-IV. New York: GuilfordPress, 1986.

Faust, D., R.A. Miner, "The empiricist and his new clothes: DSM-III in perspective",Am. J. ofPsychiatry, (1986)143, 8,962-967.

Feighner, J.P., E. Robins, S.B. Guze, R.A. Woodruff, G. Winokur, R. Munoz, "Diagnostic criteria for usein psychiatric research",Archives of General Psychiatry, (1972),26, 57-63.

First, M.B., R.L. Spitzer, M. Gibbon, J.B.W. Williams,Structured Clinical Interview for Axis I DSM-IVDisorders - Patient edition (SCID-I/P, Version 2.0). New York: Biometrics Research Dept., 1994.

Flynn, B.C., "The question of ontology: Sartre and Merleau-Ponty", in: G. Gillan (ed.),Horizons of theflesh. Critical perspectives on the thought of Merleau-Ponty. London & Amsterdam: Feffer &Simons, 1973.

Foerster, H. von, "On constructing a reality", in: W.F.E. Preiser (ed.),Environmental design research, vol.2. Strousbury: Dowden, Hutchinson & Rose, 1973.

Foerster, H. von, "Les objects, gages de comportements propres", in: B. Inhelder, R. Garcia, J. Vonèche(eds.),Hommage à Jean Piaget. Epistémologie génétique et équilibration. Neuchâtel, Paris,Montreal: Delachaux & Niestlé, 1977.

Foerster, H. von, "Objects: tokens for (eigen-)behaviors", in: H. von Foerster,Observing systems. Seaside,Cal.: Intersystems Publ., 1981.

Foerster, H. von, "Kybernetik einer Erkenntnistheorie", in: H. von Foerster,Sicht und Einsicht.Braunschweig: Vieweg, 1985.

Foerster, H. von, "Perceiving the fourth spatial dimension". Lecture given at the conference ’Evolutionand Cognition. The heritage of Jean Piaget’s genetic epistemology.’ Bergamo, 6-8 october, 1990.In: M. Ceruti (ed.),Evoluzione e conoscenza. L’epistemologia genetica di Jean Piaget e leprospettive del costruttivismo. Bergamo: Lubrina, 1992.

Folse, H.J.,The philosophy of Niels Bohr. The framework of complementarity. Amsterdam: North-Holland,1975.

Folse, H.J., "Complementarity and our knowledge of nature", in: M.E. Carvallo (ed.),Nature, cognitionand system II: On complementarity and beyond. Dordrecht, London, Boston: Kluwer, 1992.

Foulds, G.A.,Personality and personal illness. London: Tavistock, 1965.Frances, A., A.M. Cooper, "Descriptive and dynamic psychiatry: a perspective on DSM-III",Am. J. of

Psychiatry, (1980)138, 9, 1198-1202.Freud, S., "Recommendations to physicians practising psycho-analysis (1912)",Standard Edition, 12,

111-120.Gadamer, H.G.,Wahrheit und Methode. Grundzüge einer philosophischen Hermeneutik. Tübingen: Mohr,

1960 (WM). Translated by G. Barden and J. Cumming:Truth and method. New York: Crossroad,1986(a) (TM).

Gadamer, H.G.,Die Aktualität des Schönen. Kunst als Spiel, Symbol und Fest. Stuttgart: Reclam, 1977.Translated by N. Walker: "The relevance of the beautiful", in: H.G. Gadamer,The relevance of thebeautiful and other essays. Cambridge: Cambridge UP, 1986(b).

Garrison, J.W., "Husserl, Galileo, and the processes of idealization",Synthese, (1986)66, 329-338.

214

Geertz, C., "On the nature of anthropological understanding",Am. Scientist, (1975)63, 47-53.Gendlin, E.T., "The small steps of the therapy process: how they come and how to help them come", in:

G. Lietaer, J. Rombauts, R. van Balen (eds.),Client-centered and experiential psychotherapy in thenineties. Leuven: Univ. Pers Leuven, 1990.

Gergen, K.J., "Theory of the self: impasse and evolution", in: L. Berkowitz (ed.),Advances inexperimental social psychology. Vol. 17: Theorizing in social psychology: special topics. NewYork: Academic Press, 1984.

Giddens, A.,New rules of sociological method. A positive critique of interpretative sociologies. London:Hutchinson, 1976.

Giel, R., Waarom een psychiatrische diagnose? Een beknopte inleiding in het medische model en inandere modellen. Alphen a/d Rijn, Brussel: Stafleu, 1982.

GNLE (Grote Nederlandse Larousse Encyclopedie), vol. 24. Hasselt: Heideland-Orbis/’s Gravenhage:Scheltens & Giltay, 1979.

Gödel, K., "Russell’s mathematical logic (1944)", in: K. Gödel,Collected works, Vol II. Publications1938-1974. New York/Oxford: Oxford UP, 1990.

Goodman, N., (1966, 1951) "The question of classes and nominalism", in: R.J. van Iten (ed.),Theproblem of universals. New York: Appleton-Century-Crofts, 1970.

Gordon, R., "The concept of projective identification",J. Analyt. Psychology, (1965)10, 2, 127-150.Goudsmit, A.L., J.H. Mowitz, "Over inhoud en proces in psychotherapie-onderzoek. Een methode van

onderzoek naar therapeutisch handelen in ontdekkende psychotherapieën",Handelingen, (1987)1,4.

Goudsmit, A.L., "Liefdewerk", column ’De ziel op tafel’,Diemensies, Vakgroep Psychologie,Rijksuniversiteit Groningen, january 1988.

Goudsmit, A.L., "Organizational closure, the process of psychotherapy and the psychologist’s fallacy", in:G.J. Dalenoort (ed.),The paradigm of selforganization. New York: Gordon & Breach, 1989 (a).

Goudsmit, A.L., "By way of introduction: can we make a non-classical psychology?", in: A.L. Goudsmit(ed.),Self-organization in psychotherapy. Demarcations of a new perspective. NewYork/Heidelberg: Springer Verlag, 1989 (b).

Goudsmit, A.L., "On blindness and incomprehension", in: A.L. Goudsmit (ed.)Self-organization inpsychotherapy; demarcations of a new perspective. New York/Heidelberg: Springer, 1989 (c).

Goudsmit, A.L., "The black hole of psychotherapy research. Organizational closure in psychotherapeuticprocesses.", in: F. Heylighen, E. Rosseel, F. Demeyere (eds.),Self-steering and cognition incomplex systems. Toward a new cybernetics. New York: Gordon & Breach, 1990.

Goudsmit, A.L., "Consensualiteit bij autonome systemen: over autopoiesis, taal en psychoterapie".Systeemteoretisch Bulletin, (1991)9, 3, 154-193.

Goudsmit, A.L., "A one-sided boundary: on the limits of knowing organizational closure", in: G. van deVijver (ed.), New perspectives on cybernetics; self-organization, autonomy, connectionism.Dordrecht: Kluwer, 1992 (a).

Goudsmit, A.L., "Psicoterapia e tecnologia: un saggio sull’autoriferimento, sulla logica della scopertapratica e sul superamento di confini ad un solo lato", in: M. Ceruti (ed.),Evoluzione e conoscenza.L’epistemologia genetica di Jean Piaget e le prospettive del costruttivismo. Bergamo: Lubrina,1992 (b).

Goudsmit, A.L., "De l’usage de processus parallèles dans l’examen des processus autonomesthérapeutiques",Thérapie Familiale, (1993) 2, 217-230.

Gram, M.S., "Kant’s duplication problem",Dialectica, (1980)34, 1,17-59.Greenberg, J.R., S.A. Mitchell,Object relations in psychoanalytic theory. Cambridge, Mass.: Harvard UP,

1983.Greenson, R.R., "Empathy and its vicissitudes",Int. J. of Psychoanalysis, (1960)XLI , 418-424.Grinberg, L., "On a specific aspect of countertransference, due to the patient’s projective identification",

Int. J. of Psycho-Analysis, (1962)43, 436-440.Grinberg, L., "Countertransference and projective counteridentification",Contemporary Psychoanalysis,

(1979)15, 226-247.Grosskurth, P.,Melanie Klein: her world and her work. Toronto: McClelland and Stewart, 1986.Grotstein, J.S.,Splitting and projective identification. Northvale, NJ: Aronson, 1981.Grotstein, J.S., "Projective identification reappraised: Part I: projective identification, introjective

215

identification, the transference/countertransference neurosis/psychosis, and their consummateexpression in the crucifixion, the pieta, and "therapeutic exorcism"",Contemporary Psychoanalysis,(1994a)30, 4,708-746.

Grotstein, J.S., "Projective identification reappraised: part II: the countertransference complex",Contemporary Psychoanalysis, (1994b)31, 3, 479-511.

Grotstein, J.S., "The numinous and immanent nature of the psychoanalytic subject",J. of AnalyticalPsychology, (1998a)43, 41-68.

