Research report

Do you remember what you did on March 13, 1985?

A case study of confabulatory hypermnesia
Gianfranco Dalla Barba
a,b,c,d,
* and Caroline Decaix
e
a
Inserm U. 610, Paris, France
b
Universite Pierre et Marie Curie-Paris 6, Paris, France
c
AP-HP, Hopital Henri Mondor, Service de Neurologie, Creteil, France
d
Dipartimento di Psicologia, Universita` di Trieste, Trieste, Italy
e
AP-HP, Hopital Saint Antoine, Service de Neurologie, Paris, France
a r t i c l e i n f o
Article history:
Received 8 March 2007
Reviewed 25 May 2007
Revised 7 January 2008
Accepted 11 March 2008
Action editor Michael Kopelman
Published online 5 June 2008
Keywords:
Confabulation
Episodic memory
Semantic memory
Temporal consciousness
a b s t r a c t
We report on a patient, LM, with a Korsakoffs syndrome who showed the unusual ten-
dency to consistently provide a confabulatory answer to episodic memory questions for
which the predicted and most frequently observed response in normal subjects and in con-
fabulators is I dont know. LMs pattern of confabulation, which we refer to as confabula-
tory hypermnesia, cannot be traced back to any more basic and specic cognitive decit and
is not associated with any particularly unusual pattern of brain damage. Making reference
to the Memory, Consciousness and Temporality Theory MCTT (Dalla Barba, 2002), we pro-
pose that LM shows an expanded Temporal Consciousness TC, which overows the limits
of time (Do you remember what you did on March 13, 1985?) and of details (Do you re-
member what you were wearing on the rst day of summer in 1979?) that are usually re-
spected in normal subjects and in confabulating patients.
2008 Elsevier Srl. All rights reserved.
1. Introduction
Confabulation is a symptom observable in some patients with
memory impairment. At a general level, confabulation can be
referred to as a symptom observable in amnesic patients who
are unaware of their memory decit, and which consists of ac-
tions and verbal statements that are unintentionally incon-
gruous to the patients history, background, present and
future situation (Dalla Barba, 1993a).
This rather infrequent disorder is classically described in
Korsakoffs syndrome (Benson et al., 1996; Bonhoeffer, 1904;
Cermak et al., 1980; Dalla Barba et al., 1990; Korsakoff, 1889;
Mercer et al., 1977; Schnider et al., 1996a, 1996b; Talland,
1961; Wyke and Warrington, 1960). But confabulation is also
seen in patients suffering from ruptured aneurisms of the an-
terior communicating artery, subarachnoid haemorrhage or
encephalitis (Alexander and Freedman, 1984; Dalla Barba
et al., 1997a, 1997b; Delbecq-Derouesne et al., 1990; Irle et al.,
1992; Kapur and Coughlan, 1980; Kopelman et al., 1995; Luria,
1976; Moscovitch, 1989, 1995; Papagno and Muggia, 1996;
Schnider et al., 1996a, 1996b; Stuss et al., 1978), head injury
(Baddeley and Wilson, 1986; Dalla Barba, 1993b; Schnider
et al., 1996a, 1996b; Weinstein and Lyerly, 1968), Binswangers
Encephalopathy (Dalla Barba, 1993a), Alzheimers disease and
* Corresponding author. INSERM U. 610, 47 boulevard de lHo pital, Paris F-75013, France.
E-mail address: dallabarba@chups.jussieu.fr (G. Dalla Barba).
avai l abl e at www. sci encedi r ect . com
j our nal homepage: www. el sevi er . com/ l ocat e/ cor t ex
0010-9452/$ see front matter 2008 Elsevier Srl. All rights reserved.
doi:10.1016/j.cortex.2008.03.009
c o r t e x 4 5 ( 2 0 0 9 ) 5 6 6 5 7 4
frontotemporal dementia (Dalla Barba et al., 1999; Kern et al.,
1992; Nedjam et al., 2000, 2004) and aphasia (Sandson et al.,
1986). Confabulation may also be observed, on occasion, in
normal subjects (Burgess and Shallice, 1996; Dalla Barba
et al., 2002; Kopelman, 1987).
