Neurocase

ISSN: 1355-4794 (Print) 1465-3656 (Online) Journal homepage: http://www.tandfonline.com/loi/nncs20

Confabulation: Remembering another past,

planning another future

Gianfranco Dalla Barba , Jee Yun Cappelletti , Matteo Signorini & Gianfranco
Denes

To cite this article: Gianfranco Dalla Barba , Jee Yun Cappelletti , Matteo Signorini & Gianfranco
Denes (1997) Confabulation: Remembering another past, planning another future, Neurocase,
3:6, 425-436, DOI: 10.1080/13554799708405018

To link to this article: http://dx.doi.org/10.1080/13554799708405018

Published online: 17 Jan 2008.

Submit your article to this journal

Article views: 96

View related articles

Citing articles: 2 View citing articles

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=nncs20

Download by: [Universidad Nacional Colombia] Date: 31 May 2017, At: 09:55
Neuror.u.ve (1997) Vol. 3, pp, 4 2 5 4 3 6 8C:i O.\-Sortl Utiiversitjb Press 1997

Confabulation: Remembering Another Past, Planning

Another Future

Gianfranco Dalla Barball, Jee Yun Cappelletti3, Matteo Signorini3 and Gianfranco Denes3
U. 324 INSERM, Paris, 2Department of Neurology, H6pital Henri Mondor, Creteil, France and 3Department of Neurology,
University of Padua, Italy

Abstract

In this study we describe a patient, GA, who developed an amnesic-confabulatory syndrome, following a
subarachnoid haemorrhage and ischaemia due to rupture of the anterior communicating artery and
subsequent vasospasm. As far as the performance on memory tasks was concerned, GAs confabulation
was found to be restricted to the autobiographical aspect of episodic memory. Confabulation did not
manifest itself in episodic learning tasks nor on tasks tapping various kinds of semantic knowledge. In
contrast, GA confabulated in orientation in time and place tasks and also in tasks where she was required
to plan her personal future. GAs confabulation could not be accounted for in terms of an impairment of
strategic retrieval or of reality monitoring processes. It is suggested that GAS confabulation reflects a
pathological awareness of personal temporality.

Introduction
Confabulation is a symptom, observable in amnesic
patients who are unaware of their memory deficit. It
consists of actions and verbal statements that are un-
intentionally incongruous to the patients history,
background and present situation (Dalla Barba, 1993a).
This rather infrequent disorder is classically described
in Korsakoffs syndrome (BonhoefTer, 1904; Wyke and
Warrington, 1960; Talland, 1961; Mercer et al., 1977;
Cermak rt a/., 1980; Dalla Barba et al., 1990; Benson rt al.,
1996; Schnider et al., 1996), but also in patients suffering
from ruptured aneurysms of the anterior communicating
artery, subarachnoid haemorrhage or encephalitis (Luria,
1976; Stuss er al., 1978; Kapur and Coughlan, 1980;
Alexander and Freedman, 1984; Vilkki, 1985; Moscovitch,
1989; Delbecq-Derouesne et ul., 1990; Irle et ul., 1992;
Kopelman et i d , 1995; Moscovitch, 1995; Papagno and
Muggia, 1996; Schnider et al., 1996; Dalla Barba et al.,
1997), head injury (Weinstein and Lyerly, 1968; Baddeley
and Wilson. 1986; Dalla Barba, 1993b; Schnider et id.,
1996), Binswangers Encephalopathy (Dalla Barba, 1993a),
Alzheimers disease (Kern et a/., 1992) and aphasia
(Sandson rt al., 1986).
Confabulation has traditionally been considered a
disorder of memory and, in particular, a disorder of the
retrieval or post-retrieval stage from long-term memory
(e.g. Baddeley and Wilson, 1986; Moscovitch, 1989, 1995;
Johnson, 1991; Burgess and Shallice, 1996). There is
evidence, however, that considering confabulation as a
disorder confined to long-term memory is probably restric-
tive. In fact, confabulators usually make confabulatory
errors in time and space orientation tasks (e.g. Dalla Barba
et uI., 1990; Dalla Barba, 1993a,b). There is also some
clinical evidence that confabulation can also manifest itself
when the patient is required to plan his personal future.
For example, Dalla Barba (1993a) reported that on one
occasion the patient MB said that he was looking forward
to the end of the testing session because he had to go to the
general store to buy some new clothes . . . On this occasion
the patient actually attempted to leave his hospital room,
claiming that there was a taxi waiting for him downstairs
(p. 4). Dalla Barbas tentative conclusion is that the sub-
jectively experienced temporality, in its continuity and
unity, is a crucial factor in MBs confabulation.
This paper describes the study of a patient, GA, who
developed an amnesic-confabulatory syndrome, following
a subarachnoid haemorrhage and ischaemia due to rupture
of the anterior communicating artery and subsequent
vasospasm. GAs confabulation involved the auto-
biographical aspect of episodic memory, as well as her
performance in orientation in time and place tasks and in
tasks where she was required to plan her personal future.
This is the first case in which confabulation extending to