Grotstein, J.S., "Bion’s "Transformation in ’O’" and the concept of the "transcendent position"". Availablefrom http://www.sicap.it/~merciai/bion97.htm. To be published in proceedings of the ’CentennialConference on the work of Bion; Turin, july 16-19, 1997’. London: Karnac, 1998(b). (pagenumbers in the text refer to the website version)

Gustafson, J.P.,The complex secret of brief psychotherapy. New York: Norton, 1986.Gutsch, K.U.,Psychotherapeutic approaches to specific DSM-III-R categories. Springfield, Ill.: Thomas,

1988.Haley, J., "Ideas which handicap therapists", in: M.M. Berger (ed.),Beyond the double bind:

communication and family systems, theories, and techniques with schizophrenics. New York:Brunner/Mazel, 1978.

Hanson, N.R., "Notes toward a logic of discovery", in: R.J. Bernstein (ed.),Perspectives on Peirce. NewHaven: Yale Univ. Press, 1965.

Hark, M. ter,Beyond the inner and the outer. Wittgenstein’s philosophy of psychology. Dordrecht: KluwerAcademic Publishers, 1990.

Harnoncourt, N., "The performance of "Idomeneo"". Text accompanying the CD edition (1980/1984) ofMozart’s Idomeneo. Teldec, 2292-42600-2.

Heidegger, M., (1957)Identität und Differenz. Pfullingen: Neske, 1990. Translated and with an introd. byJoan Stambaugh:Identity and difference. New York, Harper & Row, 1969.

Heidegger, M.,Nietzsche. Pfullingen: Neske, 1961.Heidegger, M., "Letter on humanism" (translated by E. Lohner), in: W. Barrett, H.D. Aiken,Philosophy in

the twentieth century. New York: Random House, 1962.Heidegger, M.,The basic problems of phenomenology. Bloomington: Indiana Univ. Press, 1982.Heimann, P., "On countertransference",Int. J. of Psycho-Analysis, (1950)31, 81-85.Hilgard, E.R.,Theories of learning. New York: Appleton-Century-Crofts, 1956.Hinshelwood, R.D.,A dictionary of Kleinian thought. London: Free Association Bks, 1991.Hintikka, J., "Cogito, ergo sum: inference or performance?",Philosophical Review, (1962)71, 3-32.Hirs, W.M., Standaardclassificaties voor medische en niet-medische gegevens. Een studie naar de

samenhang van objecten, begrippen en termen in de gezondheidszorg. University of Groningen:PhD Thesis, 1987.

Hofstadter, D.R.,Gödel, Escher, Bach: an eternal golden braid. New York: Basic Books, 1979.Holton, G., "The roots of complementarity",Daedalus, (1970)99, 1015-1055.Honner, J., "The transcendental philosophy of Niels Bohr",Stud. Hist. Phil. Sci., (1982)13, 1-30.Hoofdakker, R.H. van den, "Het klinische begrip depressie",Ned. Ts. voor de Psychologie, (1978)33,

221-254.Hoofdakker, R.H. van den, "Changing concepts in psychiatry. On therapy and pathogenesis in biological

psychiatry", in: D. de Cuyper, G. Buyse (eds.),Biological psychiatry. Marching backwards into thefuture (Proceedings of the Xth Forum organized by the Belgian College ofNeuropsychopharmacology and Biological Psychiatry). Leuven: Leuven Univ. Press, 1989.

Howe, R.H., H. von Foerster, "Introductory comments to Francisco Varela’s calculus for self-reference",Int. J. of General Systems, (1975)2, 1-3.

Hume, D., "Enquiry concerning human understanding" (1777a), in: D. Hume,Enquiries concerning humanunderstanding and concerning the principles of morals. Oxford: Clarendon Press, 1975.

Hume, D., "Enquiry concerning the principles of morals" (1777b), in: D. Hume,Enquiries concerninghuman understanding and concerning the principles of morals. Oxford: Clarendon Press, 1975.

Hull, C.L., Principles of behavior. An introduction to behavior theory. New York:Appleton-Century-Crofts, 1943.

Hull, J.W., "Videogames: transitional phenomena in adolescence",Child and Adolescence Social Work J.,(1985)2, 2, 106-113.

216

Husserl, E.,Logische Untersuchungen. Halle: Niemeyer, 1913. Translated by J.N. Findlay:Logicalinvestigations. London and Henley: Routledge & Kegan Paul, 1977.

Husserl, E.,Ideen zu einer reinen Phänomenologie und phänomenologischen Philosophie. Erstes Buch(Ideen I). Halle: Niemeyer, 1922. Translated by F. Kersten:Ideas pertaining to a purephenomenology and to a phenomenological philosophy(first book). The Hague/Boston/London:Nijhoff, 1982.

IJsseling, S.,Mimesis. Over schijn en zijn. Baarn: Ambo, 1990.Ingleby, J.D., "Melanie Klein en de Britse object-relatietheorie", in: C.H.C.J. van Nijnatten (red.),

Psychodynamische ontwikkelingsmodellen, Meppel: Boom, 1992.James, W.,The principles of psychology. New York: Dover (1890), 1950.James, W.,A world of pure experience. [New York? : s.n., 1904]. Also attainable from http://

www.yorku.ca/dept/psych/classics/James/experience.htmJames, W., (1907)Pragmatism. A new name for some old ways of thinking. Cambridge, Mass.: Harvard

UP, 1975.James, W.,Essays in radical empiricism. New York: Longmans, Green & Co., 1912.Jammer, M.,The conceptual development of quantum mechanics. New York: McGraw-Hill, 1966.Jansz, J.,Person, self and moral demands. Individualism contested by collectivism. Leiden, DSWO Press,

1991.Johnson, D.K., "Reclaiming reality: a critique of Maturana’s ontology of the observer",Methodologia,

(1991)V, 9, 7-31.Joseph, B., "Projective identification - some clinical aspects", in: E. Bott Spillius (ed.),Melanie Klein

today. Developments in theory and practice. Volume 1: mainly theory. London/New York:Routledge, 1988.

Jung, C.G., (1946)Die Psychologie der Übertragung. Gesammelte Werke, Band 16. Olten: Walter, 1971.Translated by R.F.C. Hull:Psychology of the transference, in: Collected works, vol. 16: Thepractice of psychotherapy. Essays on the psychology of the transference and other subjects.NewYork: Pantheon Books, 1954.

Kafka, J.S., "The body as transitional object: a psychoanalytic study of a self-mutilating patient",Br. J. ofMedical Psychology, (1969)42, 207-212.

Kalckar, J., "Editor’s introduction/Part II", in: N. Bohr,Collected Works, vol. 6. Amsterdam: North-Holland, 1985.

Kampis, G.,Self-modifying systems in biology and cognitive science. A new framework for dynamics,information and complexity. Oxford: Pergamon, 1991.

Kant, I., Prolegomena to any future metaphysics that will be able to present itself as a science(1783). Atranslation from the German based on the original editions with an Introduction and Notes by PeterG. Lucas. Manchester: Manchester U.P., 1971.

Kaplan, A.,The conduct of inquiry. Methodology for behavioral science. New York: Harper & Row,1964.

Karasu, T.B., A.E. Skodol, "VIth axis for DSM-III: psychodynamic evaluation",Am. J. of Psychiatry,(1980)137, 5, 607-610.

Katz, M.M., "Comment on Blashfield’s article",Schizophrenia Bulletin, (1982)8, 1, 9-11.Kauffman, L.H., F.J. Varela, "Form dynamics",J. Social Biol. Structures, (1980)3, 171-206.Kaye, K., "Thickening thin data: the maternal role in developing communication and language", in: M.

Bullowa (ed.),Before speech: the beginning of interpersonal communication. Cambridge:Cambridge U.P., 1979.

Kendell, R.E.,The role of diagnosis in psychiatry. Oxford: Blackwell Sc. Publ., 1975.Kendell, R.E., "Comment on Blashfield’s article",Schizophrenia Bulletin, (1982)8, 1, 11-12.Kiesler, D.J., "Experimental designs in psychotherapy research", in: A.E. Bergin, S.L. Garfield (eds.),

Handbook of psychotherapy and behavior change. New York: Wiley, 1971.Klein, M., (1946) "Notes on some schizoid mechanisms", in: M. Klein, P. Heimann, S. Isaacs, J. Riviere,

Developments in psycho-analysis. London: Hogarth Press, 1952.Klerman, G.L., "The evolution of a scientific nosology", in: J.C. Shershow (ed.),Schizophrenia: science

and practice. Cambridge, Mass.: Harvard Univ. Press, 1978.Klerman, G.L., "History and development of modern concepts of affective illness", in: R.M. Post, J.C.

Ballenger (eds.),Neurobiology of mood disorders. Baltimore: Williams & Wilkins, 1984.

217

Klerman, G.L., "Historical perspectives on contemporary schools of psychopathology", in: T. Millon, G.L.Klerman (eds.),Contemporary directions in psychopathology. Toward the DSM-IV. New York:Guilford Press, 1986.

Kornwachs, K., "Herrschaft-Macht-System. Zur frage der objektivierung von therapeutischer Interaktion",Zeitschrift f. systemische Therapie, (1986)4, 4, 214-223.