Within confabulatory behaviour some distinctions have
been proposed. Berlyne (1972), following Bonhoeffer (1904),
distinguished between momentary and fantastic confab-
ulation. According to Berlyne (1972), momentary confabula-
tion, which invariably must be provoked, is autobiographical
in content, refers to the recent past and consists of true mem-
ories displaced in their time context. In contrast, fantastic
confabulation is spontaneous and it is autobiographical,
but entirely invented. The principal content is invariably gran-
diose and seems to be related to wish fullment and prestige
seeking (Berlyne, 1972, p. 33). Subsequently Kopelman (1987)
argued that Berlynes description was inadequate because it
failed to isolate the distinguishing features of the two types
of confabulation and proposed to focus on the modality of ap-
pearance by referring to provoked and spontaneous con-
fabulation. Provoked confabulation would reect a normal
response to a faulty memory, whereas spontaneous confabu-
lation would reect the production of an incoherent and con-
text-free retrieval of memories and associations (Kopelman,
1987, p. 1482) resulting from the superimposition of frontal
dysfunction on an organic amnesia. Schnider et al. (1996a,
1996b) have suggested that the term spontaneous confabula-
tor should only be attributed to patients who show the ten-
dency to act on their confabulations, since, they argue,
action is the only observable symptom that cannot be directly
provoked by the researcher or some other external trigger.
However, the distinction between spontaneous and provoked
confabulation has not gone unchallenged. In fact it has been
shown that spontaneous confabulations are not necessarily
fantastic and grandiose (Dalla Barba, 1993a; Dalla Barba
et al., 1997a, 1997b) and that provoked confabulations can be
fantastic and incoherent (Dalla Barba, 1993b; Dalla Barba
et al., 1998). Accordingly, the line drawnbetweenspontaneous
and provoked confabulations often appears to be quite an ar-
bitrary decision. Some authors, in fact, argue that it is not
meaningful to impose a dichotomy, and that confabulation
should be regarded as a continuous variable, ranging from mi-
nor distortions to the more fantastical (Dalla Barba, 1993b;
Fischer et al., 1995; Kapur and Coughlan, 1980).
As far as the mechanisms of confabulation are concerned,
three major approaches have been proposed.
(1) Johnson argued that confabulation reects poor source
monitoring, or reality 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 (Johnson,
1991). Damage to frontal/executive functions would result
in an impairment of judgment processes involved in real-
ity monitoring and so in confabulation. This interpretation
of confabulation is consistent with the idea that confabu-
lation is a form of source amnesia (Moscovitch, 1989;
Schacter et al., 1984) combined with misattribution of tem-
poral and spatial context (Schacter, 1987). However, in
a more recent paper, Johnson et al. (1997) demonstrated
that reality monitoring, or source monitoring, was equally
disrupted in a confabulatory patient, and in non-confabu-
lating patients with frontal lobe damage. The latter nd-
ings lead Johnson and colleagues to propose that a reality
monitoring, or source monitoring, decit may occur with
confabulation but is not the only factor involved in the
genesis of confabulation (Johnson, 1997).
(2) Moscovitch (1989, 1995), Moscovitch and Melo (1997) and
Gilboa et al. (2006) have proposed that confabulation is
the result of a decit of strategic retrieval. They propose
a distinction between two components of retrieval. One,
associative retrieval, is relatively automatic and indepen-
dent from frontal functions. The other, strategic retrieval,
is self-initiated, goal-directed, effortful and intelligent.
Within strategic retrieval processes, two further compo-
nents are hypothesised. The rst involves organising
a memory search that uses whatever knowledge is avail-
able, whether semantic or episodic. Once knowledge is re-
covered, a second strategic process is involved in
monitoring the output of the memory search and checking
whether it is consistent with other information in seman-
tic and episodic memory. When strategic retrieval is dis-
rupted, following damage in the ventromedial and
orbitofrontal cortex (Gilboa et al., 2006; Moscovitch, 1982),
both semantic and episodic confabulations should occur
if the demands on the strategic retrieval of episodic versus
semantic information are matched. Burgess and Shallices
(1996) model is consistent with Moscovitchs proposal that
confabulation is associated with decits in strategic re-
trieval, which implicate defective search and monitoring.
These hypotheses that emphasise the role of a frontal/
executive dysfunction and of the disruption of monitoring
processes in the origin of confabulation are challenged by
several types of observations: (a) two patients have been
described with a confabulatory syndrome where there were
spared executive functions and no frontal pathology (Dalla
Barba, 1993a; Dalla Barba et al., 1990); (b) the confabulating pa-
tient described by Delbecq-Derouesne et al. (1990) had a docu-
mented frontal lobe lesion but performed normally on tasks
supposed to be sensitive to frontal lesion; (c) the confabulating
patient described by Dalla Barba et al. (1997a, 1997b) showed
impaired executive functions without any evidence of struc-
tural or functional damage to the frontal lobe; (d) in some
patients confabulation affects the performance on episodic
memory tasks but not on semantic memory tasks (Dalla
Barba, 1993a; Dalla Barba et al., 1997a, 1997b, 1990), even
when demands on the strategic retrieval of episodic versus
semantic information are matched (Dalla Barba et al., 1997a,
1997b, 1999). It is also well known that only a relatively small
proportion of patients with ventromedial and orbitofrontal
lesions confabulate. Retrieval models of confabulation pro-
pose that it is a very specic set of cognitive functions that
are not measured by traditional frontal/executive tests that
is disrupted in confabulation. However, these models dont
predict what would be a good measure of the set of frontal/ex-
ecutive functions supposed to be involved in strategic re-
trieval and monitoring.
