Autobiographical Memory and Autoneotic Consciousness in a case of Semantic Dementia

22
AUTOBIOGRAPHICAL MEMORY AND AUTONOETIC CONSCIOUSNESS IN A CASE OF SEMANTIC DEMENTIA Pascale Piolino Inserm-Université de Caen, CHU Côte de Nacre, Caen, and Université René Descartes, Paris 5, France Serge Belliard CHU Pontchaillou, Rennes, France Béatrice Desgranges Inserm-Université de Caen, CHU Côte de Nacre, Caen, France Mélisa Perron CHU Pontchaillou, Rennes, France Francis Eustache Inserm-Université de Caen, CHU Côte de Nacre, Caen, and École Pratique de Hautes Études, Universitaire René Descartes, Paris 5, France Investigations of retrograde amnesia have contributed to a better understanding of the cerebral struc- tures involved in remote memory. Such studies have suggested that neocortical regions such as the ante- rior temporal lobe play a major role in both the storage and retrieval of remote episodic and semantic information. Semantic dementia (SD), characterised as a focal anterior temporal lobe atrophy, offers an opportunity to study episodic remote memory, especially in the absence of day-to-day memory dysfunctioning, which takes place in permanent amnesic syndromes. Few studies have investigated autobiographical retrograde amnesia in SD. We present the findings from a patient (AT) at the early stage of SD. First, we have compared episodic and semantic components of autobiographical memory using two specially designed fluency tasks. The results demonstrated good recall of autobiographical events from all time periods and poor retrieval of names of acquaintances, albeit to a lesser degree, with respect to recent life. Second, we have investigated strictly episodic autobiographical memory and autonoetic consciousness by means of a sophisticated autobiographical test and the Remember/Know procedure which used a more stringent criterion of episodicity. The results demonstrated a relatively good recall of autobiographical memories (whatever their nature) but poor retrieval of remote specific detailed memories compared to recent ones. Moreover, patient AT provided Remember judgements to the same extent as control subjects regardless of the time interval covered although his responses were not justified in terms of the actual contextual information retrieved beyond the last 5 years. These find- ings provide further evidence that strictly episodic recollection is restricted to the recent past in SD. These data are discussed according to their relevance to the episodic and semantic distinction and to models of long-term memory consolidation. COGNITIVE NEUROPSYCHOLOGY, 2003, 20 (7), 619–639 Ó 2003 Psychology Press Ltd http://www.tandf.co.uk/journals/pp/02643294.html DOI:10.1080/02643290242000899 619 Requests for reprints should be addressed to Pr Francis Eustache, Inserm-Université de Caen E0218, Laboratoire de Neuropsychologie, CHU Côte de Nacre, 14033 Caen Cedex, France (Email: [email protected]). The authors would like to thank Dr A. Young for reviewing the English style, Dr V. de la Sayette for the description of neuroimaging findings, and the two reviewers for their helpful comments.

Transcript of Autobiographical Memory and Autoneotic Consciousness in a case of Semantic Dementia

AUTOBIOGRAPHICAL MEMORY AND AUTONOETIC

CONSCIOUSNESS IN A CASE OF SEMANTIC

DEMENTIA

Pascale PiolinoInserm-Université de Caen, CHU Côte de Nacre, Caen, and Université René Descartes, Paris 5, France

Serge BelliardCHU Pontchaillou, Rennes, France

Béatrice DesgrangesInserm-Université de Caen, CHU Côte de Nacre, Caen, France

Mélisa PerronCHU Pontchaillou, Rennes, France

Francis EustacheInserm-Université de Caen, CHU Côte de Nacre, Caen,

and École Pratique de Hautes Études, Universitaire René Descartes, Paris 5, France

Investigations of retrograde amnesia have contributed to a better understanding of the cerebral struc-tures involved in remote memory. Such studies have suggested that neocortical regions such as the ante-rior temporal lobe play a major role in both the storage and retrieval of remote episodic and semanticinformation. Semantic dementia (SD), characterised as a focal anterior temporal lobe atrophy, offers anopportunity to study episodic remote memory, especially in the absence of day-to-day memorydysfunctioning, which takes place in permanent amnesic syndromes. Few studies have investigatedautobiographical retrograde amnesia in SD. We present the findings from a patient (AT) at the earlystage of SD. First, we have compared episodic and semantic components of autobiographical memoryusing two specially designed fluency tasks. The results demonstrated good recall of autobiographicalevents from all time periods and poor retrieval of names of acquaintances, albeit to a lesser degree, withrespect to recent life. Second, we have investigated strictly episodic autobiographical memory andautonoetic consciousness by means of a sophisticated autobiographical test and the Remember/Knowprocedure which used a more stringent criterion of episodicity. The results demonstrated a relativelygood recall of autobiographical memories (whatever their nature) but poor retrieval of remote specificdetailed memories compared to recent ones. Moreover, patient AT provided Remember judgements tothe same extent as control subjects regardless of the time interval covered although his responses werenot justified in terms of the actual contextual information retrieved beyond the last 5 years. These find-ings provide further evidence that strictly episodic recollection is restricted to the recent past in SD.These data are discussed according to their relevance to the episodic and semantic distinction and tomodels of long-term memory consolidation.

COGNITIVE NEUROPSYCHOLOGY, 2003, 20 (7), 619–639

� 2003 Psychology Press Ltdhttp://www.tandf.co.uk/journals/pp/02643294.html DOI:10.1080/02643290242000899

619

Requests for reprints should be addressed to Pr Francis Eustache, Inserm-Université de Caen E0218, Laboratoire deNeuropsychologie, CHU Côte de Nacre, 14033 Caen Cedex, France (Email: [email protected]).

The authors would like to thank Dr A. Young for reviewing the English style, Dr V. de la Sayette for the description ofneuroimaging findings, and the two reviewers for their helpful comments.

INTRODUCTION

The term semantic dementia (SD) was introducedrelatively recently by Snowden, Goulding, andNeary (1989) and since then has been used, andelaborated on, by other authors (Hodges,Patterson, Oxbury, & Funnell, 1992) to character-ise patients with semantic disorders and preserva-tion of autobiographical and day-to-day episodicmemory in the context of unilateral or bilateralanterior temporal lobe atrophy. Anecdotal reportson SD emphasised the fact that recall of recentevents is good whereas semantic knowledgebecomes restricted and highly personalised.Semantic dementia is characterised by profiles ofneuropsychological and neuroanatomical damagethat are opposite to those found in an amnesic syn-drome, in which disorders of episodic memory andlesions of the medial temporal lobe and surround-ing structures are noted. Moreover, these two syn-dromes are also opposed concerning their pattern ofretrograde amnesia: a temporally graded memoryloss in accordance with Ribot’s law in an amnesicsyndrome and a reversal gradient sparing recentmemory in SD (Graham & Hodges, 1997; Hodges& Graham, 1998; Nestor, Graham, Bozeat,Simons, & Hodges, 2002; Snowden, Griffiths, &Neary, 1996). However, studies of remote memoryin SD are still scarce and the nature of remotememory remains controversial (Moscovitch,Yaschyshyn, Ziegler, & Nadel, 1999; Nestor et al.,2002).

Snowden, Griffiths, and Neary (1994, 1995,1996) conducted the first studies of remote memoryin four patients with SD. Their findings suggestthat the recent past is more preserved than theremote past for semantic memory (such as famousnames and money systems) and for autobiographi-cal memory. For example, by studying three timeperiods (recent, past, and historic), with a surname/first-name matching task for celebrities, theseauthors have shown that the recent period is betterremembered than the past period and the pastperiod is better remembered than the historicperiod. Through the use of an autobiographicalmemory interview (devised by Kopelman, Wilson,& Baddeley, 1989) for personal semantic memory

and autobiographical incidents, Snowden et al.(1996) found that the performances for the recentperiod (last year) were better than for the tworemote periods studied (childhood and early adult-hood). Patients with Alzheimer’s disease at thestage of mild dementia showed a contrary pattern ofperformance for personal semantic and autobio-graphical incidents, with cognitive scores beingnumerically lower for the most recent past com-pared to childhood and early adulthood. Thesesame authors (Snowden et al., 1996) also carriedout a qualitative evaluation of retrograde autobio-graphical memory with photographic cues from apatient’s own family life and with semantic cuesgenerated from their own conversational output,and showed that patients reported some factualinformation, such as the names of people andplaces, as well as details of specific incidents, such asa visit from a friend about recent life. Nonetheless,early and middle-life information were more lim-ited in detail but varied in degree across thepatients. Taken together, these results are in keep-ing with the prediction of preserved current per-sonal relevant material in SD. Similar findings wereconfirmed in groups of SD with a famous namestest (Hodges & Graham, 1998) and an autobio-graphical memory interview (Graham & Hodges,1997; Nestor et al., 2002).