Kraepelin, E.,Psychiatrie. Ein Lehrbuch für Studirende und Aerzte.(sechste, vollständig umgearbeiteteAuflage) Leipzig: Barth, 1899; (fifth edition: Leipzig: Barth, 1896; eighth edition, 1913.

Kulish, N.M., "Projective identification: A concept overburdened",Int. J. of PyschoanalyticPsychotherapy, (1986)11, 79-104.

Kwant, R.C., "Merleau-Ponty’s nieuwe filosofie van de perceptie, de natuur en de logos",Tijdschr. v.Filosofie, (1989)51, 4, 669-695.

Latour, B., "The politics of explanation: an alternative", in: S. Woolgar (ed.),Knowledge and reflexivity.New frontiers in the sociology of knowledge. London: Sage, 1988.

Levenson, E.A.,The fallacy of understanding. An inquiry into the changing structure of psychoanalysis.New York/London: Basic Books, 1972.

Levenson, E., "Show and tell: the recursive order of transference", in: A. Rothstein (ed.),How doestreatment help? On the modes of therapeutic action of psychoanalytic psychotherapy. Madison,Conn.: Int. Univ. Press, 1988.

Levenson, A.E., "The uses of disorder: chaos theory and psychoanalysis",Contemporary Psychoanalysis,(1994)30, 1,5-24.

Lietaer, G., J. Rombauts, R. van Balen (eds.),Client-centered and experiential psychotherapy in thenineties. Leuven: Univ. Pers Leuven, 1990.

Linder, R., "Diagnosis: description or prescription? A case study in the psychology of diagnosis",Perceptual and Motor Skills, (1965)20, 1081-1092.

Lingis, A., "Translator’s preface", in M. Merleau-Ponty,The visible and the invisible. Evanston: North-western U.P., 1968.

Lingis, A., "Being in the interrogative mood", in: G. Gillan (ed.),Horizons of the flesh. Criticalperspectives on the thought of Merleau-Ponty. London & Amsterdam: Feffer & Simons, 1973. [NB:It is obvious that in this title the word ’mood’ has mistakenly been substituted for ’mode’. Theauthor on p. 78, note 3 refers to his own translation of Merleau-Ponty (1968, p. 103), where theterm ’interrogative mode’ is presented.]

Linschoten, J.,Idolen van de psycholoog. Utrecht: Bijleveld, 1964.Löfgren, L., "Towards system: from computation to the phenomenon of language", in: M.E. Carvallo

(ed.),Nature, cognition and system. Dordrecht: Reidel, 1988.Löfgren, L., "Complementarity in systems thinking", in:Proc. European Congress on System Science, part

I. Lausanne: AFCET, 1989.Löfgren, L., "The partiality of self-reference", in: F. Heylighen, E. Rosseel, F. Demeyere (eds.),

Self-steering and cognition in complex systems. Toward a new cybernetics. New York: Gordon &Breach, 1990.

Löfgren, L., "Cybernetics, science and complementarity", in R. Glanville, G. de Zeeuw (eds.),Mutual usesof cybernetics and science. Amsterdam: Thesis Publishers, 1991.

Löfgren, L., "Complementarity in language; toward a general understanding", in: M.E. Carvallo, (ed.),Nature, Cognition, and System II: On complementarity and beyond.Dordrecht, London, Boston:Kluwer, 1992.

London, P.,The modes and morals of psychotherapy. New York: Holt, Rinehart & Winston, 1964.Madison, G.B.,A search for the limits of consciousness. The phenomenology of Merleau-Ponty. Athens,

Ohio: Ohio U.P., 1981.Mahrer, A.R.,Experiential psychotherapy. New York: Brunner-Mazel, 1983.Mahrer, A.R.,The integration of psychotherapies. A guide for practicing therapists. New York: Human

Sciences Press, 1989.Man, P. de, "The rhetoric of blindness: Jacques Derrida’s reading of Rousseau", in: P. de Man,Blindness

and insight. Essays in the rhetoric of contemporary criticism. New York: Oxford U.P., 1971.Manson, R.,The theory of knowledge of Giambattista Vico. Hamden, Conn.: Archon Books, 1969.Margenau, H., "Reality in quantum mechanics",Phil. Sci., (1949)16, 287-302.Maturana, H.R., "The organization of the living: a theory of the living organization",Int. J. Man-Machine

218

Studies, (1975),7, 313-332.Maturana, H.R., "Biology of language: the epistemology of reality", in: G.A. Miller, E. Lenneberg (eds.),

Psychology and biology of language and thought. Essays in honor of Eric Lenneberg: conference,Ithaca, 1976. New York: Academic Press, 1978.

Maturana, H.R., "Autopoiesis: reproduction, heredity and evolution", in: M. Zeleny (ed.),Autopoiesis,dissipative structures and spontaneous social orders. Boulder, Col.: Westview, 1980 (a).

Maturana, H.R., "Man and society", in: F. Benseler, P.M. Hejl, W.K. Köck (eds.),Autopoiesis,communication, and society. The theory of autopoietic systems in the social sciences.Frankfurt/New York: Campus, 1980 (b).

Maturana, H.R., "Repräsentation und Kommunikation", in: H.R. Maturana,Erkennen: die Organisationund Verkörperung von Wirklichkeit. Ausgewählte Arbeiten zur biologischen Epistemologie.Braunschweig-Wiesbaden: Vieweg, 1982.

Maturana, H.R., "What is it to see?",Arch. Biol. Med. Exp., (1983)16, 255-269.Maturana, H.R., "The biological foundations of self-consciousness and the physical domain of existence",

in: E.R. Caianiello (ed.),Physics of cognitive processes. Proceedings of the InternationalSymposium on physics of cognitive processes, Amalfi, Italy, 16-20 june 1986. New Jersey: WorldScientific Press, 1987 (a).

Maturana, H.R., "Everything is said by an observer", in: W.I. Thompson (ed.),Gaia. A way of knowing.Political implications of the new biology. Chicago: Lindisfarne Press, 1987 (b).

Maturana, H.R., "Reality: the search for objectivity or the quest for a compelling argument",Irish Journalof Psychology, (1988)9, 1, 25-82.

Maturana, H.R., F.J. Varela,Autopoiesis and cognition. The realization of the living. Dordrecht/Boston:Reidel, 1980.

Maturana, H.R., F.J. Varela,The tree of knowledge. The biological roots of human understanding.Boston/London: Shambhala, 1987.

McMullin, E., "The conception of science in Galileo’s work", in: R.E. Butts, J.C. Pitt (eds.),Newperspectives on Galileo. Dordrecht/Boston: Reidel, 1978.

Mead, G.H., (1934)Mind, self and society. Chicago: Chicago Univ. Press, (1962).Mayer-Gross, W., E. Slater, M. Roth,Clinical psychiatry. London: Cassell, 1954.Menninger, K.,The vital balance; the life process in mental health and illness. New York: The Viking

Press, 1967 (1963).Merleau-Ponty, M.,Phénoménologie de la perception. Paris: Gallimard, 1945.Merleau-Ponty, M.,Les relations avec autrui chez l’enfant(1re partie). Paris, Centre de Documentation

Universitaire, 1951.Merleau-Ponty, M., "Préface", in: A. Hesnard,l’Oeuvre de Freud et son importance pour le monde

moderne. Paris: Payot, 1960 (a).Merleau-Ponty, M.,Signes.Paris: Gallimard, 1960 (b). Translated (by R.C. McCleary) as:Signs.

Evanston: Northwestern UP, 1964 (b).Merleau-Ponty, M.,Le visible et l’invisible. Paris: Gallimard, 1964 (a). Translated (by A. Lingis) as:The

visible and the invisible. Evanston: North-western U.P., 1968 (a).Merleau-Ponty, M.,Résumés de cours, Collège de France 1952-1960. Paris: Gallimard, 1968 (b).Miller, C.L., "Human Knowledge and God in Cusanus’ De docta ignorantia, Book I". Online publication,

available from http://www.sunysb.edu/philosophy/faculty/lmiller/DDI12.txt. Dept. of Philosophy,State University of New York at Stony Brook, 1997.

Miller, N.E., L. Luborsky, J.P. Barber, J.P. Docherty,Psychodynamic treatment research. A handbook forclinical practice. New York: Basic Books, 1994.

Millon, T., "On the past and future of the DSM-III: personal recollections and projections", in: T. Millon,G.L. Klerman (eds.),Contemporary directions in psychopathology. Toward the DSM-IV. New York:Guilford Press, 1986.

Milner, P.M., Physiological psychology. London: Holt, Rinehart, Winston, 1971.Mingers, J.Self-producing systems: implications and applications of autopoiesis. New York: Plenum

Press, 1995.Minsky, M.L., Computation. Finite and infinite machines. London: Prentice-Hall, 1967.Mowitz, J.H., A.L. Goudsmit, "Organizational closure and morphogenesis", in: G.J. Dalenoort (ed.),The

paradigm of selforganization. New York: Gordon & Breach, 1989.

