Illocutionary force and degrees of strength in language use

24
ELSEVIER $~ltrMl~ Journal of Pragmatics 33 (2001) 1791-1814 www.elsevier.com/locate/pragma Illocutionary force and degrees of strength in language use * Marina Sbis~* Department of Philosophy, University of Trieste, via dell'Universit~ 7, 34123 Trieste, Italy Received 15 October 1999; revised version 19 September 2000 Abstract In this paper, I propose to deal with mitigation/reinforcement phenomena in terms of 'degrees of strength' of speech acts, and in particular of their illocutionary force. Some aspects of the received conception of a speech act do not allow illocutionary force to vary in degree and therefore it would be inappropriate to deal with mitigation/reinforcement in tradi- tional speech-act theoretical terms. But a revised conception of the speech act as bringing about a change in the interpersonal relationship between the interlocutors might be compati- ble with the results of the research on mitigation/reinforcement and even contribute to a bet- ter understanding of these phenomena. In this perspective, mitigation and reinforcement appear not as stylistic phenomena superficially adjoined to the speech act, but as the adjust- ment and tuning of the illocutionary effect itself. In order to show how mitigation and reinforcement can be traced back to aspects of the illocutionary act and be described by the same means by which illocutionary effects can be described, I discuss examples of mitigation and reinforcement taken from recorded conversa- tions in Italian. © 2001 Elsevier Science B.V. All rights reserved. Keywords: Speech acts; lllocutionary force; Mitigation; Reinforcement; Verbal interaction A preliminary version of this paper was read at Boston University in spring 1999; I am grateful to Bruce Fraser and to the audience for comments and questions. Provisional results of my ongoing research on illocutionary force modification had formerly been presented at the Conference "L'efficacia del testo", S. Marino, 27-29 September 1996; I thank the participants for the discussion. I also thank Claudia Caffi and Carla Bazzanella for all the discussions we have had on illocutionary force modifica- tion, and two anonymous referees for their stimulating suggestions and criticisms. * E-mail: [email protected] 0378-2166/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved. PII: S0378-2166(00)00060-6

Transcript of Illocutionary force and degrees of strength in language use

ELSEVIER

$~ltrMl ~

Journal of Pragmatics 33 (2001) 1791-1814

www.elsevier.com/locate/pragma

Illocutionary force and degrees of strength in language use *

Marina Sbis~*

Department of Philosophy, University of Trieste, via dell'Universit~ 7, 34123 Trieste, Italy

Received 15 October 1999; revised version 19 September 2000

Abstract

In this paper, I propose to deal with mitigation/reinforcement phenomena in terms of 'degrees of strength' of speech acts, and in particular of their illocutionary force. Some aspects of the received conception of a speech act do not allow illocutionary force to vary in degree and therefore it would be inappropriate to deal with mitigation/reinforcement in tradi- tional speech-act theoretical terms. But a revised conception of the speech act as bringing about a change in the interpersonal relationship between the interlocutors might be compati- ble with the results of the research on mitigation/reinforcement and even contribute to a bet- ter understanding of these phenomena. In this perspective, mitigation and reinforcement appear not as stylistic phenomena superficially adjoined to the speech act, but as the adjust- ment and tuning of the illocutionary effect itself.

In order to show how mitigation and reinforcement can be traced back to aspects of the illocutionary act and be described by the same means by which illocutionary effects can be described, I discuss examples of mitigation and reinforcement taken from recorded conversa- tions in Italian. © 2001 Elsevier Science B.V. All rights reserved.

Keywords: Speech acts; lllocutionary force; Mitigation; Reinforcement; Verbal interaction

A preliminary version of this paper was read at Boston University in spring 1999; I am grateful to Bruce Fraser and to the audience for comments and questions. Provisional results of my ongoing research on illocutionary force modification had formerly been presented at the Conference "L'efficacia del testo", S. Marino, 27-29 September 1996; I thank the participants for the discussion. I also thank Claudia Caffi and Carla Bazzanella for all the discussions we have had on illocutionary force modifica- tion, and two anonymous referees for their stimulating suggestions and criticisms. * E-mail: [email protected]

0378-2166/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved. PII: S0378-2166(00)00060-6

1792 M. Sbish / Journal of Pragmatics 33 (2001) 1791-1814

1. Introduction

Research in pragmatics has shed light on a broad field of linguistic phenomena regarding 'mitigation' or 'attenuation' as well as 'strengthening' or 'reinforcement' l in language use (Fraser, 1980; Holmes, 1984; Blum Kulka et al., 1989; Bazzanella et al., 1991; Dressier and Merlini Barbaresi, 1994; Caffi, 1999). This research field is of central importance for our conception of the use of language in communication: in tackling it, we are forced to abandon oversimplifications and idealizations, and to recognize the multi-dimensional nature of language use, ranging from cognitive to emotional facets, from actional to affective ones, from the social to the subjective. The aim is to account for language use in all its irreducible complexity.

But there are risks in this enterprise. First, pragmatics might become like a Bor- gesian map: a description of language use so attentive to its multi-dimensionality and its innumerable, though still significant, details, that this description ends up by reduplicating that which it purports to analyze, thereby risking the loss of any cog- nitive and explanatory potential. Second, any attempt to analyze language use in all its stylistic complexity is liable to rely on ad hoc tools in an eclectic vein, and may therefore neglect basic theoretical issues.

In this paper, I choose an intermediate route: I consider the relationship between the mitigation/reinforcement phenomena and speech act theory. Can mitigation and reinforcement be dealt with in a speech-act theoretical framework? I contend that they can and will substantiate this claim by analyzing examples of mitigation and reinforcement in terms of 'degrees of strength' of the speech act involved (a notion first used by Holmes, 1984). In doing so, however, I do not make use of the standard, mainstream version of speech act theory (Searle, 1969, 1979; Bach and Harnish, 1979; see also Tsohatzidis, 1994). I contend that some aspects of standard speech act theory are incompatible with the features of linguistic action emerging from research on mitigation/reinforcement. Instead, I argue that a revised conception of the speech act (of Austinian inspiration) such as the one I have proposed in Sbish (1989) might be compatible with the results of the research on mitigation/reinforcement and even contribute to a better understanding of certain features underlying these phenomena. In this perspective, mitigation and reinforcement appear not as superficially adjoined to independently performed speech acts, but as closely connected to, or even identi- cal with, the adjustment and tuning of the effects that the speech act brings about.

2. Why a unified account of mitigation/reinforcement and illocutionary force is desirable

The term 'illocutionary force', a key term in speech act theory, is generally used to refer to the fact that in the uttering of a sentence, an illocutionary act of a certain

1 The term 'aggravation' has also been used (since Labov and Fanshel, 1977). I prefer 'reinforcement' because I do not believe that the upgrading of intensity is necessarily correlated with aggressivity or threat to face.

M. Sbis~ / Journal of Pragmatics 33 (2001) 1791-1814 1793

kind is performed, that is, the utterance counts as a certain kind of move in verbal interaction: a command rather than a question, an assertion rather than an apology or a promise. I will now outline some of the reasons that make it desirable to give a unified account of such a phenomenon and of mitigation/reinforcement in language u s e .

(i) On the one hand, force (a notion coming from Frege (1879, 1918) where it applies to assertions and yes-no questions) has been considered since Austin (1962) as one of the aspects of language use that make speaking the doing of something. It pertains to illocutionary acts (acts of doing something in saying something) just as meaning (sense and reference) pertains to locutionary acts (acts of saying something) (Austin, 1962: 100). On the other hand, mitigation and reinforcement in language use are tied to matters such as the adjustment of the relationzhip between the inter- locutors, the achievement of goals, the avoidance of undesirable consequences, and therefore to what we do in verbal exchanges. It would be reasonable to expect that there should be some relationship between illocutionary force and mitigation/rein- forcement, namely, that these should correspond to modifications or adjustments of what is done in saying something.