Correspondence to: Gianfranco Dalla Barba, U. 324 INSERM, Centre Paul Broca, 2ter rue dAlesia, 75014 Paris, France. Tel: + 33 I 40 78 92 17;
Fax: + 33 I 45 89 68 48; e-mail: dallabarba@broca.inserm.fr
426 G. Dalla Barba tt (11.
the entire personal temporality, i.e. past, present and
ttiturc, is experimentally documented.
The study of GAs case also provides interesting infor-
mation concerning the role of search and monitoring
processes in confabulation. Most of the current ideas
about mechanisms underlying confabulation emphasize the
role of ;I frontal executive dysfunction in this disorder
(Baddeley and Wilson, 1986; Kopelman, 1987; Moscovitch,
1989; Delbecq-Derouesne et ul., 1990; Johnson, 1991;
Moscovitch, 1995; Burgess and Shallice, 1996: Moscovitch
and Melo, 1997). These ideas are based on the common
assumption that frontal lobe neural structures are the
iinatomical correlate of cognitive functions that are
involved in organizing and controlling the cognitive
material processed by post-rolandic brain structures. As
fitr as memory is concerned, current theories assume that
the frontal lobe represents the neural substrate involved in
organizing a search in long-term memory and in evaluating
the resull of this search, which is the retrieved memory.
According to this view, confabulation would be the result
ot an impairment of these search and monitoring processes
based on frontal neural structures.
Moscovitch (1989, 1995; Moscovitch and Melo, 1997),
for example, distinguishes between two components of
retrieval. One, associative retrieval, is relatively automatic
and independent of frontal functions. The other, strategic
retrieval. is self-initiated, goal-directed, effortful and intel-
ligent. Within strategic retrieval processes. two further
components are hypothesized. The first involves organizing
;I mcniory search that uses whatever knowledge is avail-
able, whether semantic or episodic. Once knowledge is
recovered. a second strategic process is involved in moni-
toring the output of the memory search and checking
whether it is consistent with other information in semantic
and episodic memory. According to Moscovitch (1989,
1995; Moscovitch and Melo, 1997) the disruption of the
monitoring sub-component of strategic retrieval would be
responsible for confabulation.
Following a similar perspective Johnson ( 1 991) argues
that confabulation is a disorder resulting from a failure of
rcality monitoring (i.e. deciding whether a memory is a
trace of something that actually happened to you, or is a
memory of an imagined event), and suggests that reality
monitoring is a function of two kinds of judgement pro-
cesses. The first is based on the evaluation of the qualitative
characteristics of the retrieved information (e.g. the type
and amount of perceptual details); the second is based on a
more deliberate evaluation of retrieved information in
the light of other supporting memories and knowledge.
Damage to frontal structures would result in an impair-
ment of judgement processes involved in reality monitoring
and so in confabulation.
In a study based on the analysis of healthy volunteers
autobiographical recollections of recent everyday events,
Burgess and Shallice ( 1996) argue that errors in recall made
by normal subjects and those made by confabulators can
be accounted for as the result of damage t~a different
components of control processes. These are considered
part of the Supervisory System (Norman and Shallice.
1980/1986) which is held to be a function of the frontal
cortex.
These hypotheses about the origin of confabulation
are challenged by several types of observations: ( a ) two
patients have been described with a confabulatory syn-
drome where there were spared executive functions and no
frontal pathology (Dalla Barba ct ul., 1990; Dalla Barba.
I993a); (b) the confabulating patient described by Delbccq-
Derouesne and co-workers (1990) had a documented fron-
tal lobe lesion but performed normally on tasks supposed
to be sensitive to frontal lesion; (c) the confabulating
patient recently described by Dalla Barba and co-workers
( 1997) showed impaired executive functions without any
evidence of structural or functional damage to the frontal
lobe; (d) in some patients confabulation atTects the
performance on episodic memory tasks but not on
semantic memory tasks. Taken together, these findings
suggest that a retrieval control hypothesis is insufficient i n
accounting for confabulation. They also cast doubts on the
hypothetical prominent role of a frontal dysfunction i n
confabulation, and on the anatomical basis of executive
tasks.
Case report
History
GA is a 52-year-old right-handed woman with 5 years 01
education. She is married and has two children. She has
always worked as a housewife. She has no history of
psychiatric or neurologic diseases.
On 29 June 1995 she was admitted to the Neurosurgery
Department of the Treviso General Hospital suffering from
a severe headache, with no neurological signs. A CT brain
scan showed a subarachnoid haeniorrhage. An arterio-
graphy showed an aneurysm of the anterior csmmunicat-
ing artery. On 3 July she was operated on in order to clip
the aneurysm. A few hours after the operation the
patient fell into a coma. An Echo-Doppler study showed it
severe bilateral vasospasm of the brain vessels. A CT scan
performed on that occasion showed an ischaemic lesion
involving the frontal lobe bilaterally. An ( IxF)fluorodeoxy-
glucose PET study in the resting state performed I week
later showed an extensive cortico-subcortical hypo-
metabolism involving the anterior commissure, the niediul
frontal cortex, the inferior part of the left frontal lobe and
the periventricular white matter. An MRI performed on 3 I
July showed bilateral frontal degenerative areas extending
from the orbito-frontal regions to the dorsal areas, involv-
ing the anterior cingulum and the anterior two-thirds of
the corpus callosum. The patient remained in a coni;i
for approximately 3 months. Nine months after the
vascular accident the patient was referred t o the Neurology
 Confabulation and temporality 427

Department of the University of Padova for neuropsycho- Table 1. GAs scores 011 the subtests of the Wechsler Memory Scale
logical evaluation and was seen on various occasions
between March and November 1996. The neurological Mean for
GA her age SD
examination was normal. ~

Personal information 4 5.91 0.3 I

Neuropsychological examination
The patients collaboration during the testing sessions
was good. On direct questioning she was shown to be
disoriented and confabulated in time and space. She
performed well on certain tasks which are in general
considered to be sensitive to intellectual deterioration, such
as the Ravens Coloured Progressive Matrices-PM 47
where she achieved a score of 25/36 (Equivalent Score = 3),
the Mini Mental State Examinations (Folstein r t d.. 1975)
(25/30). and proverbs interpretation. On bedside tests of
oral expression, understanding of oral language, reading,
writing. neglect, praxis, calculation and short-term memory
the patient was judged to be normal.
Tests considered sensitive to frontal lesions
As G A had a lesion extensively involving the frontal lobe
structures. her performance was examined on tasks that are
traditionally considered sensitive to frontal lobe lesions.
Her performance on these tasks varied greatly. On the
Modified Card Sorting Test (Nelson, 1976) she performed
normally: she achieved five categories and produced only
four perseverative errors (score = 36.1, cut-off for frontal
lobe lesion = 50% of perseverative errors). Her perform-
ance was also normal on the Cognitive Estimates (Shallice
and Evans, 1978) and on the Tower of London (Shallice,
1982; Alamanno et ( I / . , 1987) where she achieved the
maximum score. On other tests, however, her performance
was markedly impaired. In Lurias sequencing tests (Luria.
1966) she was unable to learn a sequence of gestures of
length 2. On tasks of Verbal Fluency she was also very
defective. On Category Fluency, a test that requires gener-