219

Nagel, E., J.R. Newman,Gödel’s proof. New York: New York U.P., 1958.Neumann, J. von,Mathematische Grundlagen der Quantenmechanik. Berlin: Springer, 1932.NVAGG, "De tien meest gestelde vragen over DSM". Internal memorandum, Ca-nr. 950159 cpr/ak.

Utrecht: NVAGG, Postbus 8400, Utrecht, 1995.Ogden, T.H.,Projective identification and psychotherapeutic technique. New York: Jason Aronson, 1982.Ogden, T.H., (1989)The primitive edge of experience. London: Maresfield Library, 1992 (a).Ogden, T.H., "The dialectically constituted/decentred subject of psychoanalysis. II. The contributions of

Klein and Winnicott",Int. J. of Pyschoanalysis, (1992b)73, 613-626.Ogden, T.H., "The analytic third: working with intersubjective clinical facts",Int. J. of Psychoanalysis,

(1994a)75, 3-20.Ogden, T.H., "The concept of interpretive action",Pyschoanalytic Quarterly, (1994b)LXIII , 219-245.Okot p’Bitek, Song of Lawino. Nairobi: East African Publishing House, 1966.Olthof, J., E. Vermetten,Narratieve strategieën in psychotherapie voor kinderen en volwassenen.

Amersfoort/Leuven: Acco, 1993.Palmer, R.E.,Hermeneutics: interpretation theory in Schleiermacher, Dilthey, Heidegger and Gadamer.

Evanston, Ill.: North-Western U.P., 1969.Pask, G.,Conversation, cognition and learning. A cybernetic theory and methodology. Oxford: Elsevier,

1975.Pattee, H.H., "Laws and constraints, symbols and languages", in: C.H. Waddington (ed.),Towards a

theoretical biology, vol. 4, Edinburgh: Edinburgh University Press, 1972 (a).Pattee, H.H., "The evolution of self-simplifying systems", in: E. Laszlo (ed.),The relevance of general

systems theory. New York: Braziller, 1972 (b).Pattee, H.H., "The complementarity principle and the origin of macromolecular information",BioSystems,

(1979)11, 217-226.Pattee, H.H., Cell psychology: an evolutionary approach to the symbol-matter problem. Cognition and

Brain Theory, (1982)5, 4, 325-41.Pattee, H.H., "The measurement problem in physics, computation, and brain theories", in: M.E. Carvallo

(ed.),Nature, cognition and system II: On complementarity and beyond. Dordrecht, London,Boston: Kluwer, 1992.

Pattee, H.H., "Evolving self-reference: matter, symbols, and semantic closure",Communication &Cognition - Artificial Intelligence, (1995)12, 1-2, 9-27.

Pattee, H.H., "The problem of observables in models of biological organizations". E.L. Khalil, K.E.Boulding (eds.),Evolution, order, and complexity. London: Routledge, 1996. (available also fromURL http://ssie.binghamton.edu/people/pattee_doc/boulding.html).

Paul, G.L., "Strategy of outcome research in psychotherapy",J. of Consulting Psychology, (1967)31, 2,109-118.

Penrose, R.,The emperor’s new mind: concerning computers, minds, and the laws of physics. Oxford:Oxford U.P., 1989.

Perry, S., A. Frances, J.F. Clarkin,A DSM-III-R casebook of treatment selection. New York:Brunner/Mazel, 1990.

Peterfreund, E.,The process of psychoanalytic therapy. Models and strategies. New York: AcademicPress, 1983.

Pierloot, R.A., J.H. Thiel,Psychoanalytische therapieën. Deventer: Van Loghum Slaterus, 1986.Pitt, J.C., "Galileo: causation and the use of geometry", in: R.E. Butts, J.C. Pitt (eds.),New perspectives

on Galileo. Dordrecht/Boston: Reidel, 1978.Pitt, J.C.,Galileo, human knowledge, and the book of nature. Method replaces metaphysics. Dordrecht:

Kluwer, 1992.Plooij, F., "How wild chimpanzee babies trigger the onset of mother-infant play", in: M. Bullowa (ed.),

Before speech: the beginning of interpersonal communication. Cambridge: Cambridge U.P., 1979.Polanyi, M.,Personal knowledge. Towards a post-critical philosophy. London: Routledge & Kegan Paul,

1958.Polanyi, M.,The tacit dimension. New York: Doubleday, 1966.Popper, K.R.,The logic of scientific discovery. London: Hutchinson, 1974.Praag, H.M. van, "Two-tier diagnosing in psychiatry",Psychiatry Research, (1990)34, 1-11.Praag, H.M. van, G.M. Asnis, R.S. Kahn, S.L. Brown, M. Korn, J.M.H. Friedman, S. Wetzler,

220

"Monoamines and abnormal behaviour. A multi-aminergic perspective",Br. J. of Psychiatry, (1990)157, 723-734.

Quine, W.V.O., "Two dogmas of empiricism", in: W.V.O. Quine,From a logical point of view.Cambridge, Mass.: Harvard UP, 1953.

Quine, W.V.O., "Epistemology naturalized", in: W.V.O. Quine,Ontological relativity and other essays.New York/London: Columbia UP, 1969.

Quine, W.V.O., "On what there is", in: R.J. van Iten (ed.),The problem of universals. New York:Appleton-Century-Crofts, 1970.

Racker, H., "The meanings and uses of countertransference",Psychoanalytic Quarterly, (1957)26,303-357.

Rashevsky, N., "Topology and life",Bulletin Math. Biophysics, (1954)16, 317-348.Reichenbach, H.,Experience and prediction. An analysis of the foundations and the structure of

knowledge. Chicago: Univ. of Chicago Pr., 1938.Reichenbach, H.,The rise of scientific philosophy. Berkeley: Univ. of Calif. Pr., 1951.Reid, W.H., M.G. Wise,DSM-III-R training guide. For use with the American Psychiatric Association’s

Diagnostic and Statistical Manual of Mental Disorders (third edition-revised). New York:Brunner/Mazel, 1989.

Reid, W.H., M.G. Wise,DSM-IV training guide. New York: Brunner/Mazel, 1995.Reik, Th.,Listening with the third ear. The inner experience of a psychoanalyst. New York: Grove Press,

1948.Rice, L.N., "The evocative function of the therapist", in: D.A. Wexler, L. North Rice (eds.),Innovations

in client-centered therapy. New York: Wiley, 1974.Rijk, L.M. de, Middeleeuwse wijsbegeerte. Traditie en vernieuwing. Assen: Van Gorcum, 1981 (a).Rijk, L.M. de, Pierre Abélard (1079-1142). Scherpzinnigheid als hartstocht. Amsterdam: North-Holland,

1981 (b).Rijk, L.M. de, "John Buridan on universals",Revue de Métaphysique et de Morale, (1992) 1, 35-59.Rioch, M.J., "The work of Wilfred Bion on groups", in: J.C. Sager, H.S. Kaplan (eds.),Progress in group

and family therapy. New York: Brunner/Mazel, 1972.Rogers, C.R., B.F. Skinner, "Some issues concerning the control of human behavior",Science(1956),124,

3231, 1057-1066.Roosmalen-Wiebenga, M.W. van,Nutrition rehabilitation in the south western highlands of Tanzania; a

two way learning process. Den Haag: Pasmans, 1988.Rorty, R.,Philosophy and the mirror of nature. Princeton, New Jersey: Princeton U.P., 1979.Rosen, R., "Church’s thesis and its relation to the concept of realizability in biology and physics",Bull. of

Mathematical Biophysics, (1962)24, 375-393.Rosen, R., "On anticipatory systems: I. When can a system contain a predictive model of another?",J.

Social Biol. Structures, (1978)1, 155-162.Rosen, R., "Organisms as causal systems which are not mechanisms: an essay into the nature of

complexity", in: R. Rosen (ed.),Theoretical biology and complexity. Orlando: Academic Press,1985.

Rosen, R.,Life itself. A comprehensive inquiry into the nature, origin and fabrication of life. New York:Columbia Univ. Press, 1991.

Ruijgt, M., "Het ijkpunt. Vrouwelijke deskundigen over de DSM-III-R". Unpublished report, University ofMaastricht, dept. of Psychiatry, 1991.