(ii) There are some features of the speech act that seem to be affected by varia- tions in degree. First and most obviously, as was stated most clearly by Searle (1976), each illocutionary force is connected with the expression of an inner state. Since inner states typically vary in intensity, in order to fully characterize a speech act that expresses some inner state of the speaker, it would be sensible to take into account not only the kind of state expressed, but also its degree of intensity, and therefore mitigation/reinforcement phenomena. Second, there are speech acts the performance of which is influenced, or influences, matters of power. Sometimes what entitles the speaker to perform a speech act is his or her status within the ongo- ing interaction or 'local rank' (Diamond, 1996), which can at least in part be negoti- ated (and thus upgraded or downgraded) during the interaction itself. Other speech acts can be performed only by speakers who are already endowed with the appropri- ate status, and in these cases the speaker can either appeal to such status openly or politely avoid exhibiting it. In both kinds of cases, illocutionary force and mitigating or reinforcing strategies work together in defining or redefining the kind and level of the speaker's power as well as the extent to which it bears on the relationship between speaker and addressee.

(iii) It is generally recognized that there are different ways of performing acts having the 'same' illocutionary force: for example, there are different ways, with varying degrees of explicitness and directness, of requesting or apologizing or advis- ing or undertaking commitments. It has also been recognized that these different strategies have different effects on the interpersonal relationship between the inter- locutors: some may turn out to be disruptive, while others may mitigate the threat to 'face', avoid conflict, or be reassuring (Brown and Levinson, 1987: 68-75; Geis, 1995: 97ff., 181-183). Certainly, it is still possible to consider these different strate- gies as mere stylistic variations of an act whose essence remains untouched. But it cannot be denied that what the speaker does in speaking, changes (at least to some extent) depending on the different strategies employed.

1794 M. Sbisdt / Journal of Pragmatics 33 (2001) 1791-1814

(iv) In research on the ways in which speech acts are performed in different lan- guages and cultures (see in particular Blum Kulka et al.), the linguistic indicators of mitigation and reinforcement ('upgraders' and 'downgraders') are considered in the context of the strategies for performing speech acts as in principle distinct from illo- cutionary force indicators. However, it has also been noticed (Blum Kulka, 1985; Sbis~, 1990; Caffi, 1999) that sometimes at least, the relationship between mitigat- ing or reinforcing devices and illocutionary force is closer. There are cases in which mitigation (e.g. the use of a particle like 'please') contributes to characterizing the speech act as a directive ('Can you reach that book, please?' is more definitely a directive than 'Can you reach that book?'), while in other cases emphasis on the expressed inner state contributes to characterizing the speech act as an expressive ('Oh, it is so pretty' is more clearly a compliment than 'It is pretty'). A unified account of illocutionary force and of mitigation/reinforcement could explain these overlaps.

(v) Finally, the connections established by a unified account encompassing both illocutionary force and mitigation/reinforcement would make the methodological framework of pragmatically oriented discourse analysis more economical and per- haps more illuminating.

I do not claim that any of the points (i)-(v) above gives conclusive evidence for the desirability of a unified account of illocutionary force and mitigation/reinforce- ment. But taken together, they strongly suggest that such an account is desirable and that its viability is worth exploring. 2

2. Conditions for a unified account of illocutionary force and mitigation/rein- forcement phenomena

In order for a theory of speech acts to provide a unified account of illocutionary force and mitigation/reinforcement, that theory must be able to satisfy at least the following conditions:

(i) The conception of illocutionary force should be compatible with its being affected by variations in degree of strength.

(ii) Descriptive tools for speech acts should be able to account for variations in degree of strength.

(iii) It should be possible to describe mitigations and reinforcements by means of the same descriptive tools used to describe speech acts.

I will now argue that the standard conception of illocutionary force does not meet condition (i). Since condition (i) is logically prior to conditions (ii) and (iii), this will

2 Such viability has been relied upon rather than discussed by Bazzanella et al. (1991) and by Caffi (1999), who have distinguished between kinds of upgrading and downgrading devices on the basis of the relationship between the corresponding mitigation or reinforcement phenomena and aspects of the speech act.

M. Sbisct / Journal of Pragmatics 33 (2001) 1791-1814 1795

be enough to show that the standard conception of illocutionary force cannot be included in any unified framework encompassing mitigation/reinforcement phenom- ena without modification. Subsequently, I will argue that a reformulated conception of illocutionary force could meet all three requirements and qualify as suitable for dealing with degrees of strength in language use.

2.1. The standard conception of illocutionary force

Since Searle (1969: 46-49), the illocutionary act has generally been conceived as the act a speaker successfully performs when, uttering a sentence with a certain intention in certain circumstances, he or she gets the hearer to understand his or her intention. The speaker's communicative intention determines what illocutionary act he or she should be taken to perform and therefore what illocutionary force his or her utterance may have. The effect characteristic of the illocutionary act (its 'illocution- ary effect') has been identified with the recognition on the part of the hearer of the speaker's communicative intention.

The speaker's communicative intention must be a definite one. A speech act must have one or the other of the illocutionary points available (committing the speaker to the truth of the expressed proposition, trying to make the addressee do something, and so forth) (Searle, 1976). It does not matter here whether recognition of the speaker's communicative intention is achieved by recognizing the conventional meaning of the sentence uttered (as in Searle, 1969, and in 'direct' speech acts according to Searle, 1976) or by making inferences (as in Bach and Hamish, 1979; Sperber and Wilson, 1986; Harnish, 1994; and in 'indirect' speech acts in Searle, 1976): since in both cases, the illocutionary effect coincides with the recognition of the speaker's communicative intention or illocutionary point, it is bound to be as dis- crete as they are. 3

Aspects of the speech act which admit of degrees such as speaker's status or speaker's involvement are considered as marginal. According to Katriel and Dascal (1989), variations in degree of the speaker's involvement do not modify his or her commitment to the act he or she performs. According to Searle and Vanderveken (1985), directives display differences in degree as to the status that the speaker must have for performing them (from orders, which require the speaker's higher status, to entreaties, which require the speaker's lower status), but these differences affect the mode of achievement of the illocutionary point, not the illocutionary point itself. In no case is the illocutionary effect subject to variations in degrees of strength or describable in ways admitting of mitigation and reinforcement.

Thus, the standard conception of illocutionary force is not compatible with its being affected by variations in degree of strength and cannot meet the above-men- tioned conditions for a unified account of illocutionary force and mitigation/rein- forcement phenomena.

3 One more reason for taking illocutionary force to be discrete is available to those linguists who con- sider it as tied to sentence mood (Sadock and Zwicky, 1985; Sadock, 1994): the sentence types of a lan- guage form a system and are mutually exclusive (Sadock and Zwicky, 1985: 158-159).

1796 M. Sbisd /Journal of Pragmatics 33 (2001) 1791-1814

2.2. A reformulated conception of illocutionary force

Austin's original conception of the illocutionary act displays at least some fea- tures which distinguish it from the standard view presented above. The one most rel- evant to us is that the illocutionary act is associated with three different kinds of effect (1962: 116-117):

(i) the securing of uptake (ii) the production of a conventional effect (iii) the inviting of a response or sequel

Effect (i) amounts to bringing about the understanding of the meaning and the force of the utterance and, unless it is achieved, the illocutionary act is not actually carried out. Effect (ii) amounts to the bringing about of a state of affairs in a way dif- ferent from bringing about a change in the natural course of events: the act of nam- ing a ship 'Queen Elizabeth' makes it the case that this is the ship's name, and that referring to it by any other name will be out of order, but these are not changes in the natural course of events. Effect (iii) amounts to inviting a certain kind of subse- quent behavior; if the invitation is accepted, a certain further act by some of the par- ticipants will follow.

In the standard, Searlean-Gricean sense the 'illocutionary effect' is restricted to Austin's effect (i), but Austin himself does not suggest that restriction. Rather, he seems to imply that effect (i) is a necessary step for the successful performance of the illocutionary act, which also comprises the achievement of effects (ii) and (iii). 4 One might think that the last two effects should be dismissed as 'perlocutionary', but in Austin's view they are clearly not such: perlocutionary acts are the bringing about of changes in the natural course of events (as opposed to effect (ii), which affects a conventional state of affairs) and consist of the actual eliciting of a response (as opposed to effect (iii), which is merely the inviting of a response).