To clarify the interpretation of data concerningthe extent of the period of preservation of episodic-autobiographical memories in SD, Graham andHodges (1997) and, more recently, Nestor et al.(2002) used the Galton-Crovitz paradigm (Crovitz& Schiffman, 1974) together with a fine-grainedscoring chart, for two single cases (patients AM andJH, respectively). The results demonstrated thatthe majority of memories came from the last 2 or 3years and were specific, whereas remote memorieswere very rare and generic, or concerned cue-wordsdefinitions. However, methodological biasesinherent in the cue-word technique (i.e., the artifi-cial demands of the task, such as producing a per-sonal memory to a particular cue word that is notalways relevant to personal life) may have affectedthe results (Dritschel, Williams, Baddeley, &Nimmo-Smith, 1992; Rabbitt & Winthorpe,1988). These findings suggest that the anterior

620 COGNITIVE NEUROPSYCHOLOGY, 2003, 20 (7)

PIOLINO ET AL.

temporal cortex, which is atrophic in SD, is a repos-itory of remote memory whatever the kind of infor-mation involved (episodic or semantic).Furthermore, it is possible that these results havesome implications concerning a structural model ofautobiographical memory and may suggest that therecent past is episodic while the remote pastbecomes semantic (Cermak, 1984). Nonetheless,Graham and Hodges (1997) noted that the veryshort-term preservation of autobiographical mem-ory (for only a period of a couple of years) cannotreally be explained by a long-term semantisationprocess, unless this process occurs very rapidly. Thishypothesis may provide an explanation for the pres-ervation of remote autobiographical memory inclassic amnesic syndrome (Butters & Cermak,1986; Schmidtke & Vollmer, 1997). However, asSnowden et al. (1996) first suggested, patients withSD may have residual knowledge and memoriesabout the past that are not effectively tapped byconventional probes of a standard retrograde mem-ory assessment task. Thus, Moscovitch and Nadel(2000) argue that nominal deficits and a lack ofunderstanding of concepts, especially those appli-cable to early life (i.e., semantic prompts such asfriends, school, Christmas), which are used toprobe autobiographical memory in standard meth-ods of evaluation, are two features of the disorderthat may explain low levels of remote scores elicitedin patients with SD. Moss et al. (Moss, Capelletti,De Mornay Davies, Jaldow, & Kopelman, 2000)showed, in a patient SD, that specific lexical-semantic cues revealed memories (without anysignificant temporal gradient) that would otherwisehave been underestimated. Recently, Westmacottet al. (Westmacott, Leach, Freedman, &Moscovitch, 2001) explored autobiographicalmemory using a procedure based on family photo-graphs taken from infancy to the present as cues, inorder to minimise a verbal contribution. Onepatient with SD demonstrated a well-preservedmemory for autobiographical episodes regardless ofthe time period tested, whereas another patientwith an amnesic syndrome demonstrated a com-plete loss of episodes without temporal gradient.Unfortunately, performances of the two patientswere not compared to a control group. Thus, it is

not possible to assess if the performance of thepatient with SD was preserved when compared tocontrols at any given time period.

The aim of the present study is to address thequestion of strictly episodic remote memory in SDby investigating the retrieval of autobiographicalmemory in a single patient compared to a group ofcontrol subjects. First, the distinction between epi-sodic and semantic remote autobiographical mem-ory was examined with two fluency tasks to test thehypothesis that episodic memory is better preservedthan semantic memory (Nadel, Samsonovich,Ryan, & Moscovitch, 2000). Second, remote epi-sodic memories have been especially investigatedwith a sophisticated autobiographical memory testthat also assesses autonoetic consciousness, whichis critically involved in episodic recollection(Wheeler, Stuss, & Tulving, 1997). Autonoeticconsciousness may be defined as a feeling of re-experiencing or reliving the past by mentally travel-ling back in subjective time. The evaluation of auto-biographical memory (which takes into account theexistence of a sense of remembering) would be amore appropriate assessment technique in SD thanpreviously used methodologies (Westmacott et al.,2001).

CASE REPORT

A patient with SD according to the diagnostic cri-teria of Neary et al. (1998) has been under examina-tion at the Neurology Unit of the UniversityHospital of Rennes, France, since December 1999.Patient AT, 69 years old, right-handed, was a fur-rier until his retirement 3 years ago. According tohim and his wife, word-finding difficulties began ataround this time. These difficulties principally con-cerned problems in the recall and recognition ofproper names (persons, streets, trademarks) and offaces (personal acquaintances and famous people).He also showed, although to a lesser degree, someword comprehension difficulties, but his languagewas fluent with no grammatical or phonologicalimpairment. He could correctly repeat isolatedwords and concrete and abstract sentences. In con-trast, patient AT had no problem concerning his

COGNITIVE NEUROPSYCHOLOGY, 2003, 20 (7) 621

AUTOBIOGRAPHICAL MEMORY IN SEMANTIC DEMENTIA

day-to-day memory: he never forgot appointments,he was still autonomous and well oriented in spaceand time, he could do his shopping, and could takepart in everyday activities. Otherwise, his wifenoticed a tendency to rigidity in his thinking andobsessional traits. Anatomical MRI (Figure 1)revealed a striking and asymmetric enlargement ofthe temporal horns of both lateral ventricles, whichcoincided with a bilateral, but major right-sided,temporal neocortical atrophy and a little less promi-nent tissue loss of the medial temporal structures.

GENERAL COGNITIVEASSESSMENT

Procedure

A general cognitive examination was used toexplore semantic memory by means of categorical(names of animals) and phonemic fluency tasks(names beginning with the letter p) (Cardebat,Doyon, Puel, Goulet, & Joanette, 1990), twosemantic knowledge tasks, and the French celebri-ties questionnaire. The semantic knowledge task,according to Giffard et al. (2001) and derived from

Desgranges et al. (Desgranges, Eustache, Rioux,De la Sayette, & Lechevalier, 1996), involves threecomponents (naming, categorical knowledge, andattribute knowledge derived from names) on 30concepts belonging to four categories (animals,plants, objects, body parts). The semantic categori-sation battery (Belliard et al., personal communica-tion) consists of three subtests of categorisation(categorical, structural, and functional), concerningnames and pictures of 32 biological and 32 man-made objects. The French celebrities questionnaire(Belliard, Perron, Rouyer, Golfier, Sartori, &Edan, 2001) was constructed for patients withsemantic dementia and is an easy questionnaire ofvery famous people that consists of a subtest offamiliarity decision and a questionnaire based onrecognition tasks (dead/alive; nationality; profes-sion; specific knowledge) concerning 10 celebrities(e.g., J. P. Belmondo, A. Delon), most of them stillalive, presented by name and photograph. The firststep consists of a familiarity decision task concern-ing the famous persons presented by name. Thesecond step is a questionnaire concerning thenames classified as familiar and the third and fourthsteps consist of a familiarity decision task then aquestionnaire concerning the faces of the famous

PIOLINO ET AL.

622 COGNITIVE NEUROPSYCHOLOGY, 2003, 20 (7)

Figure 1. MRI T2-weighted coronal slices of patient AT showing a bilateral atrophy particularly affecting the temporal lobes and the righthemisphere. On the right, the atrophy of the temporal pole, inferior and lateral temporal cortex, as well as the huge dilation of the temporalhorn, are dramatic whereas the involvement of the medial temporal structrures is a little less prominent. The enlargement of the rightsylvian fissure and both frontal horns of the lateral ventricles are noticeable (right hemisphere on the left side of the slices and vice versa).

persons. Language was assessed by means of a pic-ture naming task (DO 80: Deloche & Hannequin,1997), where the subject had to name 80 black-and-white line drawings of familiar objects and theToken Test (De Renzi & Faglioni, 1975). Anepisodic verbal memory test, derived from Groberand Buschke’s procedure (1987), was conductedand involved a list of 15 words coming from the 30concepts used on the semantic knowledge task thatwere not semantically degraded (Eustache et al.,2001). Episodic visual memory was assessed bymeans of the visual retention test (Benton, 1982).Furthermore, short-term memory was evaluated bya digit span (Wechsler, 1991), executive functionsby the Wisconsin Card Sorting test (Nelson, 1976),and visuospatial abilities were measured by copyingof the Rey figure (Rey, 1960). Lastly, the subject’sgeneral intellectual level was assessed by the PM 47(Raven, Court, & Raven, 1998).

Results

Patient AT showed a normal general intellectuallevel (percentile 95) as assessed by the PM 47. Theformal neuropsychological investigation (reportedin Table 1) exhibited mild to moderate semanticmemory difficulties with anomia and an impover-ished general knowledge for concepts (viz., in thesemantic knowledge task, the semantic categorisa-tion battery, and the naming task) and for celebri-ties. Moreover, a trend towards impairment wasobserved on the category fluency task. In thesemantic knowledge task, patient AT was espe-cially impaired on the naming component revealingword-finding difficulties even when the conceptwas correctly identified from names (i.e., drome-dary, sheep, steeple, scarf). In the semantic cate-gorisation battery in which 64 concepts werescreened, he was impaired by all kinds of material(pictures or words), of category (inanimate or ani-mate) and of classification (structural or func-tional). Patient AT classified reindeer as a tool,apricot and green pepper as cooking instruments(when presented by name), and sheep, goat, andturkey as wild animals, when these were presentedas pictures. Superordinate categorisation was rela-tively better preserved compared to attribute cate-

gorisation. In the familiarity decision task and inthe recognition task of the French celebrities ques-tionnaire, his performance was abnormal for namesand for faces. For example, when presented byname, he failed to consider Prince Charles as wellknown. Even if a name (or a face) evoked a sense offamiliarity, more often than not he failed to identifythe person if no more information was available.