ating as many words as possible in 1 minute from a given
semantic category (e.g. animals), she produced an average
of four words, and on Letter Fluency, involving generating
as many words as possible in 1 minute starting with the
letters F. A and S, she produced an average of three words.
The Trial Making Test (Davies, 1968) was impossible.
To summarize, GAs performance on tests traditionally
considered sensitive to frontal lesions varied greatly: in
seven tests of this type, her performance was completely
normal on three and markedly impaired on the others.
Clinical tests of memory
On the Wechsler Memory Scale (see Table I ) GAs MQ
was 70, which is a clearly pathological score. Individual
subtest scores ranged from very defective on Visual Repro-
duction and Logical Memory to average on Digits and
Mental Control. On the Warrington Recognition Memory
Orient at io t i 4 4.96 0.24
Mental Control 5 6.63 3.04
Logical memory 4 10.15 3.52
Digits 8 10.36 2. I 4
Visual reproduction 1 10.67 3.20
Paired absociate learning 6.5 16.25 3.91
MQ 70
Delayed logical memory 0
Test (Warrington, 1984) she performed at the chance level
both on the words (25/50) and on the faces (26/50) subtests.
In another task she was presented with 50 photographs of
unknown faces and then asked for a Yes-No recognition
test in which the 50 target faces together with an equal
number of distracters were presented and she had to select
the items that she had seen before. Her performance on this
test was also at the chance level since she correctly recog-
nized 30/50 target items and made 20/50 false recognitions.
A questionnaire concerning public events relevant to the
Italian population for which normative data are available
(Costa rf NI., 1989) was given to GA. The questionnaire is
given in a recognition form and includes six questions
per biennium, from 1965 to 1985, with four choices per
question. On this test she was correct only five times out of
60, she made six false recognitions and refused to choose
any of the four alternatives 49 times. She was also very
defective in learning tasks. She was unable to learn a
supraspan spatial sequence of length corresponding to her
spatial span + 2 (De Renzi et a/., 1977a) and to learn a
10-word list after 20 presentations (De Renzi et ul., 1977b).
Semantic memory
GAs performance in tasks involving various types of
semantic knowledge was normal: definition of words,
pointing out the correctly coloured object, performing a
symbolic gesture on spoken request, the Boston Naming
Test, detecting semantic anomalies (Baddeley, 1982), were
all carried out correctly.
Examination of confabulation
G A produced most of her confabulations in informal
conversation, when answering specific questions and
spontaneously, mainly referring to autobiographical epi-
sodic memory. Confabulation was always plausible and
semantically appropriate so that a hypothetical observer
not familiar with GAs history, background and present
situation could hardly tell whether GAs reports were
confabulatory or not. For example, she would claim that
the day before, she went out shopping and then prepared a
428 G. Dalla Barba ct trl.
lable 2. GAs and normal controls percentage of correct responses and confabulations on the Confabulation Battery
Correct
Pcrsonal semantic rneiiiory 8.5
I:pisodic incmory 46
01-ientation time place 60
(icncriil scm;intic memory 73
Linguistic semantic memory 53
I dont know episodic 0 0
I dont know semantic 0 0
menl for her family. In this case it is impossible to tell
whether this is a confabulatory report unless one knows
that since her illness GA had never gone shopping nor had
she ever prepared any meals.
Confabulation Battery
To further examine GAs confabulation, we submitted her
t o the Confabulation Battery (Dalla Barba, 1993a), includ-
ing questions probing personal, general and linguistic
semantic memory. episodic memory, orientation for space
and time and questions to which the appropriate response
would be 1 dont know, both semantic (e.g. What did
Marilyn Monroes father do?) and episodic (e.g. Do you
remember what you did on 13 March 1985?). Table 2
shows GAs and 12 normal control (NC)subjects perform-
ance on the Confabulation Battery. As it clearly emerges
from Table 2, GAs confabulations were found to be
restricted mainly to episodic memory and orientation in
time and place. GAs confabulation appears to be indepen-
dent of the availability of the correct answer since she never
confabulates on I dont know questions or on Linguistic
Semantic Memory questions where she was able to provide
the correct answer only eight times out of 15.
Strategic retrieval and confabulation
Moscovitchs ( 1989, 1995: Moscovitch and Melo, 1997)
hypothesis of confabulation as a consequence of a deficit of
strategic retrieval is meant to be applied equally across all
domains of memory. In other words, confabulation should
equally affect episodic and semantic memory if both epi-
sodic and semantic memory tasks require the involvement
of the strategic retrieval process. This hypothesis seems to
have ditliculties in accounting for the present case and for
two previously described cases (Dalla Barba et a/., 1990;
Dalla Barba, 1993a,b) in which confabulations were
confined to the domain of episodic memory. However,
Moscovitch (1995) argued that Dalla Barba and
co-workers patients showed confabulation in episodic
memory but not in semantic memory because the tests
used to ;issess episodic memory made greater demands on
strategic retrieval processes than those used to assess
-
Normal controls
Confab. Correct (SD) Conlhh
5
33
40
6
0
semantic memory. Accordingly, and in order to attempt to
equate the strategic retrieval demands of semantic and
episodic memory, Moscovitch ( 1995; Moscovitch and
Melo, 1997) devised a semantic, historical ver:;ion of thc
word-cue test developed by Galton ( 1879) and revised by
Crovitz and SchiRman (1974) to complement the tradi-
tional episodic version of the test. Using thi.. techniquc
Moscovitch (1995: Moscovitch and Melo, 1997) found that
patients with damage to the ventromedial portion of thc
frontal lobe confabulated on both the episodic and thc
semantic version of the test. We used Moscovitchs tech-
nique i n order to ascertain whether the virtual absence 01
confabulation in semantic memory i n GA was ;in artetact
due to the poor involvement of strategic retrieval i n
answering the semantic questions of the Confabulation
Battery. In the episodic version of the test, G A and sevcii
control subjects were presented with eight cue-words and
asked to produce a specific detailed account of ;in auto-
biographical event related to that word. I n tlte semantic
version of the test the procedure was identical to that used
in the episodic version, except that this time i n response t o
the cue-word subjects had to describe i n detail a historical
event which occurred before they were born (see Appendix
1). Responses were scored on a 0-2 scale with two points
for a detailed description of the event. one point for ;I less
detailed description and zero points for a general statement
or if nothing was provided. All responses werc scored by
two independent raters. Agreement between them was
100%1.Memories for both semantic and personal cue-words
were verified for confabulations, using historical sourcch
for the semantic test and questioning close relatives for the
episodic test. Table 3 shows GAs and normal controls
performance on this test. As seen in the table. G A scored
poorly on both versions of the test when compared to
normal subjects. Confabulations were present only i n the
episodic version of the test. In addition, confabulations i n
the episodic test were much less frequent ( I3%,):IS com-
pared to confabulations observed in episodic niemorj
(33%) and orientation in time and place questions (40%) 01
the Confabulation Battery. This difference may reflect thc
fact that the Confabulation Battery included specific ques-
tions that were probably more eflective as retrieval cuts
than the cue-words included in the Crovitz test. As far ;IS