Ryle, G.,The concept of mind. Harmondsworth: Penguin, 1973 (1949).Sachs, D.M., S.H. Shapiro, "On parallel processes in therapy and teaching",Psychoanalytic Quarterly,

(1976)45, 394-415.Sandler, J., "De communicatie van patiënt naar analyticus: niet alles is projectieve identificatie",

Psychotherapie, (1995) 2, 59-81. (translation of: "On communication from patient to analyst: noteverything is projective identification",Int. J. of Psychoanalysis, 74, 1097-1107)

Sanford, E.C., "Personal equation",Am. J. of Psychology, (1888)2, 3-38, 271-298, 403-430.Sarbin, T.R., R. Taft, D.E. Bailey,Clinical inference and cognitive theory. New York: Holt, Rinehart &

Winston, 1960.Sartre, J.P.,L’être et le néant. Essay d’ontologie phénoménologique. Paris: Gallimard, 1943.Sartre, J.P., (1936)La transcendance de l’ego. Esquisse d’une description phénoménologique. Paris: Vrin,

221

1966.Sassen, F.,Wijsgerig denken in de middeleeuwen. Haarlem: Bohn, 1965.Schafer, R.,A new language for psychoanalysis. New Haven: Yale University Press, 1976.Schlesinger-Vaccaro, R., "A relationship between Piaget’s notion of conservation and the transference

phenomenon: a clinical example",The Psychoanalytic Review, (1983)70, 1,83-94.Schulte, D., R. Künzel, G. Pepping, Th. Schulte-Bahrenberg, "Tailor-made versus standardized therapy of

phobic patients",Adv. Behav. Res. Ther., (1992)14, 67-92.Schutz, A., (1932)The phenomenology of the social world. London: Heinemann, 1972.Schwartz-Salant, N., "Archetypal foundations of projective identification",J. of Anal. Psychology, (1988),

33 ,1, 39-64.Searles, H.F., (1955) "The informational value of the supervisor’s emotional experiences", in: H. Searles,

Collected papers on schizophrenia and related subjects. New York: Int. Univ. Press, 1965.Segal, H.,Introduction to the work of Melanie Klein. London: The Hogarth Press, 1975.Segal, H.,The work of Hanna Segal. New York: Aronson, 1981.Sheets-Johnstone, M., "Existential fit and evolutionary continuities",Synthese, (1986)66, 2,219-248.Shotter, J.,Social accountability and selfhood. Oxford: Basil Blackwell, 1984.Shotter, J., "Accounting for place and space",Environment & Planning D: Society & Space, (1985)

3,447-460.Shotter, J., "Social accountability and the linguistic creation of mind: the unaccountability of computer

conduct", in: K. Plunkett, S.F. Larsen (eds.),Computers, mind and metaphors. Aarhus/EnglewoodCliffs, NJ: Aarhus U.P./Erlbaum, 1987 (a).

Shotter, J., "Vico: the role of rhetoric and the concept of the imaginary in the making of social life".Paper given at the meeting Problems of Meaning in Modern Psychology, at ManchesterPolytechnic, March 14th, 1987 (b).

Shotter, J., "The role of the imaginary in the making of social life", appeared as: "El papel de loimaginario en la construccion de la vida social", in: T. Ibanez (ed.),El conocimiento de la realidadsocial. Barcelona: Sendai Ediciones, 1989. (page numbers in the text refer to the English draftversion)

Shotter, J., "’Getting in touch’: the metamethodology of a postmodern science of mental life",TheHumanistic Psychologist, (1990)18, 7-22.

Shotter, J., "A poetics of relational forms: the sociality of everyday social life",Cultural dynamics, specialissue on "Evolutionary models in the social sciences" (T. Ingold, ed.) (1991)4, 379-396.

Shotter, J.,Cultural politics of everyday life. Social constructionism, rhetoric and knowing of the thirdkind. Toronto: U. of Toronto Press, 1993.

Skelton, R.M., "Bion’s use of modern logic",Int. J. Psycho-Analysis, (1995)76, 389-397.Skodol, A.E.,Problems in differential diagnosis: from DSM-III to DSM-III-R in clinical practice.

Washington DC: Am. Psychiatric Pr., 1989.Spillius, E.B., "Some developments from the work of Melanie Klein",Int. J. of Psychoanalysis, (1983)64,

321-332.Spillius, E. Bott (ed.),Melanie Klein today. Developments in theory and practice. Volume 1: mainly

theory. London/New York: Routledge, 1988 (a).Spillius, E. Bott (ed.),Melanie Klein today. Developments in theory and practice. Volume 2: mainly

practice. London/New York: Routledge, 1988 (b).Spillius, E. Bott, (1988c) "Introduction to part III", in: E. Bott Spillius (ed.),Melanie Klein today.

Developments in theory and practice. Volume 1: mainly theory. London/New York: Routledge,1988.

Spillius, E. Bott, "Clinical experiences of projective identification", in: R. Anderson (ed.),Clinicallectures on Klein and Bion. London/New York: Tavistock/Routledge, 1992.

Spitzer, R.L., "Letter to the editor",Schizophrenia Bulletin, (1982)8, 4, 592.Spitzer, R.L., J.B.W. Williams, M. Gibbon, M.B. First,Structured Clinical Interview for

DSM-III-R-Personality Disorders (SCID-II, Version 1.0). Washington DC: Am. Psychiatric Press,1990.

Stegge, H., M. Meerum Terwogt, "Experimentele emotie-inductie",De Psycholoog, (1993)28, 6, 237-244.Stigler, S.M.,The history of statistics. The measurement of uncertainty before 1900. Cambridge,

Mass./London: Belknap Harvard UP, 1986.

222

Strasser, S., "Lichamelijkheid in viervoud. Het wijsgerig perspectief van Husserl, Sartre, Lacan enMerleau-Ponty", in: J.J.L. Derksen, E.L.H.M. van de Loo (red.),Het lichamelijk bestaan. Aspectenvan de lichamelijkheid in de klinische psychologie. Baarn: Nelissen, 1985.

Strauss, J.S., "Comment on Blashfield’s article",Schizophrenia Bulletin, (1982)8, 1, 8-9.Strien, P.J. van, "Naar een methodologie van het praktijkdenken in de sociale wetenschappen",Nederlands

Tijdschrift voor de Psychologie, (1975)30, 7, 601-619.Strien, P.J. van,Praktijk als wetenschap. Methodologie van het sociaal-wetenschappelijk handelen. Assen:

van Gorcum, 1986.Strien, P.J. van, "The historical practice of theory construction", in: J.F.H. van Rappard, P.J. van Strien,

L. Mos, W.J. Baker (eds.),Annals of Theoretical Psychology, vol. 8. New York: Plenum, 1993.Strien, P.J. van, "Towards a methodology of psychological practice".Theory & Psychology, 7, 683-700.Symington, J., N. Symington,The clinical thinking of Wilfred Bion. London & New York: Routledge,

1996.Tansey, M.J., W.F. Burke, "Projective identification and the empathic process: interactional

communications",Contemporary Psychoanalysis, (1985)21, 1, 42-69.Thorpe, M., "Psycho-diagnostics in a Xhosa Zionist Church". Submitted in partial fulfilment of the

requirements for Master of Arts, May Dept. of Psychology, Rhodes University (South-Africa),December 1982.

Toulmin, S.E., "The inwardness of mental life",Critical Inquiry, (1979)6, 1, 1-16.Toulmin, S.E., "Self psychology as a "postmodern" science",Psychoanalytic Inquiry, (1986)6, 459-477.Toulmin, S.E., "De grenzen vervagen. Praktische en theoretische wetenschap heben elkaar nodig" (lecture

given at Studium Generale Groningen, october 14, 1992),Intermediair, (1992)28, 49,35-41.Üxküll, J. von,Umwelt und Innenwelt der Tiere. Berlin: Springer, 1909.Varela, F.J., "A calculus for self-reference",Int. J. of General Systems, (1975)2, 5-24.Varela, F.J.,Principles of biological autonomy. New York: North-Holland, 1979.Varela, F.J., "l’Auto-organisation: de l’apparence au mécanisme", in: P. Dumouchel, J.P. Dupuy (eds.),

l’Auto-organisation: de la physique au politique. Paris: Seuil, 1983.Varela, F.J., "Das Gehen ist der Weg", in: R. Kakuska (Hrsg.),Andere Wirklichkeiten. München:

Dianus-Trikont, 1984.Varela, F.J., "Structural coupling and the origin of meaning in a simple cellular automata", in: E. Secarz,

F. Calada, N.A. Mitchinson, T. Tada (eds.),The semiotics of cellular communications in theimmune system. New York: Springer Verlag, 1987.

Varela, F.J.,Science and technology of cognition/Les sciences cognitives: tendences et perspectivesactuelles. Paris: Seuil, 1989.

Varela, F.J., "Organism, cognitive science and the emergence of selfless selves",Revue européenne dessciences sociales, (1991)XXIX , 89, 173-198.

Varela, F.J., E. Thompson, E. Rosch,The embodied mind. Cognitive science and embodied experience.Cambridge, Mass.: MIT Press, 1991.

Verden-Zöller, G.,Mother child play: the biological foundation of self and social consciousness. Passau:unpublished manuscript, Forschungsstelle f. die Ökopsychologie der frühen Kindheit, 1989.Appeared in: H.R. Maturana, G. Verden-Zöller,Amor y juego. Santiago, Chile: Instituto de TerapiaCognitiva, 1993.

Verden-Zöller, G., H.R. Maturana, "Play: the neglected path",Delfin, (1989) 12,1.Verene, D.P.,Vico’s science of imagination. Ithaca/London: Cornell U.P., 1981.Vico, G.B., [The First New Science] The principles of a new science of the nature of nations leading to

the discovery of the principles of a new system of the natural law of the gentes.(1725), in: G.B.Vico, Selected writings/ ed. and transl. by Leon Pompa. Cambridge [etc.]: Cambridge UniversityPress, 1982.