In contrast, his performance was perfect on syn-tactic comprehension (assessed by the Token Test).Short-term memory (digit span), executive func-tioning (assessed by the Wisconsin Card SortingTest), and visuospatial abilities (copy of the Rey fig-ure) were preserved. Anterograde episodic visualand verbal memory was also relatively preserved(Benton Test and the task derived from the Groberand Buschke procedure, respectively), althoughsome scores reached pathological limits (total recalland delayed recall in the episodic verbal task) giventhe small standard deviations in the control group.

EXPERIMENTALAUTOBIOGRAPHICAL MEMORYINVESTIGATION

Two tests were applied specifically to explore auto-biographical memory (Piolino, Desgranges, &Eustache, 2000): first, an autobiographical fluencytest that evaluated semantic and episodic compo-nents of autobiographical memory, and second, astrictly episodic autobiographical test associatedwith the Remember/Know procedure, which per-mits a more controlled assessment of episodicmemory. This procedure allows one to examine thecognitive processes underlying the recollection ofautobiographical events by separating the episodicfrom the semantic contributions to the memory test(Tulving, 1985) and to assess if a subject trulyrelived an episode.

Episodic and semantic autobiographicalfluency tasks

Procedure

Two autobiographical fluency tasks, derived fromDritschel et al. (Dritschel, Williams, Baddeley, &

COGNITIVE NEUROPSYCHOLOGY, 2003, 20 (7) 623

AUTOBIOGRAPHICAL MEMORY IN SEMANTIC DEMENTIA

Nimmo-Smith, 1992) and Greene et al. (Greene,Hodges, & Baddeley, 1995), were assigned to eval-uate the semantic component and the episodiccomponent of remote autobiographical memory forfive age periods: 0–17 years old, 18–30 years old,more than 30 years old, except for the last 5 years,

the last 5 years except for the last 12 months,and the last 12 months. For patient AT, these threelast time periods corresponded to 31–63 years old,64–68 years old, and 69 years old, respectively. Foreach period, the participant was given 2 min toretrieve as many items as possible for: (1) personal

PIOLINO ET AL.

624 COGNITIVE NEUROPSYCHOLOGY, 2003, 20 (7)

Table 1. Neuropsychological results: Data shown are raw scores with statistical significance according toz-scores for patient AT and mean ±SD for control subjects

Tests Patient AT Controls

Semantic memoryCategory fluency 16 27.1 (8.5)Phonemic fluency 14 19.3 (7.0)Semantic Knowledge Task (SKT)

Total score (/236) 227* 232.0 (2.0)Naming (/60) 56*** 59.3 (0.7)Categorical knowledge (/86) 85 84.7 (1.0)Attribute knowledge (/90) 86 87.9 (1.2)Semantic Categorisation Battery (SCB)Total scores (/1200) 1103*** 1185.1 (13.1)Word Score (/800) 734*** 789.2 (5.0)Picture score (/400) 369*** 395.9 (4.6)

French celebrities questionnaireFamiliarity decision task (/10) name: 9* 9.9 (0.3)

face: 6*** 9.9 (0.3)Recognition tasks (%) name: 26.6*** 88.9 (14.0)

face: 37.5*** 98.5 (1.9)LanguageNaming task (DO 80) 57*** 69

b

Token test (/36) 35 29b

Episodic memoryImmediate recall (/15)

a15 15 (0)

Total recall (/45)a

41* 43.8 (1.2)Delayed recall (/15)

a14** 14.9 (0.3)

Recognition (/15)a

13 14.6 (1.1)False recognition

a0 0.5 (0.8)

Benton testRecall (/10) 9 7

b

Delayed recall (/15) 15 11b

Short-term memoryDigit span

Forward 6 7.2 (1.9)Backward 7 5.9 (2)

Executive functionsWisconsin

Category number 6 6 (1.1)Perseveration 4 1.2 (2)

Visuospatial abilitiesRey figure copy (/36) 35 32

b

aGrober and Buschke modified task.

bPathological cut-off.

Pathological score: *p < .05; **p < .01; ***p < .001.

semantic information (semantic fluency) such asnames of acquaintances (surnames or first names ofteachers, friends, colleagues) and (2) personal auto-biographical specific single events (episodic fluency).For the two tasks, he was requested to recall infor-mation from each time period, from remote torecent past. Specific instructions were given beforeeach task. For semantic fluency, the subject wasinformed that he must recall the names of acquain-tances specific to one period and he must not givethe names of family members. For episodic fluency,the subject had to recall single autobiographicalincidents that lasted for less than 1 day and were sit-uated in a spatio-temporal and a phenomenologicalcontext (he could re-experience the incident),but instructions emphasised that no detail wasrequired. He had just to say in a few words the event(for example, I broke my leg skiing down a slope)and then pass rapidly to another event, but he wastold that should he be instructed to detail the event,then he could do so. The score for each fluency taskwas the total number of items given per time period,excluding any repetitions, or errors of time period,and any obvious generic events that were repeated(“going out with friends”) or extended (“the journeyto Paris”). Errors were highlighted by a retest,where the subject had to say the date they met eachperson and the date of each specific single event.

Results

Autobiographical fluency performances of patientAT (Figure 2a) were compared to those of 12 con-trols (mean age 70.4 ± 4.4; range 65–75 years)paired according to age and education level (mini-mum level of 8 years of primary education) andrecruited in homes for retired people. For eachscore recorded, the z-score procedure was used todetermine whether there was a significant differ-ence in performance between patient AT and thecontrol participants for each of the five time periodstested.

Results of the control group, carried out with ananalysis of variance (ANOVA), demonstrated asignificant time period effect for the semantic flu-ency, F(4, 44) = 9.51; p < .0001, and for the episodicfluency, F(4, 44) = 5.55; p < .01. Post hoc analysis(PLSD Fisher tests) indicated a lesser recall of

names for the last 5 years and the last 12 monthsperiods compared to the other time periods (p < .05to p < .0001) and a lesser recall of specific events forthe last 12 months period compared to the othertime periods (p < .05). Results of semantic fluency forpatient AT showed statistically significant differ-ences (according to z-scores) between his perfor-mance and those of the control group for each of thethree earliest time periods (0–17 years old, 18–30years old, and 31–63 years old), whereas his perfor-mances were normal compared to control partici-pants for the two most recent time periods (last 5years and last 12 months). Otherwise, episodic flu-ency was relatively well preserved compared to con-trol subjects whatever the time period assessed. A3-way ANOVA (group, time-period, task)revealed simple effects of group, F(1, 11) = 8.22, p <.05, and task, F(1, 11) = 10.53, p < .01, and a signifi-cant task by group interaction, F(1, 11) = 16.53, p <.01, showing that the performances of the patientAT (Figure 2b) were inferior to those of the con-trols for the semantic fluency (recall of names) butsimilar for the episodic fluency (recall of specificevents). The interaction group by period nearlyreached significance, F(4, 44) = 2.17, p = .09, sug-gesting that the patient AT tends towards a reversetime period effect seen in controls, i.e., a betterpreservation of recent compared to remote pastindependently of the kind of information supplied(names or events).

Strictly episodic autobiographical test

Procedure

Material and instructions. The strictly episodicautobiographical test is a semistructured question-naire based on those of Kopelman et al. (1989),Borrini et al. (Borrini, Dall’Ora, Della Sala,Marinelli, & Spinnler, 1989), and Piolino et al.(Piolino, Desgranges, Benali, & Eustache, 2002)which assesses the ability to recall detailed specificevents situated in space and time. Subjects wererequired to describe such personal events in as muchdetail as they could. The general instructions at thestart of the procedure made the nature of the recalltask clearer. The subject was asked to remember

COGNITIVE NEUROPSYCHOLOGY, 2003, 20 (7) 625

AUTOBIOGRAPHICAL MEMORY IN SEMANTIC DEMENTIA

PIOLINO ET AL.

626 COGNITIVE NEUROPSYCHOLOGY, 2003, 20 (7)

Figure 2. Patient AT’s autobiographical memory (semantic and episodic) compared to control subjects (mean and standard deviation) usingautobiographical fluency tasks, showing: (a) results of fluency tasks according to the time period examined; and (b) overall results of fluencytasks.

Pathological score: ** p < .01; *** p < .001.

0

2

4

6

8

10

12

14

16

18

20

22

0-17

years old

18-30

years old

> 30

years old

Last 5

years

Last 12

months

0-17

years old

18-30

years old

> 30

years old

Last 5

years

Last 12

months

Controls

Patient AT

*

**

*

*

*

*

Time period

Me

an

rate

dsc

ore

Semantic Fluency Episodic fluency

0

5

10

15

20

Semantic fluency Episodic fluency

Controls

Patient AT

Me

an

rate

dsc

ore

*

*

(a)

(b)

personally experienced events, and to situate themin time and space: “Describe, out loud and with asmuch detail as possible, what happened, as if youare able to relive it: What you did and felt, thecircumstances, with whom, where and how ithappened. If you are recalling, for example, holi-days at the seaside, you must avoid general descrip-tions, giving precise memories of a particular eventwhich happened on a day during these holidays.”During the task, the subject had in front of him asheet of paper indicating the period under consider-ation and the instructions. In particular, the defini-tion of the memory requested in the test (an episodewhich lasted less than a day and which he can relivewith details) and a checklist of questions he cananswer for helping him (What did he do or say?What happened? What were his thoughts, percep-tions, or emotions?) were provided. There is notime constraint but generally the test lasts 2 hours.There is a retest 15 ± 2 days after the test, whichlasts approximately 1 hour.