In my work on speech act theory (Sbish, 1984, 1987, 1989) I have argued that Austin's effect (ii), the production of changes not in the natural course of events but in conventional states of affairs, is an essential feature of the illocutionary act and can be detected not only in institutional acts such as the example provided by Austin (1962:116), but also in illocutionary acts usually performed in verbal interaction (cf. von Savigny, 1988). The production of such a conventional effect is made possible by the hearer's uptake of the act as bringing about that effect (as suggested by Austin), and 'conventionality' appears therefore to be founded on intersubjective agreement. 5

4 Strawson (1964) seems to have inferred from the fact that Austin (1962: 116) claims that unless a certain effect is achieved, the illocutionary act will not have been successfully performed, that for him the only effect essential to the illocutionary act is the achieving of uptake. On the basis of the whole con- text in which Austin's claim occurs, I do not think this interpretation is fair. 5 I am not thinking of explicitly signalled agreement. What counts as agreement may be reached by default, via the assumption that when the hearer fails to challenge the felicity of the speaker's

M. Sbisgt / Journal of Pragmatics 33 (2001) 1791-1814 1797

The conception of the illocutionary act according to which it has a conventional effect is thus committed to granting an active role to the hearer. Hence, to the stan- dard conception of illocutionary force based on the speaker's intention (which iden- tifies the illocutionary effect with the hearer's recognition of that intention), it is pos- sible to oppose a conception of illocutionary force according to which illocutionary acts have conventional effects, determined at least in part according to the hearer's uptake.

It might be useful to recall that the recognition of the hearer's active role is inde- pendently supported by studies in verbal interaction which have emphasized the cooperation between the participants in determining the meaning of the interactional event and therefore of the acts performed (Streeck, 1980; Duranti, 1988; Goodwin, 1995). If these cooperative aspects of interaction are taken seriously, it is impossible to believe that the speaker determines what he or she does by having a certain inten- tion. Indeed, a speaker may even turn out to have done (with his or her words) some- thing he or she did not intend to do. 6

A conception of illocutionary force which admits of conventional effects in the sense outlined above allows for degrees of strength. What the speaker has done (the effect of the illocutionary act) is no longer bound to mirror a discrete intention of the speaker. Rather, since there are aspects of the interpersonal relationship that are set- fled on the basis of intersubjective agreement, the conventional effects of speech acts may be considered as affecting them, and as playing a role in their adjustment and fine tuning. Thus a conception of illocutionary force which does not identify it with the speaker's communicative intention, but with the fact that utterances have con- ventional effects of certain kinds, meets condition (i) for a unified account of illocu- tionary force and mitigation/reinforcement phenomena.

As to condition (ii), regarding the tools to be used in the description of speech acts and their suitability for coping with matters of degree, I have proposed elsewhere (Sbis~, 1984, 1989, 1992) that conventional illocutionary effects may be described in terms of 'deontic modality', namely, as assignments to or cancellations from each one of the participants of modal predicates related to the necessity or possibility of actions with respect to norms: for example, there are speech acts as an effect of which it can be said of the addressee that he or she 'can' or 'may' or 'ought to' or

illocutionary act, he or she accepts it as the speaker characterizes it and when the speaker fails to reject the bearer 's uptake, he or she accepts to have done what he or she has been taken to do. If it should become manifest that default conditions do not obtain, implicit or even explicit negotiation will occur. 6 It may be objected that in such cases, what happens is mere misunderstanding. But while there are cases in which the speaker engages him or herself in a repair procedure of the kind described by Sche- gloff (1992) (and even here, negotiation may come into it), there are also cases in which he or she just lets the bearer 's uptake pass. The objection rests on a view of communication which is only apparently common sense, the 'expressivist ' view (cf. Gauker, 1994) going back at least to Locke or even to Augus- tine, according to which communication consists of the expression by the speaker of a preformed con- tent and of the recognition of the same content by the addressee. This view was attacked by Wittgenstein (1953) and ignored by Austin; research on interaction, highlighting the role of the hearer, opposes it at least potentially.

1798 M. Sbisd / Journal of Pragmatics O0 (2001) 000-000

'has to' do something. 7 The lexicon of deontic modality includes, besides modals and other verbs, a broad range of nouns (Lyons, 1977: 825) such as 'right ' , 'oblig- ation', 'entitlement', and 'commitment ' . The reason for choosing such a lexicon for the description of conventional illocutionary effects (and therefore of illocutionary acts and illocutionary forces) is that the assignment and removal to and from agents of deontic modal values is basically 'conventional ' , namely, dependent on social factors such as the relevant intersubjective agreement. Now, deontic modal notions are in many respects matters of degree: they display varying degrees of strength as well as qualitatively different shades of meaning. So obligations can be legal or moral, and more or less binding; rights may be competencies, capacities, or author- ities; the scope of their contents may vary too. Moreover, attributions of deontic modal values have intuitive connections with matters of status and power, 8 which typically come in degrees, So in the analysis of verbal interaction, the lexicon of deontic modality offers considerable potential for capturing degrees and shades. The reformulated conception of illocutionary force can thus be held to satisfy con- dition (ii) for a unified account of illocutionary force and mitigation/reinforcement phenomena.

As to condition (iii), that mitigation and reinforcement may be described by means of the same descriptive tools that are used to describe speech acts, I maintain that a great deal of the mitigating and reinforcing effects highlighted by research on interactional management, politeness, expression of emotions, and so forth, can be described completely or in part by means of the same deontic modal notions which I have proposed to use in the description of conventional illocutionary effects. The remaining part of this paper will be devoted to an attempt at showing that this is so, not by means of theoretical arguments, but by discussing examples. The discussion is not aimed at giving incontrovertible evidence in favor of the reformulated con- ception of illocutionary force I have sketchily presented. But I would like to show how this proposal can be put to work and what it is good for: in particular, how it can deal with mitigation and reinforcement phenomena.

7 I am referring here to modality as the semantic domain relating to the expression of necessity and possibility. Deontic modality is concerned with necessity and possibility interpreted as the necessity or possibility of actions or modes of behavior with respect to some kind of social rule or norm (cf. von Wright, 1972; Hugues and Cresswell, 1996; Lyons, 1977). In linguistics, deontic modality has been included in the wider notion of root modality (necessity and possibility depending on sociophysical forces or barriers) (Sweetser, 1990). Narrative semiotics has used deontic modalities ('can' and 'ought to' applied to doings) to describe the 'competencies' that differentiate the respective positions or roles of agents and are presupposed and transformed by their performances (Greimas, 1983; Greimas and Courtrs, 1982). 8 To be in a position in which one 'can' do more things than someone else is typically at the same time to be in a 'one-up' position, i.e. enjoy greater 'power'. That this is so is even more obvious for the speak- ers of those languages, such as French or Italian, in which the meaning of the substantivized infinitive of the modal verb translating both 'can' and 'may' (pouvoir, potere) corresponds to one meaning of the noun 'power'. Sociosemiotics (the application of narrative semiotics to social facts) has exploited this connection (Landowski, 1989).

M. Sbisd /Journal of Pragmatics 33 (2001) 1791-1814 1799

3. Cases of mitigation and reinforcement

The examples I am going to d iscuss are d rawn f rom a misce l l aneous corpus of recorded conversa t ions in Italian. 9 They do not consis t o f isolated, c lear-cut sen- tences (as is cus tomary in ph i losophy) , but o f textual units, cor responding to con- versa t ional turns or funct ional ly separable parts o f these. Al though cases o f mi t iga- t ion are more c o m m o n and p lay a more impor tant role in the pol i te ma na ge me n t o f interact ion, the examples presented here wil l regard re inforcement too, because m y a im is to test the descr ip t ive potent ia l o f m y approach in both direct ions.

In d iscuss ing the examples , I focus on the convent iona l i l locut ionary effect, which I take to be the core componen t o f the set o f effects o f the i l locut ionary act out l ined by Aus t in (1962: 116-117) (I wil l cal l this the ' co re i l locut ionary ef fec t ' ) . In the descr ip t ion o f the core i l locut ionary effect , I d is t inguish three ma in componen t s : the deont ic moda l va lues which the speaker appeals to in order to pe r fo rm the i l locu- t ionary act (in short, the speake r ' s ent i t lement) , the deont ic moda l values which the i l locut ionary act ass igns to the addressee (e.g. r ights, obl igat ions , or cancel la t ions o f these), and the deont ic moda l values which are cor respond ing ly ass igned to the

9 The examples quoted are drawn from various contexts: TV programs, scientific work groups, psy- chotherapeutic sessions, courtroom hearings, informal interactions. Some were recorded and transcribed by me, in part with the collaboration of my students, to whom I am very grateful. Others (examples (1), (2), (7), (11), (12), (13)) were contributions provided by Carla Bazzanella and Claudia Caffi to the cor- pus we collected while preparing Bazzanella et al. (1991; see also Bazzanella, 1994). In my research, I have also examined transcriptions published by others (Leonardi and Viaro, 1990; Gavioli and Mans- field, 1990). Although the focus of my paper is not on the Italian language as such, I quote the examples in the original Italian, because they are the source material which I have been working on, given my competence as a native speaker. The English translations are not pragmatic equivalents of the Italian examples, but aim at matching their syntactic and lexical-semantic characteristics, thus enabling the readers to identify the mitigating or reinforcing features in the original; punctuation is inserted only to make the text more readable. The Italian words to which the mitigating/reinforcing effect can be traced back and their English counterparts are both underscored, but this does not imply that the English expression used to translate an Italian mitigation or reinforcement indicator plays an analogous role in English.