Vico, G.B., The New Science. Unabridged translation of the Third Edition (1744) with the addition of"Practic of the New Science". Translated by T.G. Bergin & M.H. Fisch. Ithaca/London: CornellUP, 1986.

Vijver, G., Doelgerichtheid in cybernetica, connectionisme en cognitivisme. Tussen twee epistemologischeopties: naturalisme en constructivisme. Ghent: Rijksuniversiteit Gent, doctoral thesis, 1989.

Volkan, V.D., The need to have enemies and allies. Northvale, NJ: Aronson, 1988.Waddington, C.H., "Stabilization in systems - chreods and epigenetic landscapes",Futures, (1977)9, 2,

223

139-146.Walker, L.E.A., "Inadequacies of the masochistic personality disorder diagnosis for women",Journal of

Personality Disorders, (1987) 1, 183-187.Wallace, E.R., "What is "truth"? some philosophical contributions to psychiatric issues",Am. J.

Psychiatry, (1988),145, 2, 137-147.Wallace, W.A., "Galileo and reasoningex suppositione: the methodology of theTwo new sciences", in:

R.S. Cohen, C.A. Hooker, A.C. Michalos, J.W. van Evra (eds.),PSA 1974. Proceedings of the 1974biennial meeting; Philosophy of Science Association. Boston studies in the philosophy of science,vol. XXXII. Dordrecht/Boston: Reidel, 1976.

Wallace, W.A., "Galileo and the Doctores Parisienses", in: R.E. Butts, J.C. Pitt (eds.),New perspectiveson Galileo. Dordrecht/Boston: Reidel, 1978.

Wang, H.,Reflections on Kurt Gödel. Cambridge, Mass.: MIT Press, 1988.Watzlawick, P. (ed.),The invented reality. How do we know what we believe we know? Contributions to

constructivism. New York: Norton, 1984.Watzlawick, P., J.H. Beavin, D.D. Jackson,Pragmatics of human communication. New York: Norton,

1967.Weinsheimer, J.C., Gadamer’s hermeneutics: a reading of "Truth and Method". New Haven, Conn.: Yale

UP, 1985.Weizsäcker, C.F. von, "Komplementarität und Logik",Die Naturwissenschaften, (1955)42, 521-529,

546-555.Wexler, A., "A cognitive theory of experiencing, self-actualization, and therapeutic process", in: D.A.

Wexler, L. North Rice (eds.),Innovations in client-centered therapy. New York: Wiley, 1974.Whitaker, C., "The hindrance of theory in clinical work", in: P.J. Guerin (ed.),Family therapy: theory and

practice. New York: Gardner Press, 1976.Whitehead, A.N., B. Russell.Principia Mathematica. Cambridge: Cambridge UP, 1910-1913.Wilson, M., "DSM-III and the transformation of American psychiatry: a history",Am. J. of Psychiatry,

(1993)150, 3,399-410.Winnicott, D.W., (1951) "Transitional objects and transitional phenomena", in: D.W. Winnicott,Playing

and reality. Harmondsworth: Penguin, 1980.Winograd, T., "What does it mean to understand language?", in: D.A. Norman (ed.),Perspectives on

cognitive science: 1st Annual meeting of the Cognitive Science Society, La Jolla 1979. Hillsdale:Erlbaum, 1981.

Wittgenstein, L.,Tractatus logico-philosophicus. Frankfurt: Fischer, 1959/ London: Routledge & KeganPaul, 1963 (translation by D.F. Pears & B.F. McGuinness).

Wittgenstein, L.,Philosophische Untersuchungen. Frankfurt: Suhrkamp, 1971 (1958) / (translation byG.E.M. Anscombe)Philosophical Investigations. Oxford: Blackwell, 1976.

Young, R.M., "Benign and virulent projective identification in groups and institutions", 1992. http://www.shef.ac.uk/uni/academic/N-Q/psysc/staff/rmyoung/papers/paper3.html

Zimring, F.M., "Theory and practice of client-centered therapy", in: D.A. Wexler, L. North Rice (eds.),Innovations in client-centered therapy. New York: Wiley, 1974.

Zolo, D., "Autopoiesis. Critica di un paradigma conservatore",MicroMega, (1986) 1, 129-173.

Abstract

This book deals with the problem why it is so difficult to understand what happens in psychotherapeuticprocesses. The answer that I propose is that many psychotherapeutic events have a certain self-reference.Not only many of the therapeutic conversations pertain to the ways in which they are performed, but alsomany of the exploration processes are self-referential. It is this self-reference that impedes astraightforward empirical access and that compels us to use other techniques, some of which are in factcommon practice in supervision. The book investigates how a detached attitude, common in regularscientific research, does not advance the desired clarity of understanding.

Chapter 1 gives a general introduction. It presents some underlying ideas and it summarizes the variouschapters of the book.

In chapter 2 I introduce some concepts from Maturana’s theory of autopoietic systems. Of interest is herethe notion of autopoiesis, the ’organizational closure’ of living beings. This is a concept at the edge ofmetaphysics, in that it is defined in a way that disables immediate empirical confirmation. One cannot seethe closure closed; one has to force an opening in order to see anything. I pay attention to the idea thatautopoietic systems are capable of making distinctions as a necessary effect of their organizations. Thus,according to Maturana, making distinctions is essential to the process of staying alive. The concept ofdistinction is a major issue throughout this book.

Maturana’s ideas on how it is possible that objects are perceived by language users is taken here as apoint of departure for the elaboration in chapter 3 of some concepts beyond the theory of Maturana. Inparticular, I relate Maturana’s concept of distinction to what Merleau-Ponty wrote on the ’flesh’, which isa kind of ontological substratum for both subject and object; it is that from which both object and subjectdevelop. Furthermore, I relate Maturana’s ideas on the perception of objects to the ideas of GiambattistaVico on the poetic language of the earliest peoples. The objects thus perceived arise as a commonconstruct, a social artifact for the coordination of diverging experiences between persons. They serve as anelementary mode of avoiding conflict. I call these objects ’naive’: they do not distinguish between thatwhich is perceived and how it is perceived.

This latter distinction is put into the focus in chapter 4. I call the usage of language that has thisdistinction ’critical’. I specify two critical discourses: both are used to temper conflict and find agreementbetween persons by constructing some stable ground, to which diverging qualities can be attributed. In thefirst discourse, which I call realist, the stable ground is some abstract essence, to which concrete, butdiverging, appearances are attributed; in the second discourse, which I call nominalist, the stable ground isa concrete symbol, to which abstract, but diverging meanings are assigned.

Both discourses can be used, in radically different ways, to speak about mental entities (called ’internalobjects’), both in other persons and inoneself. This is the subject matter of chapter 5. These radicallydifferent ways are elaborated further in chapter 7.

But both discourses, when applied to the world outside oneself, could become integrated into what I call’classical science’ (chapter 6). However, this was possible only as long as the issue of self-reference didnot receive the attention of scientists. I argue also in that chapter that,with self-reference having becomean issue of interest (first in mathematics and physics), the integration between the two critical discoursesdid no longer hold.

In chapter 7, then, I elaborate how, with respect to self-reference, the two critical discourses take entirelydifferent positions, from which self-reference also looks entirely different. Nevertheless, each discourseshows a serious lacuna; neither can fully express the concept of self-reference, and I argue that bothlacunae in fact converge upon one single domain. In that chapter I introduce my concept of ’one-sided

225

boundary’, which is to be understood as a boundary to the domain of its own definition: beyond thatdomain the boundary does not exist! This is precisely the matter with the ’vagueness’ of psychotherapy: itbrings the external observer (and also the internal one, the therapist) into a position of not (or no longer)understanding. Hence, I maintain that we may appreciate our failures to understand as properties of thatwhich we are trying to grasp. This is the negative way, and I propose to apply it here in the tradition ofCusanus and of Merleau-Ponty, whose concept of ’flesh’ returns here in terms of a ’true negative’:existing as an absence to what we can perceive, not as some transcendental entity that is positively presentelsewhere. At this place in the book we are at the vertex of my argument. The chapters that follow aremeant to utilize the ideas built up thus far.

First, chapter 8 is to illustrate how in the field of psychiatric diagnosis the opposition between realist andnominalist discourses has led to different conceptions of mental disorder, one more interested inexplanations and past developments, and one more interested in future developments and treatment policy.It is maintained how incompatible these conceptions are if the observer’s (diagnostician’s) own personalcommitment is not involved. It is through making naive objects that the diagnostician is capable tointegrate the two critical discourses. Nevertheless, naive objects are not always sufficiently powerful. Inparticular, they fail when the practitioner’s very capacity to think is affected. When this happens, we arein the realm of projective identification, which is the stuff of chapter 11. The chapters that follow are toprepare that.