The same five time periods as those used in theepisodic and semantic autobiographical fluencytasks were investigated (0–17 years old, 18–30 yearsold, more than 30 years old except for the last 5years, the last 5 years except for the last 12 months,and the last 12 months). Each time period exam-ined, except the last one, was assessed by four topics(a meeting or an event linked to a person or a schooland then a professional event, a trip or a journey, afamilial event). For example: “Describe precisely aparticular event that took place during your familylife.” For the most recent period, eight questionsallowed a chronological study of the recent past (lastsummer, Christmas or New Year’s day, last month,last week, last weekend, 2 days ago, yesterday,today). Particular instructions were proposed to thesubject. For example, “describe precisely a particu-lar event that occurred last week.” If the participantcould not recollect a specific event spontaneously,cues were provided (e.g., on a day with a professoror a friend, in the primary or secondary school, inthe playground or during an exam) and/or encour-agements were given to be specific when the mem-ory was generic (e.g., do you remember a particularday?). After three trials of cueing and/or encourage-ment, another topic was proposed.

In addition, after each event recalled, the subjecthad to indicate his conscious experience accompa-nying the retrieval by means of the Remember/Knowprocedure (Gardiner, 1988, 2001; Tulving, 1985).He had to qualify his report by making a judgementof remembering, knowing, or guessing separatelyfor the content, the place, and the date of the event(what, where, when). A remember judgement meansthat the subject recollected a specific autobiograph-ical event as a re-experiencing of the source ofacquisition with details; re-experiencing an event aspart of one’s past is based on autonoetic awareness.A know judgement is like a feeling of familiarity inthe absence of recollection. This kind of responseis based on noetic awareness. A guess judgement,introduced by Mäntylä (1993), indicated that thesubject was not sure of his response. Thereafter, achecking procedure allowed one to assess if a sub-ject was able to justify each of his remember judge-ments, proving that he had effectively relived someaspects of the original event. Hence, for eachremember response provided, whatever the kind ofinformation (what, where, when), the subject hadto add contextual details from the original event ifhe had not already recalled them: his thoughts, feel-ings, or perceptions for the content (what), spatialdetails and his localisation for the place of event(where), and time of day or temporal sequence forthe date (when). The justified remember procedurewas rather similar to the approach of Moscovitch etal. (Moscovitch, Yaschyshyn, Ziegler, & Nadel,1999), who classified details retrieved for each cueword provided into various categories such as per-ceptual implication (visual-auditory), emotional,spatial, and temporal.

For the retest procedure, a cueing recall task wasproposed for each memory in a random order ask-ing the subject to recall the content and thespatiotemporal situation of the event. When it waspossible, each memory was checked by a member ofthe family.

Scoring. Each event was scored on a 4-point epi-sodic scale based on that of Baddeley and Wilson(1986). This scale took into account the specificityof the content (single or repeated event),spatiotemporal situation, and the presence of

COGNITIVE NEUROPSYCHOLOGY, 2003, 20 (7) 627

AUTOBIOGRAPHICAL MEMORY IN SEMANTIC DEMENTIA

details (perceptions, thoughts, feelings). A specificevent with details situated in time and space wasgiven a score of 4 points. A specific event withoutdetail situated in time and space scored 3 points. Arepeated, or extended, event scored 2 points or 1point according to whether or not it was situated intime and space. Absence of memory, or generalinformation about a topic, scored 0. Recollectionsscoring 3 or 4 points were consistent with the defi-nition of episodic memory, which refers to memoryfor a particular episode, lasting less than a day, andnot to a more general event, which scores 2 or 1point. The critical factor that allows one to differ-entiate between a specific event (scored 3) and aspecific detailed event (scored 4) is the inability,despite many encouragements, to add furtherdetails concerning the source of acquisition. Thus,two different total scores were recorded per timeperiod examined: (1) a global autobiographical score(max. 4 × 4 = 16), which included all the memories(specific and generic) and represents the standardepisodic memory score used in the AutobiographicalMemory Interview (AMI: Kopelman et al., 1989);(2) a strictly episodic score (max. 4 × 4 = 16) whichincludes only the specific and detailed memoriesscored 4 using a more stringent criterion. Takinginto account the specificity and the richness ofdetail allows one to assess if a subject truly relived anepisode (Moscovitch et al., 1999).

The Remember/Know procedure further permitsone to record two other total scores per time periodwhich were considered as episodic indicators: (1) aglobal remember score (max: 3 × 4 = 12) which corre-sponded to the total number of remember judge-ments provided altogether for the content, theplace, and the date of each event; (2) a global justifiedremember score (max: 3 × 4 = 12) which concernedthe number of remember judgements that wereeffectively associated with the recollection of a sin-gle event with contextual details (thoughts, feelingsor perceptions for the content, spatial details andlocalisation for the place of event and time of day ortemporal sequence for the date). Furthermore,these two episodic indicators might be studied dis-tinctly according to the kinds of information (what,where, when) considered. For example, “whatremember” score (max: 4) corresponded to the total

number per time period of remember responsesgiven for the content of each event.

For each score recorded, the results of the last 12months period, which were concerned with eightitems instead of four topics, were divided by two.Global autobiographical score, strictly episodic score,and justified remember scores were scored accordingto the information supplied in the test and corrobo-rated in the retest. Each memory was rated by twoindependent experts until a consensus was reached.

It is noteworthy that instructions of the strictlyepisodic autobiographical test were explained verycarefully to patient AT until each concept used inthis test was well understood. Furthermore, in anattempt to minimise the intervention of semanticdisorder, we did not take into account the capacityto produce specific semantic information such asthe names of people implicated in an event or thename of the street or the town where an event tookplace. All of the episodes produced by the patientwere checked with the retest procedure and thefamily, who confirmed the validity of all the memo-ries and thus established that there was no evidenceof confabulation or incorrect details.

Results

Performances of patient AT were compared tothose of 18 controls (mean age 69.4 ± 3.0; range 65–75 years), paired according to age and educationlevel (minimum level of 8 years of primary educa-tion), recruited from homes for retired people.Figure 3 illustrates the results for the global auto-biographical score and the strictly episodic score.

For the global autobiographical score, the controlgroup (Figure 3, top) showed a significant timeperiod effect, F(4, 68) = 7. 32; p < .0001, which cor-responded to the fact that the last 5 years period wasinferior compared to the other time periods. Incomparison, patient AT’s performance indicated areversal of the temporal Ribot’s gradient affectingmore remote memory compared to recent memory,but differences between AT’s performances and thecontrol group reached statistical significance(according to the z-score) only for the earliest timeperiod (0–17 years old). Though his conversationalrepertoire was often stereotyped with frequentrepetitions of an identical set of utterances (for

PIOLINO ET AL.

628 COGNITIVE NEUROPSYCHOLOGY, 2003, 20 (7)

example “we went for a walk”), patient AT providedsome information, either facts or events, pertainingto the different periods of his life, especiallydescriptive information about different residencesand professional activities, which thus indicatedthat he still knew the chronological order of differ-ent periods of his life and retained some personalsemantic knowledge. However, some information

such as the date of his wedding or the date of thebirth of his children was not produced spontane-ously but was inferred from current informationsuch as his recent wedding anniversary or thecurrent age of his children.

With respect to the strictly episodic score (Figure3, bottom), control subjects provided similarlystrictly episodic memories across all time periods,

COGNITIVE NEUROPSYCHOLOGY, 2003, 20 (7) 629

AUTOBIOGRAPHICAL MEMORY IN SEMANTIC DEMENTIA

Figure 3. Patient AT’s results on strictly episodic autobiographical memory test compared to control subjects (mean and SD): (a) global and(b) strictly episodic scores.

Pathological score: * p < .05; ** p < .01.

0

2

4

6

8

10

12

14

16

0-17 years old

18-30 years old

> 30 years old

Last 5 years

Last 12 months

**

Mean

rate

dscore

Time period

Global score

Controls

Patient AT

Mea

nra

ted

score

Time period

0

2

4

6

8

10

12

14

16

0-17 years old

18 years old

> 30 years old

Last 5 years

Last 12 months

* **

Strictly episodic score

(a)

(b)

F(4, 68) = 1.62; p = .18, whereas patient AT failedcompletely to recall such memories for the threeearliest time periods (which corresponded to thefirst 64 years of his life). In particular, some memo-ries that were generally well retrieved by controlsubjects, such as their wedding day or the birth oftheir children, were not recollected. However,patient AT was able to provide strictly episodicmemories to the same extent as control subjects forthe 12 last months period and to a lesser degree (notsignificant) for the last 5 years period. For example,for the item “last month” in the last 12 monthsperiod, patient AT recounted: “I went to Quiberon,we used to go Friday and we went for a walk on thecoast. Last month, one day with my wife, we wentto an island. We went at 9 am and returned at 5 pm.We walked around the island and we ate at theseaside. It’s very wild, we met nobody except oneperson at the end of the afternoon. When wewent back, the boat was late and we got a coffee ina café. . .” For the last 5 years period, the onlyspecific detailed event was from a memorable trip toCannes undertaken 4 years prior to testing: “beforemy retirement, I went by car to Grasse, 4 years ago,we spent a day at Cannes. At morning, we took thecar and we went for a walk on the Promenade desAnglais. At lunch time, we got a sandwich on abench. During the afternoon, we took a bus for vis-iting an underground cave. We must take carewalking along this famous river because, it’sslippery!. . .”