I have used the following transcription conventions: Beginning of an overlap * or # (in both turns) End of an overlap + or / (in both turns) Latched utterances = Short pause (up to 2") (-) Pause (between 2" and 5") (--) Slightly rising or level intonation contour Rising intonation contour 9 Falling intonation contour Lengthened syllables :, :: Incomplete words x- Low volume (^ word) Emphasis bold Non-linguistic phenomena and comments ((xxxx)) Unintelligible words xxx Omitted (...)

1800 M. Sbisd / Journal of Pragmatics 33 (2001) 1791-1814

speaker (the speaker's commitment) (Sbis~, 1984). Firstly, I will present examples of mitigation and reinforcement that affect the three components of the core illocution- ary effect. Subsequently, I briefly consider cases of mitigation or reinforcement related to two dimensions of the illocutionary act which do not straightforwardly belong (in my description) to its core effect: the expressed inner state and the invit- ing of a response. 1°

My ordering of the examples, and the related distinctions among kinds of mitiga- tion/reinforcement phenomena, differ from the set of distinctions proposed by Baz- zanella et al. (1991) mainly because here, I do not consider variations in indetermi- nacy/precision or diminution/augmentation of propositional content separatedly from variations in the core illocutionary effect. In fact, they contribute to these variations and are described by describing them. For the same reason, I shall not use the dis- tinction between 'bushes' (mitigation procedures which focus on the propositional content) and 'hedges' (mitigation procedures which center on illocutionary force indicators) proposed by Caffi (1999). This distinction is undoubtedly useful when what is at issue are the linguistic means by which the mitigating effect is achieved, but the core illocutionary effect is affected by both kinds of procedure in closely intertwined ways. That is why 'bushes' and 'hedges' overlap, as Caffi herself points out. 11

3.1. Speaker's entitlement

Unless there is evidence to the contrary, a speaker whose utterance is accepted as the performance of a certain illocutionary act is also taken to satisfy the felicity con- ditions for that act. Thus a speaker who gives an order appears as someone who has the authority to do so, a speaker who issues a judgment as someone competent to do

~0 One limitation of my analyses is that the description of the effects of iUocutionary acts is based on the analyst's (i.e. my own) reception of the textual units considered. This might seem to neglect the hearer's active role. Apart from the difficulty or even impossibility of extracting indicators of the addressee's uptake from the turns immediately following the one under consideration (it often happens that the use of mitigation or reinforcement devices elicits some cumulative response later on, and not necessarily a verbal one), it should be noted that, although the analyst is never an addressee, he or she does act as a hearer (sometimes even a ratified one, as in the case of TV talk shows). For the aims of the present investigation, it is less relevant that we should know whether the illocutionary effects are actu- ally ratified by the addressee or other participants, than that we should be able to motivate the proposed reading of the conversational turn. However, in many cases it was possible at least to make sure that the addressee's or the audience's reaction to the analyzed turn did not disconfirm the kind of reception elab- orated on in the analysis of the turn itself. In other cases, notably examples (2), (10) and (13), the effects I describe are merely purported, since the participant endowed with the highest conversational power in the context (the compare, the therapist) skips or represses them. How exactly the principle of the hearer 's active role should be applied in discourse analysis is in need of further discussion. I I As to the mitigation devices which Caffi calls 'shields ' , I must admit that the cases of mitigation in which they are involved largely escape my attempt to give a unified description of force and mitiga- tion/reinforcement. I believe that their mitigating effect cannot be captured by the description of varia- tions internal to illocutionary effects, because they involve shifts or displacements external to them, and concerning the basic indexical features of the speech act. Maybe they are akin to ' footing' (Goffman, 1981).

M. Sbisa / Journal of Pragmatics 33 (2001) 1791-1814 1801

so, a speaker who makes a promise as someone capable of doing so. Such retro- spective definitions of the deontic modal attributes of the speaker may merely con- firm them, but they can also refine or alter them (Sbis~t, 1984, 1989). Thus the 'speaker's entitlement' to perform the act can be considered as a component of its core illocutionary effect.

In many cases, this entitlement involves some kind and degree of 'power' (where 'power ' is connected with deontic possibility). There are different qualitative shades, from authority or mere influence, to legitimation, capacity, or competence (knowl- edge, the opportunity to acquire it, or mere linguistic and communicative compe- tence). Each of these shades may be affected in turn by variations in degree: for example, authority may be absolute or subject to various kinds of restrictions; per- sonal influence may be weaker or stronger and relative to a broader or narrower scope; competence may vary in range and reliability.

A speaker may want to emphasize his or her entitlement to perform a certain speech act, for example by appealing to the sources of such entitlement:

(1) io non ho mai io sono una donna di teatro ex anche (-) ma di teatro (-) h ho notato in lui (-) ho seguito tutte le trasmissioni di Piero u- un fondo di umanitgt straordinaria. I have never I am a woman of the theatre or rather ex - but of the theatre - I have noticed in him - I have seen all of Piero's programs an an extraordinary fund of humanity

In (1) a participant in a TV discussion reinforces the evaluation she is expressing ("ho notato in lui ... un fondo di umanit~ straordinaria", ' I have noticed in him ... an extraordinary fund of humanity') by listing her own credentials (her experience of the theatre, her acquaintance with the object to be evaluated).

In this and similar cases, it would be exaggerating to say that the speaker's enti- tlement is strengthened, since the speaker possesses the status required for the per- formance of his or her speech act independently of the way in which the speech act itself is formulated. But the speaker's entitlement is certainly emphasized by its being openly appealed to. On the other hand, a substantial strengthening of the speaker's entitlement may occur when his or her status or local rank is not precisely defined and he or she attempts to perform a speech act which requires a stronger kind or higher level of entitlement than previously granted. In the following exam- ple (from a TV discussion about the sale of weapons),

(2) (...) ho visto anche violato il concetto di solidariet~ internazionale, ciod vera- mente c'd una schizofrenia morale in uno stato che da un lato aiuta questi bam- bini aiuta le popolazioni del terzo mondo: e dall'altra parte le mutila per fare degli affari abbastanza schifosi. I have also seen the idea of international solidarity violated that is to say there really is a kind of moral schizophrenia in a country which on the one hand helps these children, helps the peoples of the third world, and on the other mutilates them in order to do their rather disgusting business.

1802 M. Sbis~ /Journal of Pragmatics 33 (2001) 1791-1814

the speaker, who has formerly proposed his contribution as a ' testimony' (namely, as a commissive illocutionary act, to the performance of which he is entitled by his capacity to guarantee for the authenticity of the reported experience), attempts to issue judgments (verdictive illocutionary acts) ~2 and therefore, to get his competence to issue judgments recognized. His attempt is underscored by the use of the adverb veramente ( 'really'), collocated exactly at the transition point between the testimony ("ho visto . . ." , 'I have seen ... ') and the attempted judgment ("c'O una schizofrenia morale . . ." , ' there is a kind of moral schizophrenia ...,).~3

On other occasions, speakers mitigate their entitlement by appealing to a lower status than they could claim in that situation. In the following example, the chairman of a work group attempts to direct the participants' activities not by appealing to the authority he is endowed with, but by issuing informal suggestions and advice:

(3) II mio suggerimento ~ che facc iamo un brevissimo giro presentandoci (...) sarebbe opportuno che ci presentass imo per identificarci un po ' . My suggestion is that we do a very quick round introducing ourselves (...) it would be a good idea if we presented ourselves in order to identify ourselves a bit.

The kind of entitlement required for suggestion or advice, that is, for the illocution- ary acts which the speaker of (3) presents himself as performing, is weaker than he is in a position to appeal to. This does not hinder the speaker from achieving its per- locutionary goals: not only will his directive be complied with, but he will also appear as a very kind chairman.