Table 3. GA'b and norniiil controls' performance on thc Crovitr teat
Episodic
GA 0.9
Normal controls I .9 ( 0 . 2 )
Table 4. GA's and normal controh' performance ('XI)
Recognit ion
Correct
GA 56
Normal control5 I00
the low baseline level of confabulation in the episodic
memory version of the Crovitz test is concerned. this niay
have prevented finding confabulation in the semantic
version of the same task. Yet. however low the rate of
confabulation in the episodic version of the test, the results
of this test were dissociated from the results of the semantic
version of the same task where confabulation was never
observed. Therefore, the present results, i.e. the greater
amount of confabulation in the Confabulation Battery
where strategic retrieval is likely to be less involved than in
the Crovitz test and the absence of confabulation in the
semantic version of the Crovitz test, cast serious doubts
on a strategic retrieval deficit hypothesis as the cause of
Confabulation.
Source monitoring and confabulation
It has been argued (Johnson, 1991) that confabulation
might be the result of the disruption of mechanisms
involved in monitoring the origin of information so that
imagined events are mistaken for really experienced events.
In order to test this hypothesis we administered a yes/no
recognition memory task to GA followed by a source
monitoring task. The first task was devised in order to have
a baseline control for the experimental material. I n the
study phase 16 coloured drawings of cominon objects were
presented to GA and 10 control subjects. one after the
other, for approximately 3 seconds each. Immediately
following the presentation of the 16 items. the subjects were
presented with a list of 32 printed words, where half of the
words were names of the objects seen i n the study phase
and the other half were distracters, i.e. names of common
objects not presented in the study phase. Subjects had to
indicate the words corresponding to the objects seen in the
study phase. In the study phase of the second task, follow-
ing the same procedure as the previous one. the subjects
Confabulation and temporality 429
Semantic
I' 0of Mean score "I, of
c o n lit b ti l a t ion (SD) con fa h ti lii t i o n
13 0.6 0
(1 I .8 ( 0 . 2 ) 0
o n recognition niemor) and sotircc monitoring tasks
So ti rce nion ic o r ing
False
rrcogn i t ion Corrcct 'Seen' 'Imagined'
31 62 SO 50
0 I00 50 50
were first presented with a new set of 16 coloured drawings
of common objects. They were then read a list of 16 names
of common objects and for each name they were asked to
imagine the corresponding object. Immediately afterwards
they were presented with a list of 32 printed words, where
half of the words were names of the objects seen and the
other half were names of the objects they had to imagine.
For each word, subjects were asked to say whether that
word was the name of an object they had seen or imagined
in the study phase. Table 4 shows the results of these
recognition memory and source monitoring tasks. On the
recognition memory task GA's performance was at the
level of chance. However, her poor performance on this
task cannot be attributed to a tendency to recognize as
already seen any presented item, since the rate of false
recognition was 3 1%. On the source monitoring task GA's
perforniance was also very impaired, but there was no
evidence of a tendency to attribute imagined events to
direct perception or vice versa. I n fact, regardless of
whether the response was correct or not, GA gave 50'%1
of 'seen' responses and 50%) of 'imagined' responses.
Although these results show that GA is suffering from a
source monitoring deficit, GA's confabulation cannot be
attributed to the incorrect acceptance of any item as a
memory or to the tendency systematically to consider
imagined events as really experienced events.
It might be suspected, however, that the low rate of false
recognition, in the recognition memory task, and the lack
of the tendency to attribute to visual perception what was
imagined, in the monitoring task, are due to an etrect of the
experimental material. In fact, the items that were used.
drawings and names of common objects, are 'neutral
events', i n the sense that they are irrelevant to GA's daily
life and biography. I t could be the case that GA shows
ii tendency to recognize as true memories only events,
experienced or imagined. that concern her daily life and
 xh
13
biography. It' this is tlie case. GA should recognize iis
memories both really experienced episodes as well as her
conl'itbulatiotis. I n order t o check this possibility we adniin-
istercd ;I yes/no recognition memory task to G A i n which
she \v;is presented orally with sentences referring either to
iruc episodes I'rom her recent past. collected from her
husband. o r to contabulations she had produced i n pre-
\,ious testing sessions. Sentences representing true episodes,
c.g. 'todny for lunch you had boiled meat', and sentences
representing confabtilatory cpisodes, e.g. 'today for lunch
y o u cooked spaghetti'. were selected and presented to her
in ii ranclom order. Table 5 gives tlie results of this task.
< i A recognized its true memories 86'% of her confabu-
lations. whereas she recognized a s true memories only 43%
ot'episodes she had really experienced. From these results i t
clearly emerges that GA does not show tlie tendency to
accept ;is true any candidate autobiographical memory
because 01' ii delicit of monitoring processes. She actually
seems to xihere much more to her confahulatory past than
t o lier real piist. GA's tendency to recognize her confabu-
lation. rather tlim episodes she had really experienced. as
true may bc cxplained ;IS follows: GA's confabulation
I-etlects ;I tendency to consider her general habits or per-
soniil semantic information, which are presumably more
strongly represented in long-term memory. 11s true autobio-
graphical episodes. This will be discussed more extensively
1iitcr.
Confabulation and temporality
Accord i ng to the o bserva t i on that con fa b u la tors ofte ti
confabuliite i n orientation i n time and place tasks as well as
when they iirc questioned about their personal future, it has
becn proposed (Dalla Burba. I W 3 a ) that confabulation is
not only ;I disturbance of memory but also involves all
subjectively experienced tempornlity extending to tlie per-
sonal present and t o the personal future. In order to test
this hypothesis we administered two further tasks to G A
and seven control subjects. I n the first task we selected 10
episodic memory cl uestions concet-ning her personal past
and I0 questions concerning her personal future (see
Appendix 2). Episodic memory questions and questions
concerning the I'uture were matched for proximity to the
moment when tlie questions were asked. Answers were
scored ;IS 'correct', 'wrong' or 'confabulation', according to
wliethcr they were consistent with information (collected
fi-om her liusband) concerning the patient's past, present
situation and probable future activities. I t might be argued
that a s far ;IS questions about the personal future arc
concerned, any possible answer is ;I confabulaticln. since 13)
definition the future is only 'probublc' and there is iii
principle no 'correct' mswer to questions iibotit the t'u~urc.
Yet, answers concerning the personal future c..in be dcti-
nitely confabtilatory when they s h o w ;I m:irked tliscrcpa ncy
or ii real con t rndict io t i with what ;I prccl ic tcd fit t LI re act io I 1
might be, in view of the present situation. For cwniple. to
the question 'What iire you going to d o in ii tbw miniitcs'?'.
G A answered 'I will go home to cook the suppei'. Since 1ic.i.
disease, she actually never cooked and she wiix living o w
and I I half hours from tlie hospital. To the quc5,tion 'Whiii
are you going t o do tomorrow?' she replied * I will go out
shopping alone by car'. She never did that since Iicr discaw
and she would certainly not do it the following d a y . As
shown in Table 6, GA's confabulations extended equally t o
questions concerning episodic memory iitid licr pcrson;il
future.
The second task was ii 'future' version of tlic (.'rovit/ test.
I n this task tlie same eight cue-words used in the c1;issic
episode version of the test were given to G A m l t o control
subjects. For each word. subjects were asked to produce ;I
specific, detailed account of ii personal project relutcd to
t h a t word. Subjects responses were scored on ;I 0 2 poinis
scale following tlie procedure employcd for the episodic
and semantic versions of the test. ('r~nliibit1;itions wcrc
detected according to the same criteria used in the previous
task. The results of this task iire presented #.)tiTublc 7
together with tlic results of the episodic and scm;tntic
versions of the test for comparison. The rewlts cleat-I!
show that G A also confabulated when i n response IO ii
cue-word she hod to provide ii personal project and indeccl
even more than when she was asked to provi~lc;ti1 a u t o -
biographical past event. GA's performance on these tiisks.
together with her confabulntions on orientation i n timc ;Itid
space tasks. provides evidence in support of tlic liypotlicsi:,
that confabulation involves all the subjectively cxpericnccd
temporality i ti its co ti t in ui t y and ti ni t y .
Discussion
The patient we have described sutl'crs from it rclntivcl!~purc
amnesic syndrome. in tlie sense t h a t tlie menlory impair-
ment is not associated with any other cognitive impiiirment
likely to interfere with memory pertormnnce. CiA pertimiis
well on tests of intellectual efficiency siicli iis Progrcssi\ c