(3) The third approach concerning the mechanisms involved
in confabulation predicts that confabulations are the
c o r t e x 4 5 ( 2 0 0 9 ) 5 6 6 5 7 4 567
result of a disrupted personal temporality (Dalla Barba,
1997a, 1997b, 1999, 2000, 2001, 2002). Based on clinical
and experimental, evidence, this hypothesis assumes
that in confabulating patients the knowledge of time is
preserved, in that they are aware of a past, present and fu-
ture. However, they are unable to distinguish between per-
sonal habits, repeated events and routines and true
episodic memories or true personal plans. In spite of
some important differences, this theory is not incompati-
ble with the theory proposed by Schnider and co-workers,
which posits that confabulators fail to suppress inappro-
priate memory traces, resulting in all active traces appear-
ing to relate to the present (but see Dalla Barba, 2000, 2001,
2002 for a critique of the concept of memory trace).
This paper describes the study of a patient, LM, with a Kor-
sakoffs syndrome who developed an extraordinary form of
confabulatory syndrome, which we refer to as confabulatory
hypermnesia, i.e., the tendency to provide a confabulatory an-
swer to questions for whichthe predicted and most frequently
observed response in normal subjects and in confabulators is
I dont know.
The aims of this study are to elucidate as fully as possible
the nature of the impairment of patient LM by trying to nd
answers to the following questions:
Can LMs confabulatory hypermnesia be traced back to
a specic cognitive decit?
What distinguish LM from other confabulating patients
who do not confabulate answering questions for which the
expected answer is I dont know?
2. Case report
2.1. History
LM, retired warehouseman of an airways company, is a 68-
year-old right-handed man with 8 years of education. He is
married and has ve children.
He has a history of heavy drinking which lasted more than
30 years and was stopped approximately 3 months before the
present evaluation.
Following the abrupt onset of temporo-spatial disorienta-
tion and memory loss, LM was referred to the Memory Clinic
of the Charles Foix hospital in Ivry-sur-Seine for neuropsycho-
logical evaluation. The physical, neurological and psychiatric
examination was normal. An MRI brain-scan showed a mild
diffuse cortical atrophy without evidence of any focal lesion.
The patients history of heavy drinking and the clinical picture
suggested a diagnosis of Korsakoffs syndrome.
2.2. Neuropsychological examination
The patients collaboration during the testing sessions was
good. On direct questioning he was disoriented and confabu-
lated in time and space. He scored 15/30 on the Mini Mental
State Examination (Folstein et al., 1975). On tests of oral ex-
pression, understanding of oral language, neglect, ideational,
ideomotor and constructional apraxia, and short-term mem-
ory the patient was judged to be normal. LM showed to be
impaired on some tests considered sensitive to executive dys-
function. He performed pathologically on the Trail Making
Test (Davies, 1968) [Trail Marking Test part A 120 sec, 25/25,
<10 percentile; Trail Marking Test part B 4 min, 17/25, <10 per-
centile] and on the Digit Symbol subtest of the Wechsler Adult
Intelligence Scale Revised (Wechsler, 1981) raw score 17. His
performance on Letter Fluency and on Category Fluency was
also defective, producing an average of 6 and 12 words, re-
spectively. In contrast, he performed normally on both direct
(7) and reverse (4) digit span.
2.2.1. Clinical tests of memory
On the Free and Cued Selective Reminding Test (Buschke,
1984; see Table 1) LM showed a massive decit in free recall,
whereas cued recall and recognition were relatively better
preserved. On the Logical Memory subtest of the Wechsler
Memory Scale III (Wechsler, 1997; see Table 1), he showed
very poor free recall and impaired recognition. In addition,
his performance on this test was contaminated by many con-
fabulations. The patient also had great difculty in retrieving
autobiographical episodes and showed very poor knowledge
of famous facts and famous people.
2.3. Examination of confabulation
LMproduced most of his confabulations in informal conversa-
tion, when answering specic questions and spontaneously,
mainly referring to autobiographical episodic memory. Con-
fabulation was always plausible and semantically appropriate
so that an hypothetical observer not familiar with LMs his-
tory, background and present situation could hardly tell
whether LMs reports were confabulatory or not. For example
he would claimthat on the previous day he had a soup for din-
ner and then he watched the news on television. In such
a case it would be impossible to tell whether this was a confab-
ulatory report unless one knew that he didnt have soup for
dinner and that he didnt watch television because there
was no television in the ward where he was hospitalised.