The mitigation of the speaker's entitlement may produce a shift from one kind of illocutionary act to another. For any kind of illocutionary act presupposing some 'power' on the part of the speaker, there is a corresponding kind of illocutionary act, inviting similar responses, which requires only a lower level of 'power ' or even no 'power' at all (apart from the possession of linguistic and communicative compe- tence). For example, expressions of intentions can be described as promises which do not presuppose full capacity (and cannot therefore be equally binding), but still aim at reassuring the addressee, and protests can be described as similar to criticisms except for the specific competence which the latter require. Thus, as often happens, a woman interviewed in a TV talk show chooses to express an opinion ( ' Io credo che' . . . , 'I believe that . . . ' ) instead of issuing a judgement:

(4) Compare: M a c e ne sono in natura ? ( - ) Ecco scusi signora ma in questo tipo di: cosi." ce ne sono in natura ?

12 Commissives (Austin, 1962: 157; Searle, 1976) commit the speaker to a certain course of action. Verdictives are defined by Austin (1962: 153) as consisting in the delivering of a finding, official or unofficial, upon evidence or reasons as to value or fact. ~3 The comt~re does not ratify the attempted strengthening of the speaker's entitlement, but interrupts the interviewee and redef'mes his contribution as mere 'testimony' ("Grazie della sua testimonianza", 'Thank you for your testimony'). But other hearers (the TV audience) are in principle free to align them- selves with the comp~re's uptake or deem it unfair.

M. Sbisgl / Journal of Pragmatics 33 (2001) 1791-1814 1803

But do they exist in nature? - I mean, sorry madam but in this kind of thing, do they exist in nature? Woman interviewed: Io credo che cominciano ad esserci nel senso che: no, nel senso che: gli uomini si stanno in qualche modo femminilizzando nel senso che eh stanno cominciando a conoscere: la Ioro parte sentimentale, che si sono per tanto tempo probabilmente negati (...) I believe they are starting to exist in the sense that, no, in the sense that, men are becoming in some ways more feminine in the sense that they are beginning to get to know their emotive side, which for so long they have probably suppressed (...)

Finally, the entitlement to perform certain illocutionary acts may consist of the assignment of deontic modal values not related to possibility (and therefore to some kind of 'power') , but to necessity (and particularly to obligation, as is characteristic of certain expressives such as apologizing, thanking, and congratulating). Thus a participant in a TV discussion chooses to present his judgment on a very delicate issue not as the exercise of a competence, but as if his utterance were the fulfillment of an obligation (the duty to answer a question, or perhaps to react in a morally sound way to the event referred to):

(5) Comp6re: Chi ~ *secondo lei? + Who according to you? Participant: *Chi ~, chi ~+, e non lo so. Io devo dire, (...) non si tratta di omi- cidio di stato a mio avviso, ma certamente di un suicidio con grandi respons- abilitgt istituzionali, credo di: : (-) credo di si. Who it is, who it is, oh I don' t know. I must say, (...) it 's not a state-organized murder, in my opinion, but certainly a suicide which the institutions are largely responsible for, I think - I think so.

Even directives may be formulated as presupposing an entitlement of the speaker related to necessity (typically, a condition of need). So in the following example, a psychotherapist bases his request for a detailed report (addressed to one of the fam- ily members in therapy) on his own condition of ignorance.

(6) Therapist: Ecco. Emilio. Lei deve mettersi un momentino nei miei panni. Now. Emilio. You must put yourself in my place for a moment. Family member. Si. Yes. Therapist: Lei deve cercare di aiutarmi. Perch~ io non ci sono mai stata nella vostra famiglia e auindi quando You must try to help me. Because I 've never been in your family and so when Family member: Lo so. Va bd. Io posso farle I know. OK. I can do [that] for you (Leonardi and Viaro, 1990: 112)

The obligation assigned to the addressee ("Lei deve", 'You must'), a feature which would make the speech act under consideration not merely a directive, but specifically

1804 M. Sbisd / Journal of Pragmatics 33 (2001) 1791-1814

an exercitive, 14 is here presented as deriving not from the exercise of authority as in exercitives, but from the therapist's self-avowed state of need.

3.2. Modal values assigned to the addressee

The performance of an illocutionary act influences the deontic status of its addressee by assigning to him or her, or removing from him or her, some deontic modal value. For example, orders or advice assign obligations to their addressees; a promise assigns a right to its addressee; permission may be described as removing or preventing an obligation not to do something. Here too, variations in degree may occur. In principle, they should affect all kinds of assigned deontic modal values. Our examples here are limited to the mitigation or reinforcement of assigned oblig- ations.

The addition of the alternative "o no?" ( 'or not') to polar questions can be con- sidered as a case of reinforcement of the obligation assigned to the addressee, since it emphasizes that he or she must choose between two and only two answers. Here are two examples (both from TV talk shows):

(7)

(8)

Compare: Ma tu pensi di avere la verit~ in tasca o no ? But do you think you have the truth in your pocket or not? Compare: (...) ecco secondo lei (-) Nicolas risponde ai (-) pub dire quello che crede, ehm con la superlibertd ai requisiti, (-) secondo lei dell'uomo, sexy eh: o o n o ? (...) yes in your opinion - does Nicolas correspond to - you can say what you think, with super-freedom, to the requisites - in your opinion of the sexy man eh or or not?

The turn quoted in example (7) is openly aggressive (as is indicated also by the ini- tial ma, 'but'). In contrast, in example (8) the obligation assigned is qualified as an obligation to express personal commitment (secondo lei, 'in your opinion'), and its reinforcement is counterbalanced by a concession of superlibert~ ('super-freedom'), which, however, does not dispense the addressee from the obligation to answer the yes-no question in one of the two pertinent ways.

Another case of reinforcement is found in example (6) above: the indicative mood in "Lei deve" ('You+formality must'), as opposed to the widespread use of the con- ditional mood for making requests, underscores the fact that the obligation assigned is a strict one.

The mitigation of the assigned obligation by way of qualifying its content occurs (together with the mitigation of the speaker's power) in example (3):

J4 Exercitives are iilocutionary acts characterized by the exercise of authority in favor or against some course of action (Austin, 1962: 155) and typically assign obligations to their addressees (Sbis~, 1984). Many directives (attempts to get the hearer to do something: Searle, 1976) are exercitives as well, but not all of them get the hearer to do something by imposing obligations on him or her.

M. Sbisd / Journal of Pragmatics 33 (2001) 1791-1814 1805

(3) II mio suggerimento d che facciamo un brevissimo giro presentandoci (...) sarebbe opportuno che ci presentassimo per identificarci un po'. My suggestion is that we do a very quick round introducing ourselves (...) it would be a good idea if we presented ourselves in order to identify ourselves a bit.

Following the chairman's directions will cost the participants very little effort indeed !

3.3. The speaker's commitment

Obligations concerning subsequent linguistic and non-linguistic behavior are assigned to speakers not only by commissives, but also by other illocutionary acts as a part of their core illocutionary effect. The assignment of deontic modal values is not a unilateral matter: a speaker who modifies the status of his or her addressee cannot remain unaffected by his or her own action. Thus, just as to a commitment of the speaker to do something, there corresponds the assignment to the addressee of a right to expect that behavior from him or her, to any assignment of obligations to the addressee, there corresponds a commitment on the part of the speaker not to interfere with the assigned task. Assertion, too, has been analyzed as involving a commitment on the part of the speaker (Searle, 1976; Brandom, 1994).

The speaker's commitment is sometimes reinforced, especially as a reaction to the addressee's lack of confidence in the speaker's reliability.

(9) Judge: m a l e ha dato una spinta a s ta moglie o no. a sua moglie but you gave this wife of yours a push or didn't you, your wife Defendant: no niente ne*ss-+ no nothing no Judge: *si+curo sure Defendant: sicurissimo quite sure

(10) per quel nostro intervento era stata pattuita una somma di quattrocentomila fire compenso che da lei, non abbiamo mai ricevuto (...) For that job we did a fee of four hundred thousand lire was agreed upon, which we never received from her (...)

In (9) (from a courtroom heating) a defendant reinforces his commitment in answer to the judge's expression of mistrust by using a superlative (sicurissimo, 'quite sure', lit. 'very sure'). He is not merely expressing his confidence in the truth of the reply he has just given, but assuring the judge that his commitment to that reply is serious. Also in (10) (from a TV show), the addition of an informatively superfluous 'mai (+emphasis)' ( 'never') underscores the speaker's commitment to what he asserts.