In order to clarify the results of patient AT, wehave studied each event score for all the periodsexamined. Figure 4a illustrates the percentagenumber of responses made by patient AT in eachevent scoring category. It was clear that in the mostrecent time period (last 12 months), patient AT wasmore likely to produce specific events (62.5%), withmost of them being detailed (50%). The pattern ofrecall for the last 5 years period (AT’s age 65-68)was rather heterogeneous, with a memory fromeach scoring category. In contrast, beyond the last 5years period, the majority of memories were generic(multiple events, repeated or extended) and only afew were specific without details (“I remember theday I was in my father’s tanker before crossing a

bridge. A policeman came to warn my father of agrave accident on the bridge.” Q: “Does this remindyou of any specific details?” A: “I just clearlyremember the policeman coming, I was 18”), thusindicating a change in quality of memories overtime. These specific memories (2 out of 12 remotememories provided) were scored 3 because patientAT was unable to provide details even with a gooddeal of encouragement. However, if we consideredthis time all specific memories (with and withoutdetails) which concern events that occurred onlyonce (though they may have been rehearsed later),findings indicated that patient AT could recall spe-cific events until the age period 18–30 years old,corresponding to a lifespan of 51 years (for the fivetime periods from the most recent the ratio was62.5%; 25%; 25%; 25%; 0%, respectively). On thecontrary, control participants scored 4 on at least40–50% of their memories. The results detailed foreach of the eight items tested for the last 12 monthsperiod (Figure 4b) indicated that patient AT wasable to describe detailed events about the previousdays (last month, 2 days ago, yesterday, and today)whereas he did not procure enough details abouta recent specific event such as Christmas Eve orNew Year’s Day or a recent time such as lastsummer or last week, but only vague events.Nevertheless, this pattern was similar to that of thecontrol subjects.

The results on the measures of state of awarenessare presented in Table 2 with pathological scoresindicated according to the values of z-scores. Find-ings showed that the global remember score was nor-mal whatever the time period examined whereasthe global justified remember score was at a pathologi-cal level for the three earliest time periods. Further-more, whatever the type of information (what,where, when), remember scores for patient AT dem-onstrated the same profile as control subjects. Incontrast, his justified remember scores were signifi-cantly impaired for each type of information for theearliest time period and thereafter only for thespatiotemporal information. Thus, phenomeno-logical details for the content of the event, (what)including emotion, perception, and thoughts, werebetter preserved than contextual details for the

PIOLINO ET AL.

630 COGNITIVE NEUROPSYCHOLOGY, 2003, 20 (7)

place and the date of the event. Figure 5 illustratesthe ratio of global justified remember judgements toglobal remember judgements made by patient AT asa function of the time period examined, showingthat patient AT could not provide as much episodic

details (feelings, thoughts or images, localisation ortime) as control subjects prior to the last 5 yearsperiod. Nevertheless, the 18–30 years old periodwas less impaired than the two surroundingperiods.

COGNITIVE NEUROPSYCHOLOGY, 2003, 20 (7) 631

AUTOBIOGRAPHICAL MEMORY IN SEMANTIC DEMENTIA

Figure 4. Responses made by patient AT on the strictly episodic autobiographical memory test: (a) percentage number of responses in eachscoring category (1, 2, 3, 4) for each of the five time periods tested; (b) detailed scores for each of the eight items tested in the last 12 monthperiod. (Scores: 1 = vague response; 2 = generic event; 3 = specific event; 4 = specific detailed event. The % numbers of the last 12 monthperiod, which concerns eight items instead of four, are divided by two.)

Time period

0

10

20

30

40

50

60

70

80

90

100

0–17 years old

18–30 years old

>30 years old

Last 5 years

Last 12 months

score 1 score 2 score 3 score 4

%num

ber

0

1

2

3

4

Sum

mer

Chri

stm

as

last

month

last

week

last

wee

kend

2day

sag

o

yest

erday

today

Time period

(a)

(b)

DISCUSSION

The aim of the present study was to investigate, in asingle case of SD (patient AT), the distinctionbetween episodic and semantic remote autobio-graphical memory and the integrity of strictly epi-sodic remote autobiographical memory andautonoetic consciousness.

Neuroimaging data (MRI) showed an asym-metric enlargement of the temporal horns of thelateral ventricles, which coincides with a majorright-sided temporal neocortical atrophy and

with a little less prominent tissue loss of the medialtemporal structures. Even if these results do notcorrespond to the typical profile of SD (Neary etal., 1998), neither the feature of relative medialtemporal atrophy nor the involvement of thefrontal lobe exclude the diagnosis of SD (Galton etal., 2001; Mummery, Patterson, Price, Ashburner,Frackowiak, & Hodges, 2000; Rosen et al.,2002). Thus, patient AT’s neuroimaging datasuggest a temporal variant of frontotemporaldementia (Edwards-Lee et al., 1997; Hodges et al.,1999).

PIOLINO ET AL.

632 COGNITIVE NEUROPSYCHOLOGY, 2003, 20 (7)

Time period

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0–17 years old

18–30 years old

> 30 years old

Last 5 years

Last 12 months

Controls

Patient AT

***

***

**

Figure 5. Ratio of global justified remember judgement to global remember judgement in each time period explored : Patient AT’s resultscompared to control subjects (mean and SD).

Pathological score: ** p < .01; *** p < .001.

Table 2. Patient AT’s autonoetic consciousness results on strictly episodic autobiographical memory test

0–17 18–30 >30 Last 5 Last 12years old years old years old years months

Global remember score (/12) 8 7 8 9 9Global justified remember score (/12) 0** 2* 1* 6 7What remember score (/4) 3 3 4 4 3Where remember score (/4) 3 3 3 3 3When remember score (/4) 2 1 1 2 3What justified remember score (/4) 0* 1 1 3 2Where justified remember score (/4) 0* 1 0* 2 2.5When justified remember score (/4) 0* 0* 0* 1 2.5

Pathological scores: *p < .05; **p < .01.

A general neuropsychological assessmentindicated a mild stage of SD, with difficulties ongeneral semantic knowledge, especially on famouspeople knowledge, and a preservation of phono-logic and syntactic aspects of language, visual-spa-tial, and perceptual abilities. In the context of adamage to the medial temporal lobe in patient AT,the relative preservation of anterograde episodicmemory may be surprising but not unprecedentedgiven the asymmetry of the damage (Smith &Bigel, 2001).

Using two autobiographical fluency tasks, wefound evidence of poor retrieval of names ofacquaintances albeit to a lesser degree with respectto recent life and good recall of autobiographicalevents from all time periods, which seeminglyfavours an episodic/semantic distinction in patientAT. However, our use of an original detailedinvestigation of strictly episodic remote memoryallowed us to demonstrate a good recall of all auto-biographical memories (whatever their nature)except for childhood and the adolescence period,but poor retrieval of remote, and not recent,specific detailed memories. Measures of patientAT’s recollective ability, via a remember/knowjudgement, revealed evidence of normal subjectiveremembering over the whole of his life, althoughhis remembering was not justified in terms of theactual contextual information retrieved beyondthe last 5 years. The apparent discrepancy betweenthe results of those two investigations (i.e., fluencyand questionnaire) merits comment and deliversinsights that question the issue of remote episodicmemory in SD.

The data of the autobiographical fluency taskscontrasts with a number of previous studies thathave investigated episodic and semantic autobio-graphical memory in SD using the Autobiographi-cal Memory Interview (Kopelman et al., 1989) andhave shown a low level of score elicited in patientswith SD in both personal semantic and autobio-graphical incidents for all time periods albeit to alesser degree with respect to recent life (Graham &Hodges, 1997; Nestor et al., 2002; Snowden et al.,1996). Nevertheless, our findings correspond to theresults of Westmacott et al. (2001), which havestressed in a single case (EL) a well-preserved

memory for episodes and a reverse temporal gradi-ent (to the detriment of remote past) for personallyrelevant names recalled from personal photographs.According to these authors, such a differentiationin remote memory between two kinds of informa-tion (specific events versus names of acquaintances)may support the distinction between episodic ver-sus semantic memory in SD. However, a few pointsare worthy of attention before assuming such a dis-tinction in our study.

First, can the deficits observed in the autobiographi-cal semantic fluency task be applied in general to AT’spersonal semantic knowledge? The names of people,whether famous or not, appeared particularlyaffected in patient AT, as demonstrated by theFrench celebrities test. Moreover, this category ofknowledge was more affected when compared toother ones tested with a fluency task (names of ani-mals, names beginning with the letter p). Never-theless, AT’s personal semantic deficit wasextended to the names of places where he went, hisdifferent addresses, or important dates along with arelative sparing of current information. Takentogether, these data are consistent with the viewthat several categories of personal semantic infor-mation are deficient in this patient.