Table 7. G A . 5 a n d iioriiiiil controls' performance on the ('1-ovitr test
GA
Mean score
(SD)
Episodic 0.9
Seinantic 0.6
F tit ti re 0.5
Matrices and the MMSE a s well as on tests of short-term
memory. She has an extensive lesion involving the i'rontal
lobe from the orbital to the dorsal areas, yet her perform-
ance on tests considered sensitive to frontal lobe lesions
varied greatly. ranging from normal on the Nelson's
version of the Wisconsin Card Sorting Test, the Cognitive
Estimates and the Tower of London. to massively impaired
on Verbal Fluency, Luria's sequencing tests and the Trail
Making Test. Her amnesia involves memory for both
personal and public episodes, a s well as learning of new
information as reflected by her poor performance on the
Wechsler Memory Scale and on tests requiring the retrieval
or recognition of personal or impersonal episodes. In
contrast. dthough her performance on tasks tapping
various types of general knowledge is poor. but not
dramatically impaired. her confahulatory syndrome does
not extend to tasks involving the retrieval of information
frotii general knowledge. In addition, she never confabu-
lates when she is required to answer questions. semantic
and episodic, to which the appropriate response would be 'I
don't know'. Taken together these findings suggest that
GA's confabulation does not reflect ii tendency to fill a gap
in memory (Barbizet, 1963; Talland. 1965; Sandson rt NI.,
1986: Kopeliiian. 1987) in order to avoid embarrassment.
since it ii ppears to be tot ;i 11y independent of the a v;i i la b i I i t y
of the correct answer. In other words, GA does not
confabulate in response to those questions to which she
does not have an answer.
Strategic retrieval and confabulation
One of the iiiiiis of this study was to assess the validity of
Moscovitch's hypothesis ( 1989, 1995; Moscovitch and
Melo. 1997) of confabulation a s the result of an impair-
ment of strategic retrieval. According to Moscovitch, coii-
fa hula t i oti sh o Lild ma ti i fest itself eq Lial I y i t i episodic a iid
semantic mcmory if episodic and semantic memory tasks
require the involvement of strategic retrieval. Moscovitch
(1995) argued that a 'reason for the greater prevalence of
cont'abulation about episodic memory is that retrieval of
episodic memories in the laboratory and i n real life are
likely to make greater demands on strategic processes than
retrieval of semantic memories' (p. 16). In support of the
idea that confabulation is the result of ii strategic retrieval
deficit. Moscovitch (1995) Moscovitch and Melo (1997)
provides evidence of patients who confabulate both in the
Confabulation and temporality 431
Normal coii t rol s
":, of
con fa b ula t ion
I3 1.9 ( 0 . 2 )
0 1.8 ( 0 . 2 )
50 I .h (0.3)
episodic and the semantic version of the Crovitz test. In the
present study we were not able to replicate Moscovitch's
results. In fact, the study of GA's case shows that she does
not confabulate on seniantic memory tasks even when a
considerable amount of strategic retrieval is required, a s in
the semantic version of the Crovitz test. Further evidence
against ;I strategic retrieval deficit account of confabulation
conies from GA's performance on the Cognitive Estimates
test. This test requires that the subject answer questions
which d o not have a readily available correct answer (e.g.
'How many camels are there i n Holland'?'). and the subject
in order to provide a plausible answer. is constrained to
make ii judgement involving the activation. co-ordination
and control of several kinds of semantic information. Thus,
strategic retrieval is expected to be prominently involved i n
the prodiiction of a plausible answer on this test. However,
GA not only does not confabulate on this test. but her
performance is perfectly normal. So. according to the
present results, a strategic retrieval deficit hypothesis as the
origin of confabulation is not contirmed. Consequently we
argue that confabulation confined to episodic memory. as
in GA's case and. possibly, i n other studies (Dalla Barba
c't trl., 1990; Delbecq-Derouesne e / rrl.. 1990; Dalla Barba.
1993a.b: Dalla Barba r'/ ol., 1997) is an empirically proved
phenomenon and not the result of an experimental artefact
due to poor involvement of strategic retrieval.
Confabulation and monitoring processes
A second issue we wanted to address in this study concerns
the relationship between confabulation and processes
involved in iiionitoring the origin of information. Johnson
( 1991) argues that confabulation reflects a disruption of
processes involved in 'reality monitoring' (i.e. deciding
whether a memory represents something that actually
happened to you. or is a memory of an imagined event). A
disruption of this type of process would result in mistaking
imagined events for really experienced events. thus deter-
iii i ti i ng con fa b ul a t i o n . This kind of per spect i ve , however,
does not account for the fact that confabulation. as in the
present case and other studies (Dalla Barba P / ( I / . , 1990:
Delbecq-Derouesne e/ t i / . , 1990; Dalla Barba. 199321,b:
Dalla Barba ct a/.. 1997). can be confined to the autobio-
graphical aspect of episodic memory while it is absent in
episodic learning tasks and in tasks that require retrieval of
non-episodic information. I n fact, if ii disruption of reality
131 G. Dalla Barba ~t rrl.
monitoring were responsible for confabulation. confusion
between re:ally experienced and imagined events and infor-
motion should not be limited to autobiographical episodic
rclrievul but should extend across all memory domains.
Further doubts on the relationship between a deficit of
reality monitoring and confabulation are cast by GAs
pcrlormance on recognition and source monitoring tasks.
G A shows source amnesia as reflected by her performance
ar chance on the source monitoring task. However, she
does not systematically attribute to visual perception what
M ~ I Simagined or vice versa. In addition, on recognition
memory tasks she does not show a tendency to recognize as
~ l l r c i i d yseen any item she is presented with, as reflected by
rhe lo\+ rate of false recognition on the recognition memory
task. Thcreforc. dthough G A shows a source monitoring
deficit. it scenis restrictive to trace back GAs confabulation
to a general deficit of processes involved in reality moni-
toring. or to H systematic bias to recognize its part of her
past experience any present information. One possibility, as
suggested elsewhere (Dalla Barba, 1993a), is that distinct
reality monitoring processes iire specific to different types
o l ~retrieval. so that confabulation confined to episodic
memory would retlect the disruption of reality monitoring
processes operating in episodic memory retrieval, while
reality monitoring processes operating in other types of
retricval ;ire intact. This type of account, however, is
compromised by GAs performance on tlie task in which
she was asked to recognize sentences referring to true
cpisodes from her recent past or to confabulation she had
produced in previous testing sessions. The results of this
task show that GA recognizes her confabulations as true
memories, whereas she is at the level of chance in recogniz-
ing really experienced episodes. I f a deficit of processes
involved in reality monitoring were present in G A and
responsible for her confabulation, one would expect the
siinie pattcrn of yes/no responses both for sentences refer-
ring to confabulations and for those referring to true
in cniot-i es .
Confabulation and temporality
The most interesting aspect of GAs case study is that
her coiiliibulation does not only involve the retrieval of
personal past episodes and present orientation in time and
place, which is a known characteristic of some confabu-
lators, but nianifests itself also when she is asked to plan
her personal future. The extension of GAs conpdbulation
to a11 tlie three dimensions of temporality-past, present
and future-seems to be something more than a fortuitous
coincidence. GA confabulates when she has to answer
questions like What did you do yesterday? Where are you
now? What are you going to d o tomorrow? but she never
confiibulates answering questions concerning historical
events o r in traditional laboratory learning tasks. In this
sort oftask her poor performance is in fact due to errors of
omission and not to intrusions or even fleeting confabu-
lation. Also. she is very good on the Tower of London.
which is a planning task. What all these ta
common, and at variance with tasks where GA contabu-
lates, is that they do not involve subjective temporality, in
the sense that they do not require the involvenient 01 the
personal history, present situation and probable future 01
the subject who is performing them. I n contrast. <;A
.
cot i Fdbula t es when subjective tempo r;i I it y is i n vo I vcd i .c .
when she has to address her personal past, present and
future. Recently, Conway and Tacchi (Conway ;I nd Tacchi.
1996) described a patient. OP. who cotlf>tbtlliitt:d in auto-
biographical memory. They interpret 01s cot.tfiibuliitiotl
within the framework of Conways theory of autobio-
graphical memory (Conway, 1992; Conway, 1990; Conwiil
and Rubin. 1993). suggesting that their patients confabu-
lation is the result of a disruption of the self due t o thc
damage of frontally based control processes of memory
(Burgess and Shallice, 1996; Johnson. IWl). In their cieu
OPs confabulation is motivated by the purpose of protect-
ing the self in a time of stress and isolation. Thc problem
with this interpretation is that none of the elements invoked
to explain confabulation is conscious: the st:lf. control
processes and the purpose to protect the self are all
unconscious. We have shown elsewhere (Dalla Harba.
1993; Dalla Barba, 1997) that considering control processc\
of memory as operating outside consciousness poses tii:ijor
theoretical problems as this would necessitate the attribu-
tion of intentionality to the unconscious. The sittnc argu-
ment applies to the hypothesis of an unconscious sell :in11
of an unconscious purpose to protect this self. 10 iittributc
intentionality to an unconscious process ineatis to proc ide
it with the property of being . s i i h j w / , that is, 1.0 attributc
consciousness to the unconscious. Like i n ,.I game 01
Chinese boxes, the subject would contain another subjcci.
an honri~ncu/ir.r,provided with a shady and inaccessible
consciousness, that is busy solving problems such iis reject-
ing false memories and providing conscious conscioiisnehs
with the result of its accurate selection. According to these
criteria, can the homunculus distinguish betwci-n true and
false information? Does not the hoinunculus ;iIso need an
unconscious 11om u ncu 1us that he1ps it to Ji s t i ngu i sli
between true and false memories? Like this the Chinese
boxes game continues endlessly, and what we arc Iclt with
is just a r d i r c t i o r i d itzfiuimi.
When GA confabulates, she acts like somebody who is
aware of her past, present and future. The problem.
however, is that she is aware of a past, a present. and ;I
future that are false, in the sense that they do n o i
represent ii past, present and future coherent with her
actual situation. I t has been suggested that conliibulation
reflects a disturbance of chronology or of time-order
memory, so that real memories are confused iii tcmpor;rl
context (Korsakoff, 1889; Schnider ct ( I / . , 1996; Talland.
1961; Van der Horst, 1956; Victor o f r i / , , 1971). T h i h
interpretation, however, is challenged by rhe f ~ c ttlivi
inany amnesic and non-amnesic patients have disordered