The speaker's commitment is often downgraded, so as to lower the cost of the speech act, thereby making it more easily defensible. There are two main strategies

1806 M. Sbisd / Journal of Pragmatics 33 (2001) 1791-1814

for doing so. In the following example, from a psychotherapeutic session, the speaker uses both.

(11) Father: perO logicamente siamo sempre li nel merito (--) come giustamente io h o l e mie idee e penso di essere dalla parte della ragione lei ha le sue idee piuttosto dura But obviously we are back to the same problem basically - like quite rightly I have my ideas and I think I 'm in the right she has her ideas she's rather hard Little girl: ((cries out)) Father: ma molto dura almeno secondo il mio principio (...) but very hard at least according to my way of seeing things (...)

At first, the speaker doesn't commit himself to the straightforward claim that his wife is hard, but to a weaker one ("piuttosto dura", 'rather hard'); then he switches to calling her "molto dura" ( 'very hard'), but relativizes his commitment to his own subjective perspective. In both cases, the criticism he levels at his wife is easier to defend than an unrestricted or absolute one.

Sometimes, mitigated commitment parallels the mitigation of the speaker's enti- tlement. Let us reconsider our example (4):

(4) Woman interviewed: Io credo che cominciano ad esserci nel senso che: no, nel senso che: gli uomini si stanno in qualche modo femminilizzando nel senso che eh stanno cominciando a conoscere: la loro parte sentimentale, che si sono per tanto tempo probabilmente negati (...) I believe they are starting to exist in the sense that, no, in the sense that, men are becoming in some ways more feminine in the sense that they are beginning to get to know their emotive side, which for so long they have probably suppressed (...)

Besides presenting her speech act as an expression of belief rather than an act of judgment, the speaker uses the modal adverb probabilmente ( 'probably') to down- grade her commitment to the truth of the clause 'che si sono per tanto tempo (...) negati' ( 'which for so long they have (...) suppressed'), and the adverbial expression in qualche modo ('in some ways') to introduce vagueness into her expressed belief that men are becoming more feminine, thus making it easier to defend. ~s

3.4. Degrees of strength in expressed attitudes and inner states

Variations in degree affecting expressed attitudes and inner states may also be accounted for within the framework of the standard conception of the speech act.

is Piuttosto dura ('rather hard') and in qualche modo ('in some ways') might be classified as 'bushes' (Caffi, 1999), i.e. mitigation devices which focus on the propositional content. In my perspective, 'bushes' offer one strategy for downgrading the obligations one might have to face as a consequence of one's own speech act.

M. Sbisgt / Journal of Pragmatics 33 (2001) 1791-1814 1807

However, I consider them within the alternative framework I am proposing, in order to show their connections with the core illocutionary effect and its description in deontic modal terms.

In my framework, the 'expression' of attitudes and of inner states coincides with the speech act's licensing the hearer to assign a certain inner state (with a certain degree of intensity) to the speaker. Degrees of intensity may come into the picture in two ways: they may affect either the inner state expressed, or its expression, which may be foregrounded or backgrounded. Example (12) (from a psychotherapeutic ses- sion) is a case of the former kind, in which the intensity of an expressed inner state is reinforced:

(12) Therapist: ecco e che: (-) quando vi hanno detto di venire qui tutta la famiglia che cosa avete pensato well and what - when they told the whole family to come here what did you all think Mother: io ho pensato the era la cosa giusta perchd secondo me i bambini ne risentono tanto della situazione: : tesa in famiglia I thought it was the right thing because in my opinion the children are so much affected by the tense situation in the family

By using "tanto" ('so much') instead of "molto" ( 'much'), the speaker expresses, and invites hearers to ascribe to her, an intensified worry for her children. This effect is added to the core illocutionary effect of the performed speech act, but does not produce any substantial transformation of it: the speaker is still giving an explana- tion of her decision to initiate family psychotherapy.

When reinforcement concerns the expression of an inner state as opposed to the inner state expressed, a closer connection with the core effect of the illocutionary act may result: the speech act as a whole may qualify as an expressive. We have already met this phenomenon in example (4), in which the mention of the speaker's inner state of belief ("Io credo che", 'I believe') turns an act of judgment into the expres- sion of an opinion.

In example (5), in which the speaker's entitlement is presented as a matter of obligation rather than competence or authority, a similar foregrounding of the expression of inner states can be noted:

(5) Compare: Chi ~ *secondo lei ? + Who according to you? Participant: *Chi O, chi ~+, e non lo so. Io devo dire, (...) non si tratta di omi- cidio di stato a mio avviso, ma certamente di un suicidio con grandi respons- abilitgl istituzionali, credo di: : (-) credo di si. Who it is, who it is, oh I don't know. I must say, (...) it's not a state-organized murder, in my opinion, but certainly a suicide which the institutions are largely responsible for, I think - I think so.

1808 M. Sbisgt / Journal of Pragmatics 33 (2001) 1791-1814

The connection of the mitigation of the speaker 's entitlement with the foregrounding of the expression of inner states is not surprising, because both are features of expressive or behabitive illocutionary acts. ~6

The expression of not especially intense inner states is not perceived as a kind o1 mitigation, unless intense inner states are expected. As to the backgrounding of the expressive dimension of the speech act, it may coincide with an attempt to appear objective or formal, as in the following example (again from a psychotherapeutic session), in which the speaker, in complaining about his wife 's behavior, attenuates his involvement by using impersonal constructions:

(13) (...) non riesce cosl a: : ad amalgamarsi e dire cosi: : : questa cosa magari facciamola in questo modo e non in quest'altro: (-) ciod magari al momento ragiona pure e si dice si facciamo questa cosa al posto di quest'altra: : : poi ovviamente la si lascia a sd lei ritorna al punto di prima: : she doesn ' t manage you know to amalgamate herself and say you know this thing let 's do it like this and not like that - that is at that moment she may even be reasonable and one says yes let 's do this rather than something else, then obviously one leaves her to herself and she goes back to where she was before.~7

3.5. Perlocutionary goals o f the speech act

The way in which a speech act is formulated may also manifest the degree of intensity of the speaker 's intention or desire to achieve a certain perlocutionary goal, and the attempt to achieve the perlocutionary goal may itself be made in weaker or stronger ways. One reinforcement phenomenon to be mentioned in this connection is insisting.

Insisting is necessarily sequential and involves a series of attempts to achieve the same goal (basically, the elicitation of a certain response on the part of the addressee). The speech act is repeated and reformulated, and sometimes resort is had to explicit performative utterances. The increasing strength of the attempt to achieve the perlocutionary goal may be connected with reinforcements of the core illocu- tionary effect, since the reformulations may also strengthen the speaker 's entitlement or commitment or the modal values assigned to the addressee. In the following example (from a TV show acting out pretend courtroom hearings):

~6 Expressives (Searle, 1976) have the illocutionary point of expressing some inner state of the speaker. Behabitives (Austin, 1962: 160) consist in reactions to somebody's behavior or to events affecting some- body. Although their definitions differ, in practice the two terms cover the same domain. I have stressed (Sbis?~, 1984) that these illocutionary acts do not presuppose any kind of 'power' on the part of the speaker, but rather, a condition of obligation or debt. ~7 The impersonal constructions used in this example might be described as 'shields' (Caffi, 1999). There may be a connection between the backgrounding of the expression of inner states and mitigation by means of 'shields': the only way to radically background the expression of the speaker's subjectivity is to conceal the enunciating subject from view.

M. Sbisd / Journal of Pragmatics 33 (2001) 1791-1814 1809

(14) Defendant: signor giudice io assolutamente non pago perch~ loro mi hanno offeso: la mia persona mi hanno: messo nel ridicolo con apprezzamenti dici- amo: (-) volgari. (-) sicch~ (Anon: :)= your honour I am absolutely not going to pay because they have offended me, my person, they have (-) made me look ridiculous with let's say - vulgar - comments, so I 'm not Plaintiff: =ma no= but no Defendant: anzi chiedo dei danni morali in fact I demand moral damages

the initial refusal to pay is reformulated, by means of a performative utterance, as a demand to be refunded which presupposes a strengthened entitlement of the speaker and tries not only to deprive the plaintiff of his alleged rights, but also assigns him an obligation.