Second, is the autobiographical episodic fluency tasksensible enough to investigate episodic memory?Although some authors (Dristchel et al., 1992;Greene et al., 1995) have used this task to assessepisodic autobiographical memory, the originalprocedure did not provide a precise measure of epi-sodic memory since all autobiographical eventsproduced were considered whatever their nature.Our modified version of this procedure proposes animprovement on previous instructions and scoringand takes into account only specific events,although no further details were requested, con-trarily to the autobiographical memory test. Never-theless, comparison of the good recall ofautobiographical events in the episodic fluencytask with the relative good recall of global auto-biographical memories in the autobiographicalmemory test, which takes into account all types ofrecalls, specific and generic, may suggest thatseveral events reported in the fluency task werein fact semanticised events (generic memories).

COGNITIVE NEUROPSYCHOLOGY, 2003, 20 (7) 633

AUTOBIOGRAPHICAL MEMORY IN SEMANTIC DEMENTIA

Nonetheless, it is noteworthy that the global scorecorresponds well to the “standard” episodic mem-ory score studied in controlled assessments such asthe Autobiographical Memory Interview(Kopelman et al., 1989). Thus, the episodic/semantic distinction in patient AT’s remote mem-ory depends on the definition of episodic autobio-graphical memory applied (“standard” or strict).

The fact that global autobiographical memory isrelatively preserved in patient AT is not unprece-dented in SD. However, although the global scoreswere not significantly altered compared to controlsubjects (except for the 0–17 years time period), it isnoteworthy that patient AT revealed a pattern ofincreasingly higher scores, reflecting recall of morespecific memories, from the remote to the recentpast consistent with previous studies (Graham &Hodges, 1997; Nestor et al., 2002; Snowden et al.,1996). A number of patients have been reported ashaving normal performances or only a mild impair-ment of remote autobiographical memory. In thecontext of selective right-sided temporal corticaldamage, patient VH, reported by Kitchener andHodges (1999), was rather similar to patient AT. Inaddition, patient DM (Graham et al., 1998) andpatient IH (Moss et al., 2000), who had predomi-nantly left-sided temporal damage, showed amilder remote memory impairment than other sin-gle-case studies (Graham & Hodges, 1997; Nestoret al., 2002). These findings suggest that patientswith SD may manifest mild remote autobiographi-cal impairment depending on the stage of their dis-ease and their degree of temporal lobe damage,most particularly whether this occurs unilaterally orbilaterally.

With respect to the strictly episodic autobio-graphical score, which uses a more stringent crite-rion consistent with the most up-to-date definitionof episodic memory (Tulving, 2001; Wheeler et al.,1997), our results highlight that, contrarily tothe global score, and the episodic fluency score, theproduction of specific detailed single events isrestricted to a short period of 5 years. This findingcorroborates the two single-case studies carried outby Graham and Hodges (1997; case AM) andNestor et al. (2002; case JH). We notably demon-strate a clear recency effect (for the last 12 months)

consistent with the temporal cut-off for a couple ofyears reported in patient AM, although some spe-cific nondetailed memories persist in patient ATthrough his young adulthood period. The study ofthe state of awareness during recollective experi-ence carried out with the Remember/Know proce-dure substantiates these findings. The evidencethat patient AT felt he could mentally travel back insubjective time to the same extent as controls cor-roborates the case report of Westmacott et al.(2001) and suggests that autonoetic consciousness,which is “the major distinguishing characteristic ofepisodic memory” (Conway, 2001; Wheeler et al.,1997), is preserved at the early stage of SD. Indeed,unlike amnesic patients whose access to the contentof memories does not give rise to recollective expe-rience (Knowlton, 1998; Knowlton & Squire,1995; Levine et al., 1998; Tulving, 1985), patientAT thinks he is actively remembering and that hecan focus his attention directly on his own subjec-tive experiences. However, the checking procedureused in our test, which allows one to assess if a sub-ject truly relived a specific event and its encodingcontext, showed that patient AT was unable tojustify each of his remember judgements beyond thelast 5 years period. Therefore, this feeling of relivingthe past seemed to be more “subjective” thanobjectively demonstrated by the content of patientAT’s memories.

Previously, opposite patterns of retrogradeamnesia between SD (reverse temporal Ribot’s gra-dient) and bi-hippocampal amnesic syndrome(Ribot’s gradient) have been interpreted as a strongsupport for the time-limited involvement of themedial temporal lobe in the storage and retrieval ofinformation (Graham, 1999; Graham & Hodges,1997; Hodges & Graham, 1998; Murre, Graham,& Hodges, 2001; Snowden et al., 1996). Indeed,“standard models” of consolidation (McClelland,McNaughton, & O’Reilly, 1995; Murre, 1996;Squire & Alvarez, 1995) based upon hippocampal-neocortical interaction propose that, over time(from 2 years to more than 10 years depending onthe authors), retrieval of information stored inneocortical areas becomes independent of themedial temporal lobe. The “standard model” ofconsolidation may provide an explanation for the

634 COGNITIVE NEUROPSYCHOLOGY, 2003, 20 (7)

PIOLINO ET AL.

temporal cut-off observed in patient AT for thesemantic fluency task and the strictly episodic scoreby suggesting that the retrieval of recent declarativeknowledge (episodic and semantic) is initiallydependent upon the medial temporal lobe (partiallypreserved in AT), whereas the retrieval of more dis-tant knowledge is subtended by neocortical net-works (affected in AT).

This observation is in keeping with the idea thatthe temporal neocortex is a crucial site for the stor-age of knowledge and past autobiographical experi-ence (Graham & Hodges, 1997; Hodges &Graham, 1998; Murre et al., 2001), which suggeststhat the alteration of remote memory in SD may beexplained by a deficit of storage (Nestor et al.,2002). Since the role of the anterior temporal lobe iswell known in semantic memory, this may implythat remote autobiographical memory has acquireda semantic nature (Cermak, 1984). Nevertheless,convergent data also provide arguments that theneocortical temporal lobe is crucial for the recollec-tion of autobiographical specific memories. Resultsobtained in patients with cerebral damage(Calabrese et al., 1996; Markowitsch, Calabrese,Haupts, Durwen, Liess, & Gehlen, 1993) as well asPET-based studies on normal subjects (Fink,Markowitsch, Reinkemeir, Bruckbauer, Kessler, &Heiss, 1996; Markowitsch et al., 1997), have con-firmed the major involvement of the right anteriortemporal lobe (which is damaged in patient AT) inecphorizing (i.e., associating and binding retrievalcues to the retrieval itself) old autobiographicalmemories. On the contrary, other functionalneuroimaging studies have stressed the involve-ment of the left side fronto-temporal regions (rela-tively more preserved in patient AT) in ecphorizingautobiographical memory (Conway et al., 1999;Maguire, Henson, Mummery, & Frith, 2001;Maguire & Mummery, 1999; Mummery et al.,2000). Therefore, the pattern of remote strictly epi-sodic memory and justified remembering in patientAT may be interpreted by a deficit of retrievalinstead of storage. According to a recent construc-tive model of autobiographical memory (Conway,2001; Conway & Pleydell-Pearce, 2000), a com-plex cyclic retrieval process that depends on thefrontal lobe (especially the left side) allows one to

access sensory/perceptual event-specific knowledge(situated on the posterior regions) through the per-sonal knowledge base (lifetime periods and generalevents situated on the fronto-temporal regions,especially in the right side). Hence, in keeping withthe interpretation of Kitchener and Hodges (1999)for the well-preserved autobiographical memory inpatient VH, our findings may suggest that theremaining healthy tissue in patient AT, especiallyin the left hemisphere, is sufficient to ensure therecollection of recent and some remote autobio-graphical memories (often generic and a few spe-cific ones without details) and subjective sense ofremembering, but prevents access to remote epi-sodic details because personal semantic memory isessential to recollect remote episodic memory.

It is noteworthy that an alternative model ofmemory consolidation has been proposed andtermed the “multiple memory trace model”(Moscovitch & Nadel, 1999; Nadel &Moscovitch, 1997), in which it is suggested thatepisodic recollection always depends on themedial temporal lobe. This is because this struc-ture is essential for the long-term availability ofparticular aspects of memory whereas personalsemantic information becomes independent of itover time. According to this latter theory, remoteepisodic memory would be relatively preservedwhatever the time interval in SD, if tested appro-priately (not too dependent on verbal abilities), asargued by Westmacott et al. (2001) from their sin-gle case (patient EL). However, this theory mayalso explain the report of a reversed temporal gra-dient in SD by suggesting an additional deficit ofaccess. The differentiation of impairment betweennames and specific events (demonstrated with thetwo fluency tasks) and the relative well-preservedglobal autobiographical memory and subjectivesense of remembering in recent, as well as remote,autobiographical memory (demonstrated with theautobiographical memory test) would concur withthe “multiple trace model.” However, one must becautious about the interpretation of such data.First, it depends on the definition of episodicautobiographical memory applied (“standard” orstrict). Second, it depends on the assumptionthat the episodic nature of the information relies

COGNITIVE NEUROPSYCHOLOGY, 2003, 20 (7) 635

AUTOBIOGRAPHICAL MEMORY IN SEMANTIC DEMENTIA

essentially upon the function of the medial tempo-ral lobe while other regions, particularly theprefrontal lobe, are involved in episodic memoryand autonoetic consciousness (Levine et al., 1998;Wheeler et al., 1997). These two structures(medial temporal and prefrontal lobes), which arenot spared in patient AT, are both involved in theretrieval of memories. Therefore, with our single-case study and more generally in SD, it is difficultto especially infer, as suggested by several authors(Moscovitch & Nadel, 1999; Nadel et al., 2000;Westmacott et al., 2001), that persistent remoteautobiographical memory, or sense of remember-ing, would argue in favour of the permanent role ofthe medial temporal lobe in retrieval.