Consciousness Long-Term
Storage System
I
- 1 I example, that tells that Y comes before Z and after X. They
Fig. 1. A hypothesis concerning the relation between memory. conscious-
ness and temporality. TC = temporal consciousness; KC = knowledge
consciousness: X. Y and Z = aspecific modifications of the Long-Term
Storage System. varying from less ( X ) to more stable (Z).
chronology. yet do not confabulate (Milner and Petrides,
1984; Nichelli, 1993; Stuss r t ul., 1978; Vriezen and
Moscovitch, 1990). Accordingly, it does not seem that
GAs pathological awareness of her personal temporality
can be simply traced back to a disorder of time-order
memory. Yet it remains to be explained of what GA is
aware, if not of real memories in a wrong temporal context,
when she is confabulating about her past, her present and
her future. GAs confabulation is typically plausible, or
semantically appropriate (Dalla Barba, 1993a) in the sense
that a hypothetical observer not familiar with GAs history,
background and present situation could hardly tell whether
GAs reports are confabulatory or not. She would claim,
for example, that the day before she went shopping or that
after the testing session she will prepare a meal for her
family-which are acts consistent with her life before her
illness. So this pattern of behaviour suggests that, at
variance with classic amnesic patients that are lost in the
present, in the sense that they are no longer aware of a
personal past and of a personal future (see for example
Tulving (1 ul., 1988; Tulving, 1989). GA is perfectly aware
of her personal temporality. but she mistakes her habits or
personal semantics for her real past, present and future. In
this sense G A is remembering another past and planning
knot her future.
What GAs case shows is that confabulation involves all
the subjectively experienced temporality in its continuity
and unity extending from the past through the present to
the future. If this is the case, we need to explain what
subjectively experienced temporality is. One way to attempt
to answer this question is to make a set of assumptions that
are shown schematically in the model represented in Fig. 1.
Assuinption I
Events produce atemporal and aspecific modifications i n
the Long-Term Storage System (LTSS). These modifi-
cations, represented in Fig. 1 as X, Y, and Z, are atemporal
in the sense that they do not contain any information
concerning time. They do not represent the past, the
Confabulation and temporality 433
present or the future, nor are they organized according to
the order of succession, i.e. there is nothing in Y, for
are aspecific in the sense that they do not contain any
information specifying that they are representing episodes,
meanings, rules, procedures, algorithms, etc.
Assiriiiption 2
The modifications in the LTSS can be more or less stable
and more or less vulnerable depending on a number of
variables. These variables act at the encoding and storage
level and include, among others, attention at encoding,
emotional value of encoded and stored event, depth of
encoding, rehearsal and repeated experience of the same
event or similar events.
Assuniption 3
Consciousness means to be conscious q/ sorilething in LI
.spec.ijzc N U J . That means that consciousness is not an
aspecific dimension that passively receives and becomes
aware of different types of already specified information,
but rather that different types of consciousness exist, each
representing an original and irreducible way of addressing
the world. Different types of consciousness include, among
others, temporal consciousness (TC) and knowing con-
sciousness (KC). TC means to become aware of something
as part of a personal past, present or future. KC means to
become aware of something as a meaning or as an element
of impersonal knowledge or information.
Assuniption 4
TC can, and usually does, interact with the less stable and
more vulnerable modifications of the LTSS in order to
remember the past, be oriented in the present and plan the
future.
According to this model, confabulation in episodic memory
(e.g. Dalla Barba, 1993a; Dalla Barba rt ul., 1990). in
orientation in time and place and in personal future
planning tasks, as in GAs case, is the result of a condition
in which TC is still intact but cannot use less stable
modifications of the LTSS but only more stable ones, in
order to remember the personal past, be oriented in the
present and plan personal future. In other terms, this is a
condition in which TC is still working but is no longer able
to accomplish its usual task, i.e. to operate a sort of fine
grain search in the LTSS and use less stable modifications
in order to set up a personal temporal workspace. What
T C does instead in this condition is to use more stable
modifications of the LTSS, with the result that habits or
personal semantic information are considered in a personal
temporal framework. When asked what they have done the
previous day or what they are going to d o the following
day, confabulating patients of this type typically answer
with memories and plans that they usually have in their
daily life. Although admitted to the hospital, they will say,
434 C. Dalla Barba ct cd
for example, that the previous day they went out shopping
and that the following day they will be visiting some friends.
acts that presumably were part of their routine life.
I t could be argued that patients who confabulate in
episodic memory, orientation and planning tasks are not
necessarily conscious of a confabulatory past, present and
future but rather they simply produce the more plausible
iitlswer without having a subjective experience of remem-
bering, of being in that place at that time or of planning
their actions. If this were the case our account of confabu-
lation in the past, present and future would be dismissed
because TC would not play any role at all in confabulation.
Yet there is evidence that patients who confabulate actually
do become aware of their confabulatory past, present and
future. For example [Dalla Barba (1993a) following the
procedure described by Tulving (1985)], the patient MB
was asked to attribute a 'remember' or a 'know' judgement
to his confabulation in episodic memory task; he system-
atically gave 'remember' judgements. Also, the same
patient showed he was ready to carry out his confabulatory
plans (see also Baddeley and Wilson, 1986; Moscovitch,
1989). I n addition, from a clinical point of view, confabu-
lating patients do not look like subjects who produce their
'best guess' in answering questions, but rather they seem to
adhere completely to their confabulatory reports. There-
fore. although more experimental evidence is suitable, it
does not seem to be misleading to consider confabulation
i n episodic memory, orientation and future planning tasks
a s a condition in which TC can interact only with more
stable modifications of the LTSS.
I t is well known that patients sometimes also confabulate
i n some semantic memory tasks (e.g. Moscovitch, 1995;
Moscovitch and Melo, 1997). According to our hypothesis,
confabulation in semantic memory reflects a condition in
which KC is still there but is unable to use more stable
modifications of the LTSS. When this condition occurs
patients confabulate i n tasks tapping general semantic
knowledge, such as knowledge of public and historical facts
and knowledge of famous people (e.g. Moscovitch, 1995;
Moscovitch and Melo, 1997), but they do not confabulate
i n semantic tasks that require the use of more stable
modifications of the LTSS, such as word definition tasks.
Sonie patients, however, also confabulate in word defi-
nition tasks (Dalla Barba, 1993b). They also usually
produce confabulation in episodic memory tasks that
are described as fantastic (Berlyne, 1972), implausible
(Baddeley and Wilson, 1986), incoherent (Kopelman, 1987)
or semantically anomalous (Dalla Barba, 1993a). Confabu-
lation in word definition tasks and fantastic confabulation
in episodic memory tasks reflect a condition in which the
modifications in the LTSS are degraded so that the material
used by TC and KC is per s o incoherent. The result is that
the semantic content of confabulation in episodic memory
is anomalous so that confabulation takes a fantastic and
bizarre form and so does confabulation in word definition
and other semantic tasks.
Acknowledgements
This article is a modified version of part of iin invited paper
presented at the XV European Workshop on Cognittvc
Neuropsychology, Bressanone, Italy. January 1097. and at
the Cognitive Neuroscience and Memory C'onferencc.
Stockholm, June 1997. We are grateful to manly people 111
the audience at those conferences and to Martin Conwit!
and an anonymous reviewer for helpful comment:, on i i n
earlier draft. This work was partially supported by thc
Institut National de la Sante et de la Rechercht: M ~ d i c a l .
References
Alamanno N. Della Sala S. Loiacona M. Pasetti C. Spinnlcr H. Prohlcni
solving ability in aging and dementia: noriii;itiw data or ii non-\erbal
test. Italian Journal of Neurologic;tl Scicncc 19x7: 8: I I 10.
Alexander MP. Freedman M. Amnesia after :interior coiiiiiitiiiic;itiii~
artery aneurysm rupture. Neurology 1984: 34: 752 7.
Baddeley A. Your memory. A user's guide. London: M ultinicdi;i
Publications, 1982.
Baddeley A. Wilson B. Amnesia, autobiographical iiieniory and c o i i I ' ~ i h ~ ~ -
lation. In: Rubin. DC, editor. Autobiographical niemor:y. Canihritlgc.
Cambridge University Press, 1986: 225-52.
Barbizet J . Defect o f memorizing o E hippocampal maminill;iry origin
Journal of Neurology. Neurosurgery and Psychiatry 1963; 26: 127 3 5 .
Benson D F , Djenderedjirin A, Miller BL ( ' I trl. Neural b
lation. Neurology 1996; 46: 1239-43.
Berlyne N. Confabulation. British Journal o f Psychiatr) 1972; 120. 31 'J
Bonhoeffer, K. Der Korsakowschc Syrnptonicnkoinplcu in w n c n 1 % ~ -
ziehungen zu den verschiedenen Kr~tnkhcitsfornien.Allgemeinc Zcit t i n ?
Psychiatrie 1904; 61: 744-52.
Burgess PW. Shallice T. Conlitbulntion and the control 1.11' rccolltcti~ii~
Memory 1996; 4: 3 5 9 4 1 I
Cermak LS. Uhly B. Reale L. Encoding specificity in the iilcoht)I~c
Korsakoff patient. Brain and Cognition 19x0: I I : I I 9 27.
Conway MA. A structural model ol' autobiographical mtmoi-!
In: Conway MA, Rubiii DC, Spinnler H. Wagenaai- W A . editor\
Theoretical perspectives on autobiographic;il memory. Ihrdi-ech13'flit
Netherlands: Kluwer Aciidemic Publishers. 1991: I67 94.
Conway MA. Autobiographical memories mid aut~)bi1,gr;ipliie~il
knowledge. In: Rubin DC. editor. Remembering o u r past: studit\ i n
autobiographical memory. Cambridge: Cambridge Clnivcrbity h e > \ .
1996: 67-93.
Conway MA. Rubin DC. The structure o f ati[obiogra[illic;II meinor) 111
Collins AE, Gathercole SE, Conway MA. Morris PEM. etlitoir
Theories of memory. Hove. LJK: Lawrence Erlbauni Associ;itca. 1 0 c j 7
103 37.
Conway MA, Tacchi PC. Motivated confithulation. Ncui ocitsc IWO. 2 .
32539.
Costa M, De Renzi E. Fnglioni P. U n questionario ilaliano per 1 ~ 1
studio della memoria retrograda. Archivio di Psicologia. Ncurologia I'
Psichiatria 1989: 5 0 : 735-55.
Crovitz HF. Schiffman H. Frequency o f episodic nieiiioric~;is l'iiiiction I ) I
their ages. Bulletin of the Psychonomic Society 1974: 4: S17 I X .
Dalla Barba G. Confabulation: knowledge and recollectioii espcricncc.
Cognitive Neuropsychology 1993a: 10: I 20.
Dalla Barba G. Difercnt patterns of confabulation. Cortex l993b. 2 9
567- 8 I .
Dalla Barba G. Confabulation and consciousness. In: X V Europc;iii
Workshop on Cognitive Neuropsychology, Bressanonc. lt;rIk. 1097.
Dalla Barba G. Cipolotti L. Denes G. Autobiogruphic;il ni8:mory l o s s ;ind
confabulation in Korsakotrs syndrome: ;I case report. Cortcx 1900; 10
525-34.
Dalla Barba G. Boise M-F, Bartolomeo P. Bachoud-LOLi A-C' C'oiil'aliii-
lation following rupture of posterior communicating arter) . C ' o r i ~ > \
1997; 33: 563-70.
Davies A. The influence or age oii Trail Making Tcsi pcr-li)rrii;iric.~~
Journal of Clinical Psychology 1968: 24: 96 8.
De Renzi E. Faglioni P, Previdi P. Spatial ineniory and hemispheric IOCLI\I
of lesion. Cortex 1977a: 13: 424 33.
 Con fa b ula t i on and tempo ra 1it y 4 35