As to mitigation, it is not clear whether cases of mitigation of the desire or the attempt to achieve the illocutionary act 's perlocutionary object (the goal conven- tionally associated with the performed illocutionary act; Austin, 1962:118) can be detected at all. Sometimes, particularly in formal contexts, speech acts may be per- formed simply in order to fulfill one's role, without any real interest in the achievement of their associated perlocutionary goals; but the absence of such an interest need not be linguistically marked and thus falls beyond the scope of our research. Moreover, if an illocutionary act is interactionally taken as successfully performed, it does invite the associated response, and whether the speaker actually cares to elicit it or not does not seem to influence any aspect of the illocutionary effect.

Finally, speakers may have a wide range of additional perlocutionary goals beyond those conventionally associated with the illocutionary acts they perform, and speech acts may bring about an enormous range of perlocutionary sequels, whether intended or unintended. We may wonder whether this domain, too, is affected by mitigation/reinforcement. In Bazzanella et al. (1991), the avoidance or intensification of conflictual perlocutionary sequels has been considered as belonging to the dimensions of the speech act that are liable to be mitigated or reinforced. In consideration of the role played by mitigation with respect to polite- ness, it might even be claimed that perlocutionary sequels are what mitigation is pri- marily about.

I do not think that this is correct. Rather, a distinction should be drawn between the purposes that mitigation or reinforcement may serve and the mitigation/rein- forcement phenomena themselves. Mitigation and reinforcement achieve the tuning of the interactional relationship by adjusting the core illocutionary effect or associ- ated aspects of the illocutionary act. Mitigation strategies are often (or even regu- larly) aimed at downgrading conflictual perlocutionary sequels, but this outcome is mediated by the fine-tuning of the interactional relationship at the illocutionary level.

Let us consider here a wider stretch of a conversational turn we have already quoted:

1810 M. Sbisgt /Journal of Pragmatics 33 (2001) 1791-1814

(5) Chi ~, chi d, e non 1o so. Io devo dire, non sono: : d'accordissimo con quello che dice la figlia, la quale, la figlia di Lombardo, la quale rispetto naturalmente il il dolore, non si tratta di un xxx omicidio, non si tratta di omicidio di stato a mio avviso, ma certamente di un suicidio con grandi responsabilitgt istituzionali, credo di: : (-) credo di sL Who it is, who it is, oh I don't know. I must say, that I 'm not fully in agreement with what the daughter says, Lombardo's daughter, whose naturally whose pain I respect, it's not a xxx murder, it's not a state-organized murder, in my opinion, but certainly a suicide which the institutions are largely responsible for, I think - I think so.

This turn contains a great deal of mitigating devices, some of which have been dis- cussed above. The interactional problem the speaker is facing is how to give his interpretation of the event under consideration (the suicide of a Marshal of the Cara- binieri in a mafia context), which comes close to an accusation to state institutions, while avoiding being scornful of other persons involved or of the institutions and exposing himself too much. The solution the speaker adopts involves: mitigation of the speaker's entitlement (which is founded not on some kind of 'power', but on an obligation: "io devo dire", 'I must say'), foregrounding of expression of attitudes ("non sono d'accordissimo", ' I 'm not fully in agreement'; "la quale rispetto natu- ralmente il dolore", 'naturally whose pain I respect'), restriction of the speaker's commitment (by means of a litotes: lack of full agreement is mentioned rather than disagreement or even mere lack of agreement), relativization of the speaker's com- mitment to truth to the speaker's subjective perspective ("a mio avviso", ' in my opinion'), redefinition of the assertion made as an expression of belief ("credo", 'I think'). The overall mitigating effect on the degree of conflictuality is brought about by means of the foregrounding of the expression of attitudes in a context of miti- gated speaker's entitlement and restricted or relativized speaker's commitment. That is to say that this and similar cases of mitigation, though fundamentally aimed at the management of conflict (which is not itself a matter of illocutionary effect), can be described as based on variations in the degree of strength of the illocutionary acts involved, and therefore by reference to the deontic modal values mobilized by the core illocutionary effect.

4. C o n c l u d i n g r e m a r k s

I hope that the discussion of examples has shown that a wide range of mitiga- tion/reinforcement phenomena can be described in a revised speech-act theoretical framework envisaging illocutionary force as the language-triggered introduction of changes in the 'conventional' level of the interpersonal relation. Such a conception of the speech act appears to possess a descriptive or even explanatory potential that the standard conception lacks.

I would now like to append some brief remarks on the implications of the kind of analysis that I have been proposing for mitigation/reinforcement phenomena.

M. Sbis~t / Journal of Pragmatics 33 (2001) 1791-1814 1811

(i) It has often been noted that in actual discourse, mitigation is overridingly more frequent than reinforcement, and perhaps more interesting for pragmatic theory. This might lead us to study mitigation as a phenomenon in itself. In my proposal, on the other hand, mitigation and reinforcement are in principle symmetrical: they are of the same nature, since both are to be explained as variations in the illocutionary effects of speech acts. This connection between mitigation and reinforcement is con- firmed by the observation that certain reinforcements (for example, the foreground- ing of the expression of attitudes) combine with mitigations (for example, the miti- gation of the speaker's entitlement) to yield a complex result. This can be so because both mitigation and reinforcement affect the illocutionary effect and can be described in deontic modal terms.

(ii) Illocutionary acts need not exhibit a one-to-one correspondence with individ- ual, syntactically complete sentences (as has also been noted by Geis, 1995). Some- times, one illocutionary effect is brought about by a sequence of connected sen- tences, while at other times, one sentence may give rise to a cluster of illocutionary effects. Against the standard conception of the illocutionary act as basically reflect- ing the communicative function of the sentence, it is consistent with the approach defended here to say that illocutionary effects are produced by textual units consid- ered as structured wholes. This should not be taken to mean, however, that each tex- tual unit has to be internally consistent and homogeneous. It may well happen that the overall physiognomy of a textual unit emerges from a heterogeneous structure. In our example (5), discussed above, there is at least one feature which cannot be traced back to mitigation: "certamente" ('certainly'). Its collocation after the connective ma ('but') and before the formulation of the speaker's main claim ("un suicidio con grandi responsabilit~ istituzionali", 'a suicide which the institutions are largely responsible for') hints at a reinforcement of the speaker's commitment to that spe- cific claim, which might seem inconsistent with the mitigating devices that precede or follow. Still, we feel that the speaker is not inconsistent. It is no coincidence that soon after the (rash) reinforcing device certamente ('certainly') it becomes conve- nient, if not compulsory, for him to redefine his speech act as an expressive ("credo", 'I think'). The mitigating devices used counterbalance the reinforcement of the speaker's main claim without canceling it.

(iii) As to the relationship between illocutionary force indicating devices and mit- igation/reinforcement devices, the conception of mitigation/reinforcement proposed here supports the hypothesis of a fuzzy border between these two groups of linguis- tic expressions or strategies, with overlaps. However, it is not merely the case that some linguistic expressions are 'multifunctional', serving the distinct functions of mitigating/reinforcing speech acts and of facilitating the recognition of their illocu- tionary force (Blum Kulka, 1985). At least in some of our examples (2), (4), (5), (14), reinforcement or mitigation are not only associated with the use of certain lin- guistic expressions which also indicate illocutionary force, but are part of the dynamics by which illocutionary force is determined. It is the overall physiognomy of a speech act that permits us to take it (and describe it) as having a certain kind of illocutionary force, and which at the same time sets the degree of strength or inten- sity of the various components of the illocutionary effect: you cannot have the one

1812 M. Sbisd* / Journal of Pragmatics 33 (2001) 1791-1814

thing without the other. Each linguistic expression or textual strategy participating in the production of the illocutionary effect contributes in its own way to the overall physiognomy of the speech act. So if the task of discourse analysis is to identify the features of a text which motivate (or disconfirm) the reading that text is given, an analysis of illocutionary acts must appeal to all the features in the text which con- tribute to bringing about a particular illocutionary effect, whether they are standard illocutionary force indicators or indicators of a finer adjustment and tuning.