In conclusion, we report, in a patient with SD, apoor retrieval of remote personally relevant namesand specific detailed memories contrasting with arelative good recall of global autobiographicalmemory and subjective remembering over thewhole of his life. Our observations support the viewthat remote autobiographical memory may be rela-tively preserved in SD and subtended by a distrib-uted network that may assist the cerebral structuresdamaged in SD. Nonetheless, strictly episodicremote memory was clearly impaired in our patient.

Manuscript received 11 December 2001Revised manuscript received 11 October 2002

Revised manuscript accepted 13 November 2002

REFERENCES

Baddeley, A. D., & Wilson, B. (1986). Amnesia, auto-biographical memory and confabulation. In D. C.Rubin (Ed.), Autobiographical memory (pp. 225–252).Cambridge, UK: Cambridge University Press.

Belliard, S., Perron, M., Rouyer, F., Golfier, V., Sartori,E., & Edan, G. (2001). Présentation d’une batteriefrançaise de personnes célèbres. Résultats auprès depatients atteints de démence sémantique. RevueNeurologique (Paris), 157, 10–14.

Benton, A. L. (1982). Test de rétention visuelle de Benton.Paris: Les Editions du Centre de PsychologieAppliquée.

Borrini, G., Dall’Ora, P., Della Sala, S., Marinelli, L., &Spinnler, H. (1989). Autobiographical memory,

sensitivity to age and education of a standardisedinquiry. Psychological Medicine, 19, 215–224.

Butters, N., & Cermak, L. S. (1986). A case of study ofthe forgetting of autobiographical knowledge. In D.C. Rubin (Ed.), Autobiographical memory (pp. 253–272). Cambridge, UK: Cambridge University Press.

Calabrese, P., Markowitsch, H. J., Durwen, H. F.,Widlitzek, H., Haupts, M., Holinka, B., & Gehlen,W. (1996). Right temporofrontal cortex as criticallocus for the ecphory of old episodic memories. Jour-nal of Neurology, Neurosurgery, and Psychiatry, 61,304–310.

Cardebat, D., Doyon, B., Puel, M., Goulet, P., &Joanette, Y. (1990). Evaluation lexicale formelle etsémantique chez des sujets normaux. Performances etdynamique de production en fonction du sexe, de l’âgeet du niveau. Acta Neurologica Belgica, 90, 207–217.

Cermak, L. S. (1984). The episodic-semantic distinctionin amnesia. In L. R. Squire & N. Butters (Eds.),Neuropsychology of memory (pp. 55–62). New York:Guilford Press.

Conway, M. A. (2001). Sensory-perceptual episodicmemory and its context: Autobiographical memory.Philosophical Transactions of the Royal Society ofLondon. Series B: Biological Sciences, 356, 1375–1384.

Conway, M. A., & Pleydell-Pearce, C. W. (2000). Theconstruction of autobiographical memories in theself-memory system. Psychological Review, 107, 261–288.

Conway, M. A., Turk, J. D., Miller, S. L., Logan, J.,Nebes, R. D., Meltzer, C. C., & Becker J. T. (1999).A positron emission tomography (PET) study ofautobiographical memory retrieval. Memory, 7, 679–702.

Crovitz, H. F., & Schiffman, H. (1974). Frequency ofepisodic memories as a function of their age. Bulletinof the Psychonomic Society, 4, 517–518.

Deloche, G., & Hannequin, D. (1997). Test dedénomination orale d’images DO 80. Paris: LesEditions du Centre de Psychologie Appliquée.

De Renzi, E. D., & Faglioni, P. (1975). Verbal compre-hension in aphasic and in normal subjects with ashortened version of the token test. Rivista diPatologia Neurosa e Mentale, 96, 252–269.

Desgranges, B., Eustache, F., Rioux, P., De la Sayette,V., & Lechevalier, B. (1996). Memory disorders inAlzheimer’s Disease and the Organization of HumanMemory. Cortex, 32, 387–412.

Dritschel, B., Williams, J. M. W., Baddeley, A. D., &Nimmo-Smith, J. (1992). Autobiographical fluency.

636 COGNITIVE NEUROPSYCHOLOGY, 2003, 20 (7)

PIOLINO ET AL.

A method for the study of personal memory. Memoryand Cognition, 20, 135–140.

Edwards-Lee, T., Miller, B., Benson, D., Cumming, J.,Russel, G., Boone, K., & Mena, I. (1997). Thetemporal variant of frontotemporal dementia. Brain,120, 1027–1040.

Eustache, F., Desgranges, B., De la Sayette, V., Lalevée,C., Giffard, B., Piolino P., Viader, F., & Baron, J. C.(2001). Contributions de la tomographie parémission de positons à la neuro-anatomiefonctionnelle de la maladie d’Alzheimer. RevueNeurologique, 157, 377–383.

Fink, G. R., Markowitsch, H. J., Reinkemeier, M.,Bruckbauer, T., Kessler, J., & Heiss, W. D. (1996).Cerebral representation of one’s own past: Neuralnetworks involved in autobiographical memory.Journal of Neuroscience, 16, 4275–4282.

Galton, C. J., Patterson, K, Graham, K., Lambon Ralph,M. A., Williams, G., Antoun, N., Sahakian, B. J., &Hodges, J. R. (2001). Differing patterns of temporalatrophy in Alzheimer’s disease and semantic demen-tia. Neurology, 57, 216–225.

Gardiner, J. M. (1988). Functional aspects of recollectiveexperience. Memory and Cognition, 16, 309–313.

Gardiner, J. M. (2001). Episodic memory and autonoeticconsciousness: A first-person approach. PhilosophicalTransactions of the Royal Society of London. Series B:Biological Sciences, 356, 1351–1362.

Giffard, B., Desgranges, B., Nore-Mary, F., Lalevée, C.,de la Sayette, V., Pasquier, F., & Eustache, F. (2001).The nature of semantic memory deficits in Alzhei-mer’s disease: New insights from hyperprimingeffects. Brain, 124, 1522–1532.

Graham, K. S. (1999). Semantic dementia: A challengeto the multiple-trace theory. Trends in CognitiveScience, 3, 85–87.

Graham, K. S., & Hodges, J. R. (1997). Differentiatingthe roles of the hippocampal complex and the neocor-tex in long-term memory storage: Evidence from thestudy of semantic dementia and Alzheimer’s disease.Neuropsychology, 11, 77–89.

Graham, K. S., Lambon Ralph, M. A., & Hodges, J. R.(1997). Determining the impact of autobiographicalexperience on “meaning”: New insights from investi-gating sports-related vocabulary and knowledge intwo cases with semantic dementia. CognitiveNeuropsychology, 3, 534–544.

Graham, K. S., Pratt, K. H., & Hodges, J. R. (1998). Areverse temporal gradient for public events in a singlecase of semantic dementia. Neurocase, 4, 461–470.

Greene, J. D., Hodges, J. R., & Baddeley, A. D. (1995).Autobiographical memory and executive function inearly dementia of Alzheimer type. Neuropsychologia,12, 1647–1670.

Grober, E., & Buschke, H. (1987). Genuine memorydeficits in dementia. Developmental Neuropsychology,3, 13–36.

Hodges, J. R., Patterson, K., Oxbury, S., & Funnell, E.(1992). Semantic dementia. Progressive fluent apha-sia with temporal lobe atrophy. Brain, 115, 1783–1806.

Hodges, J. R., & Graham, K. S. (1998). A reversal of thetemporal gradient for famous person knowledge insemantic dementia: Implications for the neuralorganisation of long-term memory. Neuropsychologia,36, 803–825.

Hodges, J. R., Patterson, K., Ward, R., Garrard, P., Bak,T., Perry, R., & Gregory, E. (1999). The differentia-tion of semantic dementia and frontal lobe dementia(temporal and frontal variants of fronto-temporaldementia) from early Alzheimer’s disease: A compar-ative neuropsychological study. Neuropsychology, 13,31–40.

Hodges, J. R., & Graham, K. S. (2001). Episodicmemory: Insights from semantic dementia. Philo-sophical Transactions of the Royal Society of London.Series B: Biological Sciences, 356, 1423–1434.

Kitchener, E. G., & Hodges, J. R. (1999). Impairedknowledge of famous people and events and intactautobiographical knowledge in a case of progressiveright temporal lobe degeneration: Implications forthe organization of remote memory. CognitiveNeuropsychology, 16, 589–607.

Knowlton, B. J. (1998). The relationship betweenremembering and knowing: A cognitive neuroscienceperspective. Acta of Psychology (Amsterdam), 98, 253–265.

Knowlton, B. J., & Squire, L. R. (1995). Rememberingand knowing: Two different expressions of declara-tive memory. Journal of Experimental Psychology:Learning, Memory, and Cognition, 21, 699–710.