Next, in chapter 9, I present the construction of naive objects in terms of self-referential explorationprocesses. It is maintained that the task of exploration in psychotherapeutic sessions cannot be performedas a technological procedure. This is argued by means of Gadamer’s concept of ’phronesis’ (prudentaction), which is to be seen as a kind of intermediary between general rules, or general knowledge, andunique individual situations; and by means of Gadamer’s ideas on ’playing’. According to Gadamer, in theperformance of a piece of art (e.g. theatre, music) there is a particular collapse of (what I called) criticaldiscourse; the observer becomes naive, and can no longer distinguish the piece of art from the way inwhich it is performed, nor from the way the artist had meant it to be. Gadamer calls this the aestheticnon-distinction. Transposing this idea to the field of psychotherapy, we find that the processes ofexploration are qualified by a similar kind of aesthetic non-distinction. Exploration goes by means ofnaive objects, and is self-referential. This cannot be replaced by a procedure, applied by a ’behaviortechnologist’ or ’behavior engineer’. It would eliminate the very aspect of exploration.

In chapter 10, then, I proceed by giving some case material from my own experiences with psychotherapyresearch that attempts to do justice to the self-referential and closed qualities of therapeutic systems. Thischapter is meant to clarify and illustrate some of the phenomena of getting trapped in a closed system, aswell how the observer’s or researcher’s absence of understanding and its exploration can be made useful.

In chapter 11 I discuss the concept of ’projective identification’. Unlike the self-reference discussed inchapter 9, here we encounter situations in which a decrease of the therapist’s (as well as the externalobserver’s) capacity to clearly think is a property of the thing to be observed! Whereas in processes ofexploration critical perception was impossible, in favor of naive perception that could flourish in thetherapist and the patient, matters are different here. In processes of projective identification naiveperception and, with it, exploration are halted and can be resumed only to the extent the therapist (orobserver) is capable of enduring (’containing’) some of the negative experiences that the patient isincapable to endure. These negative experiences are of an ’undigested’ quality, they are ’raw experiences’,like radical atoms ungraspable without changing and binding them. It is here that the negative way, theexploration of one’s incapacity to fully understand, shows to be most fertile.

Chapter 12 gives some final considerations on the self-reference of psychotherapeutic processes and onthe impossibility to catch it in positive terms.

Samenvatting

Dit boek gaat over de vraag waarom het zo moeilijk is om te begrijpen wat er gebeurt inpsychotherapeutische processen. Het antwoord dat ik voorstel is dat veel psychotherapeutischegebeurtenissen gekenmerkt worden door een zekere zelf-verwijzing. Niet alleen gaan veel therapeutischegesprekken over de manier waarop ze gevoerd worden, maar ook zijn veel van de exploratieprocessenzelf-verwijzend. Juist deze zelf-verwijzing belemmert een rechtstreekse empirische toegang en dwingt onsom andere technieken te gebruiken, waarvan sommige gangbaar zijn in supervisies. Het boek onderzoekthoe de objectiverende houding die gebruikelijk is binnen wetenschappelijk onderzoek, hier niet bijdraagttot de gewenste helderheid van begrip.

Hoofdstuk 1 geeft een algemene inleiding. Enkele achterliggende gedachten worden gepresenteerd en dehoofdstukken van dit boek worden kort samengevat.

In hoofdstuk 2 introduceer ik enkele begrippen uit Maturana’s theorie van autopoietische systemen. Vanbelang is hier het begrip autopoiese, de ’organisationele sluiting’ van levende wezens. Dit begrip bevindtzich op de rand van de metafysica omdat het gedefinieerd is op een wijze die een rechtstreekse empirischebevestiging ervan onmogelijk maakt. Men kan de sluiting niet zien in gesloten toestand; men moet eenopening forceren om iets te kunnen zien. Ik besteed aandacht aan het idee dat autopoietische systemen instaat zijn onderscheidingen te maken, hetgeen een noodzakelijk onderdeel is van hun organisatie. VolgensMaturana is het maken van onderscheidingen essentieel om in leven te blijven. Het begrip’onderscheiding’ keert voortdurend terug in dit boek als thema.

Uitgaande van Maturana’s ideeën over hoe objecten kunnen worden waargenomen door taalgebruikers,worden in hoofdstuk 3 enkele begrippen uitgewerkt die buiten het kader van Maturana’s theorie vallen. Inhet bijzonder breng ik Maturana’s begrip ’onderscheiding’ in verband met hetgeen Merleau-Ponty schreefover het ’vlees’, hetgeen een soort ontologisch substraat is voor zowel object als subject; het is datgenewaaruit zowel object en subject zich ontwikkelen. Bovendien breng ik Maturana’s ideeën over dewaarneming van objecten in verband met de ideeën van Giambattista Vico over de poëtische taal van devroegste volken. De objecten die zo worden waargenomen ontstaan als een gezamenlijk construct, eensociaal artefact ten behoeve van het coördineren van ervaringen die divergeren tussen personen. Ze dienenals een elementaire manier om conflicten te vermijden. Ik noem deze objecten ’naïef’: ze maken geenonderscheid tussen datgene wat wordt waargenomen en de manier waarop het wordt waargenomen.

Het laatstgenoemde onderscheid staat centraal in hoofdstuk 4. Ik noem het taalgebruik dat dit onderscheidmaakt ’kritisch’. Ik specificeer twee kritische discoursen: beide worden gebruikt om conflicten tereguleren en om overeenstemming te bevorderen tussen personen, doordat er een stabiele basis wordtgecreëerd waaraan niet met elkaar strokende kwaliteiten kunnen worden toegekend. In het eerste discours,dat ik realistisch noem, is de stabiele basis een abstracte essentie, waaraan concrete, doch divergerende,verschijnselen kunnen worden toegekend; in het tweede discours, dat ik nominalistisch noem, is destabiele basis een concreet symbool, waaraan abstracte, doch divergerende, betekenissen kunnen wordentoegekend.

Beide discoursen kunnen op geheel verschillende, uiteenlopende wijzen worden gebruikt, om te sprekenover psychische entiteiten (’interne objecten’ genaamd), zowel in anderen als in onszelf. Dit is hetonderwerp van hoofdstuk 5. Deze radicaal andere wijzen worden verder uitgewerkt in hoofdstuk 7.

Wanneer beide discoursen worden toegepast op de wereld buiten onszelf, dan zouden ze echter kunnenworden geïntegreerd tot wat ik noem ’klassieke wetenschap’ (hoofdstuk 6). Dit was echter slechtsmogelijk zolang wetenschappers zich niet bezighielden met zelf-verwijzing. In dat hoofdstuk laat ik danook zien dat, toen zelf-verwijzing eenmaal in de belangstelling kwam te staan (het eerst in de wiskunde ende natuurkunde), de integratie tussen de twee kritische discoursen niet langer stand hield.

227

In hoofdstuk 7 geef ik vervolgens aan hoe de twee kritische discoursen geheel andere posities innemen tenaanzien van zelf-verwijzing, waardoor zelf-verwijzing een geheel verschillend aanzien krijgt. Nietteminvertoont elke discours een belangrijke lacune; geen van beide kan het begrip zelf-verwijzing vollediguitdrukken, en ik laat zien dat beide lacunes in feite samenkomen in een enkel gebied. In datzelfdehoofdstuk introduceer ik mijn begrip ’eenzijdige grens’, dat een grens is van het domein van zijn eigendefinitie: buiten dat domein bestaat deze grens niet! Dit is precies het geval bij de ’vaagheid’ vanpsychotherapie: het brengt de externe waarnemer (alsook de interne waarnemer, de psychotherapeut) ineen positie van niet (meer) begrijpen. Op basis daarvan beweer ik dat we onze mislukkende pogingen omte begrijpen anders kunnen waarderen, en wel als eigenschappen van datgene waarop we trachten greep tekrijgen. Dit is de negatieve weg, en mijn voorstel is om deze hier toe te passen in de traditie van Cusanusen van Merleau-Ponty: als iets dat bestaat als afwezigheid van hetgeen we kunnen waarnemen, niet alseen transcendentaal ding dat elders op een positieve manier aanwezig is. Op deze plaats in het boekbevinden we ons op het scharnierpunt van mijn redenering. De hierop volgende hoofdstukken beogen detot dusver opgebouwde ideeën te gebruiken.

Om te beginnen laat ik in hoofdstuk 8 met betrekking tot psychiatrische diagnostiek zien hoe detegenstelling tussen realistische en nominalistische discoursen heeft geleid tot twee uiteenlopendeopvattingen over psychische stoornissen: een opvatting die meer georiënteerd is op verklaringen envoorgeschiedenis, en een opvatting die meer georiënteerd is op toekomstige ontwikkelingen enbehandelbeleid. Deze opvattingen zijn onverenigbaar met elkaar als de persoonlijke betrokkenheid van dewaarnemer/diagnosticus buiten spel blijft. Doordat de diagnosticus naïeve objecten construeert kan hij detwee kritische discoursen met elkaar integreren. Maar naïeve object niet altijd toereikend. Ze schietenvooral tekort wanneer het vermogen van de hulpverlener om te denken wordt aangetast. We bevinden onsdan in het rijk van de projectieve identificatie, waarmee we ons in hoofdstuk 11 speciaal zullenbezighouden. De volgende hoofdstukken dienen ervoor om dat voor te bereiden.