De Renri E. Faglioni P. Ruggerini C. Provc di iiieiiioriii \crb;ile Tul\ing L:. Schacter DL. McLacIiliin DR. Moscovitch M Priming
di inipiego clinico per lii diagnosi clinica di :iiiinesiit. Archivio (11 of seniitntic ailtobiographical knowledge: H ciibe study o f retrograde
Psicologiii. Neurologia e Psichiatria 1997h: 38: 303 18. amnesia. Brain and Cognition 1988: 8: 3 2 0 .
Delbecq-Derouesne J. Beauvois MF. Shallice T . P r e s c r ~ ~reciill
l \erstis V;in clcr H o i - s t L. Lc sens de la teinpor;ilisatioii pour la niL;moirc ct pour
impaired recognition. Brain 1990: 113: 1045 153. I'orientation. Ecolution Psychintrique 1956: 1: 18') 205.
Folstein MF. Folstein SE. McHtigh PR. 'Mini Mental State': a practical Victor M. Adams RD. Collins G H . The Weriiicke Korsakolf syndrome.
method for grading t h e cogniti\e state of patients foi- t h e clinician. Philadelphiu. PA: Davis. 1971,
Journal 01' Psychiatric Research 1975: 12: 189-98. Vilkki J . Amnesic syndrome after surgery for antcrioi- coiiimiinic;rting
Galton F. Psychometric cxperirnents. Brain 1879: 2: IJC)6 2 . artery ;tneurysm. Cortes 1985: 21: 431 44.
Irk E. Wowra B. Kunert HJ. Kunze S. Memorq disturbances fnllowing Vriexn ER. Moscovitch M. Memory f o r temporal order and con-
anterior communicating artery rupture. Annals of Neurology I 997: 3 I ' ditional ;issociative-lcarning i n paticnts w i t h Parkinson's disease.
473 80. Netii-opsycliologiit 1990: 28: I283 93.
Johnson MK. Reality monitoring: evidence from confahulation i n organic Warrington EK. Recognition memory test. Windsor. LIK: NFER-Nelson.
brain disease patients. In: Prigatano GP. Schitcter DL..editors. Aware- 19x4.
ness o f dclicit after brain in.jury. New York-Oxfnrd: Oxl'nrd Llniversit!, Wcinstcin E. Lyerly 0. Conl"ibul;ttion folloning hr;iin injury. Archive5 of
Press. 1991 General Psychiatry I96X: 18: 348-54.
Kaptir N. Coughlan AK. Confahulation after frontal lobe dysftinction. Wyke M. Warrington E. An experimental analysis of conl'abulation i n il
Journal 01' Neurology. Neurosurgery and Psychiatry 1980: 43: 461 - 3 . case of KorsakolTa syndrome using ii t;ichistoscopic method. Journal of
Kern R. Van Grop W. CLininiings J. Brown W. Osato S. Confabulation i n Neurology. Neurosurgery and Psychiutry 1960: 23: 327 - 3 3 .
Alzheimer's Disease. Brain a n d Cognition 1997: 19: I72 82.
Kopelnian M D. Two types of confabulation. Journ:il of Neurnlogy. R i ~ c ~ c , i ~0' 1c1d 13 J////c,. 1997; rc.sr/h/rri/tc~rlO / I 20 ,-lu,yl/.s/, 1997:
Neurosurgery and Psychiatry 1987: 3 1 : 1482-87. clc.c.r'ptt'rl 0 1 1 3 .s"f"lllhc~r. I997
Kopelinan MD. Guinan EM. Lewis PDR. Delusional memory. confabu-
lation, a n d frontal lobe dysfunction: ii case study in De CIPranlhault's
syndrome. Neurocase 1995: I : 71 7.
Korsakotf SS. Etude iiic;dico-psychologiquesur tine foi-me dcs ni;hrlirs de
la meninire. Revue Philosophique 1889; 28: 501 -30.
Luria AR. Higher cortical fiinctions in man. London: T ~istock.L 1060.
Luria AR. The neuropsychology of iiiemory. New Y o r k : John Wilq .
1976. Appendix I . The list of w ~ r d stised t o cue episodic (;iiitohiogr:ipliical) and
Mercer B. Wapner W. Gardner H. B e n m i F. A 5t~1d)of conf~thulation. wniintic (historical) memorie\ i n the CrovitL test
Archives of Neurology 1977: 34: 479 3 3 .
Milner. B. Petrides. M. Behavioural eHects on frontal lobe lesions in nxtn. Episodic/' 1. tit urc' Semantic
Trends i n Neuroscience 1984: 7: 403-7.
Moscovitch M. Confabulation and the frontal system: strategic \ c r s t ~ s
associative retrieval in neuropsychological theories of memor). In: Train Assassination
Roedinger HL. Craik F1. editors. Varieties of memory and conscious- M ountaln Attack
ness: Essaq i n honor of Endel Tulving. Hillsdale. NJ. Lawrence Letter Discovery
Erlbaum. 1989. Mone) Train
Moscovitch M. Confithulation. I n : Schacter DL. editor. Memory Cinema War
distortion. Cambridge. MA: Harvard University Press. 1095: 2 2 6 5 1 Resta tirant President
Moscovitch M. Melo B. Strategic retrieval and the frontal lobes: evidence H ospi t ;II Kevolutinn
from conktbulation and amnesia. Ncuropsycliologia 1997; 35: I(1I7 34. MiirkcI King
Nelson HE. A modified card sorting test s e n s i t n e lo frontal defects.
Cortex 1976: 12: 313 23.
Nichclli P. Human temporal information processing. In: BolleI F. Appendix 2. Questions concerning personal past ;tiid future (bold type
Grafnian G . editors. Handbook of ncLIropsycliolog!. Ani.;terd;lln: s l i o ~ squestions that elicited confabulation)
Elsevier. 1993.
Norman DA. Shallice T. Attention t o action: willed and autnniittic control
-
of behavior. I n : Center for Human Information Processing (Techiiic;iI Po\/
report No 99). 1980. Reprinted i n revised form in: Davidson RJ. I . What did you do a few minutes ago?
Schwarz GE. Shapiro D. editors. Consciousness and self-regtilntion. 2. Who did you see this morning?
Advances in research and theor). Vol. 4. New York and London: 3. What did you eat for dinner yesterday?
Plenum Press. 1986: 1 18. 4. What did you do yesterday?
Papagno C. Muggin S. Conlibulation: dissociation between e v e r y h y life _5 . What did y o t i d o the d a y befnrc yesterday'!
and neuropsychological performance. Neurocase 1996: 7. I I I I X . 6 H o w did y o u spend last Christmas'?
Sandson J. Albert ML. Alexander M . Confabulation i n aphusia. Cortcs I. Do you remember the last time you went to see a doctor?
1986; 2': 6 2 1-6. X. Do you rememher the last time ) n i l went to visit a friend o r i~
Schnider A. von Diniken C. Gutbrod K. The mechanisms o f spontaneous relative'!
and provoked confabulations. Brain 1996: I 19: 1365-75. 9 Do you reinember the last time you went to it restaurant'!
Shallice T. Specific impairments of pliinning. Philosophical Transactions 10. Do you remember the last time you took vacation?
of the Royal Society of London B Biological Science 1982: 29X:
199 209. Fti/i/,
Yl'