Kopelman, M. D., Wilson, B. A., & Baddeley, A. D.(1989). The autobiographical memory interview: Anew assessment of autobiographical and personalsemantic memory in amnesic patients. Journal of Clin-ical and Experimental Neuropsychology, 11, 724–744.

Kroll, N. E. A., Markowitsch, H. J., Knight, R. T., &Von Cramon, Y. (1997). Retrieval of old memories:The temporofrontal hypothesis. Brain, 120, 1377–1399.

COGNITIVE NEUROPSYCHOLOGY, 2003, 20 (7) 637

AUTOBIOGRAPHICAL MEMORY IN SEMANTIC DEMENTIA

Levine, B., Black, S. E., Cabeza, R., Sinden, M.,McIntosh, A. R., Toth, J. P., Tulving, E., & Stuss,D. T. (1998). Episodic memory and the self in a caseof isolated retrograde amnesia. Brain, 121, 1951–1973.

Maguire, E. A., Henson, R. N. A., Mummery, C. J., &Frith, C. D. (2001). Activity in prefrontal cortex, nothippocampus, varies parametrically with the increas-ing remoteness of memories. NeuroReport, 12, 441–444.

Maguire, E. A., & Mummery, C. (1999). Differentialmodulation of a common memory retrieval networkrevealed by positron emission topography. Hippocam-pus, 9, 54–61.

Mäntylä, T. (1993). Knowing but not remembering:Adult age differences in recollective experience.Memory and Aging, 21, 379–388.

Markowitsch, H. J. (1995). Which brain regions are crit-ically involved in the retrieval of old episodic memory.Brain Research Reviews, 21, 117–127.

Markowitsch, H. J., Calabrese, P., Haupts, M., Durwen,H. F., Liess, J., & Gehlen, W. (1993). Searching forthe anatomical basis of retrograde amnesia. Journalof Clinical and Experimental Neuropsychology, 15,947–967.

Markowitsch, H. J., Thiel, A., Kessler, J., vonStockhausen, H. M., & Heiss, W. D. (1997).Ecphorizing semi-conscious episodic episodic infor-mation via the right temporopolar cortex—a Petstudy. Neurocase, 3, 445–449.

McClelland, J. L., McNaughton, B. L., & O’Reilly, R.C. (1995). Why there are complementary learningsystems in the hippocampus and neocortex. Psycholog-ical Review, 102, 419–457.

Moscovitch, M., & Nadel, L. (1999). Consolidation andthe hippocampal complex revisited: In defense ofthe multiple-trace model. Current Opinion inNeurobiology, 8, 297–300.

Moscovitch, M., & Nadel, L. (2000). Multiple-tracetheory and semantic dementia: Response toK. S. Graham (1999). Trends in Cognitive Sciences, 3,87–89.

Moscovitch, M., Yaschyshyn, T., Ziegler, M., & Nadel,L. (1999). Remote episodic memory and retrogradeamnesia: Was Endel Tulving right all along? In E.Tulving (Ed.), Memory, consciousness, and the brain:The Tallinn conference (pp.331–345). Philadelphia,PA: Psychology Press.

Moss, E., Cappelletti, M., De Mornay Davies, P.,Jaldow, E., & Kopelman, M. (2000). Lost for wordsor loss of memories: Autobiographical memory in

semantic dementia. Brain and Language, 74, 350–354.

Mummery, C. J., Patterson, K., Price, C. J., Ashburner,J., Frackowiak, R. S. J., & Hodges, J. R. (2000). Avoxel based morphometry study of semantic demen-tia: Relationship between temporal lobe atrophy andsemantic dementia. Annals of Neurology, 47, 36–45.

Murre, J. M. J. (1996). TraceLink: A model of amnesiaand consolidation of memory. Hippocampus, 6, 675–684.

Murre, J. M. J., Graham, K. S., & Hodges, J. R. (2001).Semantic dementia: Relevance to connectionnistmodels of long-term memory. Brain, 124, 647–75.

Nadel, L., & Moscovitch, M. (1997). Memory consoli-dation, retrograde amnesia and the hippocampalcomplex. Current Opinion in Neurobiology, 7, 217–227.

Nadel, L., Samsonovich, A., Ryan, L., & Moscovitch,M. (2000). Multiple trace theory of human memory:Computational, neuroimaging, and neuro-psychological results. Hippocampus, 10, 352–368.

Neary, D., Snowden, J. S., Gustafson, L., Passant, U.,Stuss, D., Black, S., Freedman, M., Kertesz, A.,Robert, P. H., Albert, M., Boone, K., Miller, B.,Cummings, J., & Benson, D. F. (1998).Frontotemporal lobar degeneration: A consensus onclinical diagnostic criteria. Neurology, 51, 1546–1554.

Nelson, H. A. (1976). A modified card sorting test sensi-tive to frontal defects. Cortex, 12, 313–324.

Nestor, P. J., Graham, K. S., Bozeat, S., Simons, J. S., &Hodges, J. R. (2002). Memory consolidation andthe hippocampus: Further evidence from studies ofautobiographical memory in semantic dementia andfrontal variant frontotemporal dementia. Neuro-psychologia, 40, 633–654.

Nolde, S. F., Johnson, M. K., & D’Esposito, M. (1998).Left prefrontal activation during episodic remember-ing: An event-related fMRI study. NeuroReport, 9,3509–3514.

Piolino, P., Desgranges, B., Benali, K., & Eustache, F.(2002). Episodic and semantic remote autobiograph-ical memory in ageing. Memory, 10, 239–257.

Piolino, P., Desgranges, B., & Eustache, F. (2000). Lamémoire autobiographique: Théorie et pratique.Marseille, France: Solal.

Rabbitt, P., & Winthorpe, C. (1988). What do oldpeople remember? The Galton paradigm reconsid-ered. In M. M. Gruneberg, P. E. Morris, & R. N.Sykes (Eds.), Practical aspects of memory: Vol. 2. Clini-cal and educational implications (pp. 301–307).London: Wiley.

PIOLINO ET AL.

638 COGNITIVE NEUROPSYCHOLOGY, 2003, 20 (7)

Raven, J. C., Court, J. H., & Raven, J. (1998). Progres-sive matrices. Paris: Editions et Applicationspsychotechniques.

Rey, A. (1960). Test de la figure complexe de Rey. Paris: LesEditions du Centre de Psychologie Appliquée.

Rosen, H. J., Gorno-Tempini, M. L., Goldman, W. P.,Perry R. J., Schuff, N., Weiner, M., Feiwell, R.,Kramer, J. H., & Miller, B. L. (2002). Patterns ofbrain atrophy in frontotemporal dementia andsemantic dementia. Neurology, 58, 198–208.

Schmidtke, K., & Vollmer, H. (1997). Retrograde amne-sia and semantic memory deficits. Neuropsychologia,35, 505–518.

Smith, M. L., & Bigel, M. G. (2001). The temporal lobesand memory. Philosophical Transactions of the RoyalSociety of London. Series B: Biological Sciences, 356,49–65.

Snowden, J., Goulding, J., & Neary, D. (1989). Semanticdementia: A form of circumscribed cerebral atrophy.Behavioural Neurology, 2, 167–182.

Snowden, J., Griffiths, H., & Neary, D. (1994). Seman-tic dementia: Autobiographical contribution to pres-ervation of meaning. Cognitive Neurospychology, 11,265–288.

Snowden, J., Griffiths, H., & Neary, D. (1995). Auto-biographical experience and word meaning. Memory,3, 225–246.

Snowden, J., Griffiths, H., & Neary, D. (1996). Seman-tic-episodic memory interactions in semantic demen-tia: Implications for retrograde memory function.Cognitive Neuropsychology, 13, 1101–1137.

Snowden, J., Griffiths, H., & Neary, D. (1999). Theimpact of autobiographical experience on meaning:Reply to Graham, Lambon Ralph, and Hodges.Cognitive Neuropsychology, 16, 673–687.

Squire, L. R., & Alvarez, P. (1995). Retrograde amnesiaand memory consolidation: A neurobiologicalperspective. Current Opinion in Neurobiology, 5, 169–177.

Tulving, E. (1985). How many memory systems arethere? American Psychologist, 40, 385–398.

Tulving, E. (2001). Episodic memory and commonsense: How far apart? Philosophical Transactions of theRoyal Society of London. Series B: Biological Sciences,356, 1505–1515.

Tulving, E., Kapur, S., Craik, F. I. M., Moscovitch, M.,& Houle, S. (1994). Hemispheric encoding/retrievalasymmetry in episodic memory: Positron emissiontomography findings. Proceedings of the NationalAcademy of Science of the USA, 91, 2016–2020.

Wechsler, D. (1991). Echelle clinique de mémoire de Wechs-ler révisée. Paris: Editions du Centre de PsychologieAppliquée.

Westmacott, R., Leach, L., Freedman, M., &Moscovitch, M. (2001). Different patterns of auto-biographical memory loss in semantic dementia andmedial temporal lobe amnesia: A challenge to consol-idation theory. Neurocase, 7, 37–55.

Wheeler M. A., Stuss, D. T., & Tulving E. (1997).Toward a theory of episodic memory: The frontallobes and autonoetic consciousness. PsychologicalBulletin, 121, 331–354.

COGNITIVE NEUROPSYCHOLOGY, 2003, 20 (7) 639

AUTOBIOGRAPHICAL MEMORY IN SEMANTIC DEMENTIA