(iv) Since mitigation and reinforcement have obvious connections with politeness, the view I have put forward suggests some refinements for politeness theory. One widely accepted idea about politeness is that it is strategically connected with mat- ters of 'face' (Goffman, 1967; Brown and Levinson, 1987). But as used in pragmatic research, 'face' is a highly ambiguous notion. Brown and Levinson (1987: 61-62) give a general definition of it in terms of wants that participants have, that is, in terms of psychological attitudes, while in their own (and others') finer descriptions it turns out to be related to notions such as imposition, freedom from imposition, right, claim, debt, respect, competence, ratification or approbation of desires, legiti- mate expectation and violation of expectations, which are all either directly deontic or analyzable in deontic modal terms. If mitigation/reinforcement phenomena are to be analyzed in deontic modal terms, as I have claimed, this ambiguity in the notion of 'face' should be resolved in favor of the centrality of deontic modal notions as opposed to psychological attitudes. It would then be seen more clearly why and how variations in the degree of strength of speech acts play such a central role in polite- ness strategies. This might in turn induce us to recognize that the relationship between politeness strategies and illocution is closer than is usually admitted.

References

Austin, John L., 1962. How to do things with words. Second revised edition 1975. Oxford: Oxford Uni- versity Press.

Bach, Kent and Robert M. Hamish, 1979. Linguistic communication and speech acts. Cambridge, MA: MIT Press.

Bazzanella, Carla, 1994. Le facce del parlare. Firenze: La Nuova Italia. Bazzanella, Carla, Claudia Caffi and Marina Sbish, 1991. Scalar dimensions of illocutionary force. In:

I.~. ~agar, ed., Speech acts: Fiction or reality?, 63-76. Ljubljana: IPrA. Blum-Kulka, Shoshana, 1985. Modifiers as indicating devices: The case of requests. Theoretical Lin-

guistics 12: 213-229. Blum-Kulka, Shoshana, Juliane House and Gabriele Kasper, eds., 1989. Cross-cultural pragrnatics:

Requests and apologies. Norwood, NJ: Ablex. Brandom, Robert B., 1994. Making it explicit. Cambridge, MA: Harvard University Press. Brown, Penelope and Stephen C. Levinson, 1987. Politeness. Some universals in language usage. Cam-

bridge: Cambridge University Press. Caffi, Claudia, 1999. On mitigation. Journal of Pragmatics 31: 881-909. Diamond, Julie, 1996. Status and power in verbal interaction. Amsterdam: Benjamins. Dressier, Wolfgang and Lavinia Merlini Barbaresi, 1994. Morphopragmatics. Berlin: de Gruyter. Duranti, Alessandro, 1988. Intention, language, and social action in a Samoan context. Journal of Prag-

matics 12: 13-33. Fraser, Bruce, 1980. Conversational mitigation. Journal of Pragmatics 4: 341-350. Frege, Gottlob, 1879. Begriffsschrift. Halle: Nebert.

M. Sbisd /Journal of Pragmatics 33 (2001) 1791-1814 1813

Frege, Gottlob, 1918. Der Gedanke. Eine logiscbe Untersuchung. Beitr~ge zur Philosophie des deutschen Idealismus 1(2): 58-77.

Ganker, Christopher, 1994. Thinking out loud. An essay on the relation between language and thought. Princeton, NJ: Princeton University Press.

Gavioli, Laura and G. Mansfield, 1990. The PIXI corpora. Bookshop encounters in English and Italian. Bologna: Clueb.

Geis, Michael L., 1995. Speech acts and conversational interaction. Cambridge: Cambridge University Press.

Goffman, Erving, 1967. Interaction ritual: Essays on face to face behavior. New York: Garden City. Goffman, Erring, 1981. Forms of talk. Philadelphia, PA: University of Pennsylvania. Goodwin, Charles, 1995. The negotiation of coherence within conversation. In: M.A. Gernsbacher and

T. Giv6n, eds., Coherence in spontaneous text, 117-137. Amsterdam: Benjamins. Greimas, Algirdas J., 1983. Du sens 1I. Paris: Seuil. Greimas, Algirdas J. and Joseph Court6s, 1982. Semiotics and language. An analytical dictionary (Eng-

lish translation). Bloomington, IN: Indiana University Press. Harnish, Robert M., 1994. Mood, meaning and speech acts. In: S.L. Tsohatzidis, ed., Foundations of

speech act theory, 407-459. London: Routledge. Holmes, Janet, 1984. Modifying illocutionary force. Journal of Pragmatics 8: 345-365. Hugues, George E. and M.J. Cresswell, 1996. A new introduction to modal logic. London: Routledge. Katriel, Tamar and Marcelo Dascal, 1989. Speaker's commitment and involvement in discourse. In: Y.

Tobin, ed., From sign to text, 275-295. Amsterdam: Benjamins. Labov, William and David Fanshel, 1977. Therapeutic discourse. New York: Academic Press. Landowski, Eric, 1989. La soci6t6 r6fl6chie. Paris: Seuil. Leonardi, Paolo and Maurizio Viaro, 1990. Conversazione e terapia. Milano: Cortina. Lyons, John, 1977. Semantics. Cambridge: Cambridge University Press. Sadock, Jerrold, 1994. Toward a grammatically realistic typology of speech acts. In: S. Tsohatzidis, ed.,

Foundations of speech act theory, 393-406. London: Routledge. Sadock, Jerrold and A.M. Zwicky, 1985. Speech act distinctions in syntax. In: T. Shopen, ed., Language

typology and syntactic description, vol. 1: Clause structure, 151-195. Cambridge: Cambridge Uni- versity Press.

Savigny, Eike von, 1988. The social foundations of meaning. Berlin: Springer. Sbis~, Marina, 1984. On illocutionary types. Journal of Pragmatics 8:93-112. Sbish, Marina, 1987. Speech acts and context change. In: T. Ballmer and W. Wildgen, eds., Process lin-

guistics, 252-279. Ttibingen: Niemeyer. Sbis~, Marina, 1989. Linguaggio, ragione, interazione. Per una teoria pragmatica degli atti linguistici.

Bologna: I1 Mulino. Sbish, Marina, 1990. Speech acts and the expression of affect. Grazer Linguistische Studien 33/34:

279-295 Sbish, Marina, 1992. Affetto e diritto come dimensioni dell'interazione verbale. In: C. Galimberti, ed.,

La conversazione. Prospettive sull'interazione psico-sociale, 185-203. Milano: Guerini. Schegloff, Emanuel, 1992. Repair after next turn: The last structurally provided defense of intersubjec-

tivity in conversation. Americal Journal of Sociology 97: 1295-1345. Searle, John R., 1969. Speech acts. Cambridge: Cambridge University Press. Searle, John R., 1975. Indirect speech acts. In: P. Cole and J.L. Morgan, eds., Syntax and semantics 3:

Speech acts, 59-82. New York: Academic Press. Searle, John R., 1976. A classification of illocutionary acts. Language and Society 5: 1-23. Searle, John R., 1979. Expression and meaning. Cambridge: Cambridge University Press. Searle, John R. and Daniel Vanderveken, 1985. Foundations of illocutionary logic. Cambridge: Cam-

bridge University Press. Sperber, Dan and Deirdre Wilson, 1986. Relevance. Second edition 1995. Oxford: Blackwell. Strawson, Peter F., 1964. Interaction and convention in speech acts. The Philosophical Reviewer 73:

439-460. Streeck, Jiirgen, 1980. Speech acts in interaction: A critique of Searle. Discourse Processes 3: 133-154. Sweetser, Eve, 1990. From etymology to pragmatics. Cambridge: Cambridge University Press.

1814 M. Sbisdt / Journal of Pragmatics 33 (2001) 1791-1814

Tsohatzidis, Savas, ed., 1994. Foundations of speech act theory. London: Routledge. Wittgenstein, Ludwig, 1953. Philosophische Untersuchungen (with English translation). Oxford: Black-

well. Wright, Georg H. von, 1972. An essay in deontic logic and the general theory of action. Amsterdam:

North Holland.

Marina Sbis/l has studied philosophy of language and semiotics, with particular attention to ordinary language philosophy and to pragmatic issues such as speech acts, presupposition and implicature. Her research interests include the use of pragmatic notions in discourse analysis. She has co-edited J.L. Austin, How to do things with words (with J.O. Urmson, 1975). Her recent publications in English include 'The feminine subject and female body in discourse about childbirth' (European Journal oJ Women's Studies 3(4), 1996) and 'Presupposition, implicature and context in text understanding' (in: P. Bouquet et al., eds., Modeling and using context, Berlin: Springer, 1999). She teaches Philosophy of Language at the University of Trieste.