Sh;illice T. Evans M. The involvement of the 1.rontal lobe i n cognitive I . What are you going to do in a few minutes?
estimation. Cortex 1978: 14: 294 303. 2. Who are you going to see this evening?
Stuss DT. Alexander MP. Liherman A. Levine H. An extraordinary l n r m 3. What are you going to eat this evening?
of confabulation. Neurology 1978: 28: I 160-72. 4. What are you going to do tomorrow?
Talland GA. Confabulation in the Wcrnickc Korsakolf syndronie. 3. What ;ire y o t i going to do the day after t o m o r r o w ?
Journal of Nervous and Mental Disease 1961: 137: 361 -81. 6. Htm will yoti spend next Christnias?
Talland G A . Deranged memory. New York: Academic Press. 1965. I. When will he the nest time you visit a doctor?
Tulving, E. Memory and consciousness. Canadian Psychology 1985: 26: 8 . W h t n will be the next time you visit ;I friend?
1-12, 9. When will he the nest time y o u go to ii restaui-ant?
Tulving E. Remembering and knowing the past. American Scientist I9X9: 10. When will be the nest time you go on Facation?
361-6.
Confabulation: Remembering 'another'
past, planning 'another' future

G . Dalla Barba, J. Y. Cappelletti,

M. Signorini and G . Denes
Abstract
111 t l i i h s t u d y \\c desci-ihc ;I p;rticnt. GA. ulio developed iui amnesic-
L.l,iit'~ihiil;itory s) ndi.ome. following 21 suharachnoid haemorrhage due to
I tipturc ( 1 1 the anterior c(iiiii1itiiiic;itiiig artery. As far ;is tlie performance
1111 mcnioi-y t a s k s h i i s concerned. GA's conlihulation WLIS found to be

rc\iricted to the ;ititohiograpliic.al aspect of episodic memory. Confabu-

l;ition did not manilest itsell' in episodic learning tasks nor on tasks
t;iiiping \ iiriotis kinds 01' sem;intic knouledgc. I n contrast. GA confabu-
I;itcd in orientation in time ;ind place tasks and a l s o in tiisks where she was
I . ~ L I L I I I . ~ ~ t o pliiii Iicr pcrsoniil ftittirc. GA's conlhhulatiori could not he
;i~cuuntcdIOr in term5 o f :in impairment ol'str;itegic retrieval or o f reality
iiionitorinp 1iroccsscs. I t is htrggested that GA's confahulation reflects ii
pathologiciil iiw;ireiicss of pcrsoii;iI temporalit y.

Journal
wclll c1ciise 1097: 3 : 425 36

Neurocase Reference Number:

00.3

Primary diagnosis of interest

Siih;ir;iclinoiil haemorrhagr. I-:rontal strake

Author's designation of case

(i .I

Key theoretical issue

~'oiiflihul;itionrellecth ii pathological iiwi~reiiessof persoii;il temporal-
It!

Ac.1. I I . I W ~ \ c~iiif~i1~iil;ition:
: iiiiinesia: consciousness: temporality

Scan, EEG and related measures

c'r. hiiti. wr
Standardized assessment
I'M-47. Wcchsler Memory Scale. Mini-Mental State Examination. Modi-
lied C'arJ Sorting Test. Cognitive Estimates. Tower of London. Luria's
Sequencing ' l c s t s . Letter Fluency. Category Fluency. Trail Making Test,
Digit Span. C'orsi Test. Warrington Recognition Memory Test. Public
l . \ c i i t a Qurstionn;iire, 10-hord Lisl Learning. Boston Naming Test.
Seni;intic Anomalies Detection. Confabulation Battery

Other assessment
C ' I m itc Test (moililied). Personal PastlFuttire Interview. Source Monitor-
1ng l C \ l

Lesion location
MIII: 1iIatcr;il frontal degenerative areas extending from the orhito-
l'rontiil regions to tlie dors;il iirciis. involving the anterior cingulum and
t h e :interior two-thirds of t h e corpus collosum

Lesion type
L';l\clll~ll

Language
I:nglish