Vervolgens behandel ik in hoofdstuk 9 de constructie van naïeve objecten in termen van zelf-verwijzendeprocessen van exploratie. Ik stel dat het exploreren in psychotherapeutische zittingen niet mogelijk is in devorm van een technologische procedure. Dit wordt beargumenteerd met behulp van Gadamer’s begrip’phronese’ (verstandig handelen), dat begrepen kan worden als een soort intermediair tussen algemeneregels of algemene kennis enerzijds, en anderzijds unieke individuele situaties; alsook met behulp vanGadamer’s gedachten over ’spelen’. Volgens Gadamer verdwijnt er bij het uitvoeren van een kunstwerk(bijv. toneel, muziek) het kritische discours (zoals ik dat noemde); de waarnemer wordt naïef, en kan hetkunstwerk niet langer onderscheiden van de manier waarop het wordt uitgevoerd, noch van de manierwaarop de kunstenaar het had bedoeld. Gadamer noemt dit de esthetische niet-onderscheiding. Toegepastop het terrein van psychotherapie, betekent dit dat de exploratieprocessen gekenmerkt worden dooreenzelfde soort esthetische niet-onderscheiding. Exploratie vindt plaats met behulp van naïeve objecten enis zelfverwijzend. Dit kan niet worden vervangen door een procedure, die dan zou worden toegepast dooreen ’gedragstechnoloog’ of een ’gedrags-ingenieur’. Het zou de act van exploreren in zijn kern tenietdoen.

In hoofdstuk 10 presenteer ik vervolgens casuïstisch materiaal op basis van mijn eigen ervaringen metpsychotherapieonderzoek, waarin getracht wordt recht te doen aan de zelf-verwijzende en geslotenkenmerken van therapeutische systemen. Dit hoofdstuk beoogt verhelderend en illustratief te zijn voorsommige verschijnselen van vast komen te zitten in een gesloten systeem. Het laat ook zien hoe hetontbreken van begrip bij de waarnemer of onderzoeker door middel van exploratie ten nutte kan wordengemaakt.

In hoofdstuk 11 behandel ik het begrip ’projectieve identificatie’. In tegenstelling tot de zelf-verwijzingdie in hoofdstuk 9 werd besproken, hebben we hier te maken met situaties waarin een verminderdvermogen (bij de therapeut of bij de externe waarnemer) om helder te denken, een eigenschap is van hetwaargenomene! Was kritische perceptie onmogelijk in processen van exploratie, ten gunste van eenbloeiende naïeve waarneming bij therapeut en patiënt, hier, in processen van projectieve identificatiekomen naïeve waarneming en daarmee ook exploratie juist tot stilstand, en kunnen pas weer op gangkomen naarmate de therapeut (of waarnemer) in staat is een deel van de negatieve ervaringen te verdragendie de patiënt juist niet kan verdragen. Deze negatieve ervaringen hebben de kwaliteit van ’onverteerd’; ze

228

zijn ’rauw’, en net als radicale atomen zijn ze ongrijpbaar zonder ze te veranderen en ze te binden. Hiertoont zich de negatieve weg, de exploratie van ons eigen onvermogen om tot een goed begrip te komen,als het meest vruchtbaar.

In hoofdstuk 12 worden enkele slotoverwegingen gegeven over de zelf-verwijzing vanpsychotherapeutische processen, en over de onmogelijkheid om die in positieve termen te vangen.

Stellingen behorend bij het proefschrift ’Towards a negative understanding of psychotherapy’, teverdedigen op 26 oktober 1998 om 14.45 uur te Groningen, door Arno Goudsmit

(goudsmit@mailbox.unimaas.nl)

1Psychotherapeutische processen kennen een zekere zelf-verwijzing. Deze bemoeilijkt empirischonderzoek volgens de gangbare methoden. Deze zelf-verwijzing is tevens de blinde vlek vangedragstechnologen, die menen dat therapeutische procedures van tevoren kunnen wordengespecificeerd.

2Parallelprocessen tussen psychotherapie en supervisie zijn een beproefd en veel gebruikt middelvoor de supervisor om toegang tot de gesuperviseerde therapie te krijgen. In empirisch onderzoeknaar psychotherapieën kan men evenzeer van dit middel gebruik maken.

3De negatieve wegverschilt van Hume’s skepticisme en van Popper’s kritisch rationalisme omdathet hier gaat om de onjuiste ideeën of hypothesen als manifestaties te nemen van datgene wat menonderzoekt.

4In veel empirisch onderzoek wordt een complot gesmeed met behulp van hypothesen en variabelen.De variabelen zijn dan de commitments waar de onderzoeker in de loop van het onderzoek spijtvan krijgt; de hypothesen zijn de cosmetische oplossingen om het onderzoek alsnog geslaagd tedoen lijken. Onderzoek kan pasradicaal empirischworden (verg. James, 1912) als ook het falenzelf respectabel mag zijn.

5Het verschil tussen de binnenzijde en de buitenzijde van een eenzijdige grens is dat binnen hetverschil tussen binnen en buiten niet bestaat, en buiten wel. De eenzijdige grens is daarmee degrens van zijn eigen domein van geldigheid, en men passeert hem in de ene richting door hem tevergeten, in de andere richting door hem opnieuw te beseffen.

6Ontdekkende psychotherapieën zijn vaak op zoek naar naieve objecten; als deze voor essenties vande klachten worden aangezien, dan kan een misplaatste diepzinnigheid ontstaan die dit soorttherapieën kwetsbaar maakt voor nominalistische kritiek (verg. hoofdstuk 8, dit proefschrift).

7In onderhandelingen met fundamentalisten (ongeacht of het gaat om aanhangers van religieuze ofseculiere opvattingen en doctrines) dient een ’beschaafde’ partij die zichzelf vindt passen binnen detraditie van de Verlichting, te beseffen dat de tegenpartij strijdt voor de zuiverheid en deschoonheid van een ideaal, en dat de gemeenschappelijke constructie van kritische objecten, zoals

compromissen en nuanceringen (zie hoofdstuk 4, dit proefschrift), daaraan per definitie afbreukdoet. Idealen waar fundamentalisten voor strijden, zijn naieve objecten (zie hoofdstuk 3, ditproefschrift), en kunnen niet adequaat in termen van kritische objecten worden beschreven. Defundamentalistische overtuigingen zijn dan ook geen ’opvattingen’ in een kritisch discours.

8De psychologie bevindt zich in een grondslagencrisis. Alleen door het vak zelf te beoefenen énerover te reflecteren, kunnen uitwegen worden gevonden. Historiserende benaderingen kunnenhierbij weliswaar behulpzaam zijn, maar zijn in principe nooit toereikend. Het vak ’grondslagenvan de psychologie’ kan ook bestaan zondermirror of nature (verg. Rorty, 1979).

9Strikte protocollen en procedures voor psychotherapeutische behandelingen zijn vergelijkbaar meteffectief berekenbare procedures in de getaltheorie. Ze werken voor specifieke klachten, zoalscomputerprogramma’s specifieke taken kunnen uitvoeren. De overtuiging dat voor alle klachten eentherapeutische procedure kan worden gespecificeerd (omdat elke klacht door eenbepaaldmechanisme zou worden veroorzaakt) berust op dezelfde misvatting als de overtuiging dat vooralle natuurlijke processen een berekenbare procedure bestaat (de Church-Turing hypothese, ziehoofdstuk 6, dit proefschrift). Dit is een meer formele verwoording van wat beoefenaars van’softe’, niet-geprotocolliseerde behandelwijzen binnen de psychotherapie bedoelen, wanneer ze inde bekende geitenwollige en bloemige taal hun bezwaren tegen ’onpersoonlijke’ techniekenkenbaar maken.

10De menselijke intuïtie is een van de laatst overgebleven wonderen, die niet doornatuurwetenschappelijke benaderingen zijn weg-verklaard.

11Autorijlessen zijn een geschikte plaats voor het experimenteel bestuderen van de processen vanindividuatie en separatie tussen leraar en leerling.

12’Geestige stellingen’ bij een academisch proefschrift zijn een soort literair genre, waarbij debeoogde geestigheid het wint van de daadwerkelijke. De tweede zin van deze stelling beoogt dattevens te illustreren.

13"Wetenschappelijk onderwijs is pas van kwalitatief hoog niveau als de afgestudeerdenprobleemloos kunnen worden ingezet op de arbeidsmarkt. Dit zei C. Herkströter, oud-topman vanShell, vanmiddag bij de opening van het universitaire jaar aan de Erasmus UniversiteitRotterdam."(NRC-Handelsblad, 7 september 1998)Evenzo: het bedrijfsleven is pas van kwalitatief hoog niveau als het zich belangeloos inzet voor debevordering van zuiver wetenschappelijk onderzoek en onderwijs.