2.4. Confabulation Battery
To further examine LMs confabulation, we submitted him to
a new version of the Confabulation Battery (Dalla Barba,
1993a). The Confabulation Battery involves the retrieval of
various kinds of information and consists of 165 questions,
15 for each of the following domains:
Table 1 Memory test results
LM NCs
means
(standard
deviation, SD)
Free and Cued
Selective
Reminding Test
Immediate cued recall 13/16 15.2 (1.4)
Total free recall 6/48 28.9 (6.8)
Recognition 13/16 15.7 (.5)
Logical Memory Immediate recall 11/75
Delayed recall 6/50
Recognition 16/30
c o r t e x 4 5 ( 2 0 0 9 ) 5 6 6 5 7 4 568
(1) Personal Semantic Memory (age, date of birth, current ad-
dress, number of children, etc.).
(2) Episodic Memory. Episodic, autobiographical questions.
(3) Orientation in Time and Place.
(4) Linguistic Semantic Memory. Items 1630 of the Wechsler
Adult Intelligence Scale vocabulary subtest were selected
for a word denition task.
(5) Recent General Semantic Memory. Knowledge of facts
and people, which have been repeatedly reported in the
news during the last 10 years. For example, Who is Ben
Laden?
(6) Contemporary General Semantic Memory. Knowledge of
famous facts and famous people from 1940 to 1990. For
example, What happened in Paris in May 1968?
(7) Historical General Semantic Memory. Knowledge of fa-
mous facts and famous people before 1900. For example,
What happened in 1789?
(8) Semantic Plans. Knowledge of issues and events likely to
happen in the next 10 years. For example, Can you tell
me what you think will be the most important medical
breakthrough likely to take place in the next 10 years?
(9) Episodic Plans. Personal events likely to happen in the fu-
ture. For example, What are you going to do tomorrow?
(10) I dont know Semantic. These were questions tapping
semantic knowledge and constructed so as to receive
the response I dont know from normal subjects. For
example, What did Marilyn Monroes father do?
(11) I dont know Episodic. These questions tapped episodic
memory and were constructed so as to receive the re-
sponse I dont know by normal subjects. For example,
Do you remember what you did on March 13, 1985?
2.4.1. Procedure
Questions from the 11 domains were randomised and pre-
sented to the patient and to 10 age and education matched
normal controls (NCs). Responses were scored as correct,
wrong, I dont know, and confabulation. For episodic
memory, responses were scored correct when they
matched information obtained from the patients relatives.
Correct responses were self-evident for semantic memory
questions. For I dont know questions, both Semantic and
Episodic, an I dont know response was scored as correct.
Since there is no sufciently acceptable external criterion ca-
pable of dening confabulation, for its detection an arbitrary
decision necessarily had to be made. In order to distinguish
between a wrong response and a confabulation a clear-cut de-
cision was adopted only for answers to questions probing ori-
entation in time. In this case the most strict criterion was
chosen: answers to questions regarding the current year, sea-
son, month, day of the month, day of the week and hour of the
day were judged to be confabulations only if erring for more
than 5 years, 1 season, 2 months, 10 days, 3 days or 4 h, respec-
tively. Answers to the other questions of the Confabulation
Battery were independently rated as correct, wrong, and
confabulation by four different raters, and interrater reliabil-
ity was 100%. It must be emphasised that the decision as to
whether an answer was wrong or confabulatory was never
puzzling, although it may have been made on an arbitrary or
subjective basis. As far as questions concerning personal
and semantic plans are concerned, it might be argued that
any possible answer is a confabulation, since, by denition,
the future is only probable and there is in principle no cor-
rect answer to questions about the future. Yet, answers con-
cerning the future can be denitely confabulatory when they
show a marked discrepancy or a real contradiction with
what a predicted future event might be, in view of the present
situation. For example, although he was hospitalised and de-
spite the fact that there wasnt any television in the ward, to
the question What are you going to do to night?, LM an-
swered Ill have dinner with my wife and then watch the
news at the television.
3. Results
LMs and NCs performance on the Confabulation Battery is
reported in Table 2.
3.1. Correct responses
LM produced signicantly fewer correct responses than NC to
questions probing Personal Semantic Memory, Episodic Mem-
ory, Orientation in Time and Place, Recent Seneral Semantic
Memory [t(9) 15, p <.001; estimated percentage of normal
subjects falling below LMs scores .0%], Contemporary
Table 2 LMs and NCs performance (%) on the confabulation battery
Correct Confabulation Wrong IDK Total recall
LM NC LM NC LM NC LM NC LM NC
Personal Semantic Memory 73 100 20 0 0 0 7 0 93 100
Episodic Memory 33 100 53 0 7 0 7 0 93 100
Orientation in Time and Place 53 100 13 0 33 0 0 0 100 100
Linguistic Semantic Memory 66 81.4 (15.7) 0 0 33 9.9 (9.9) 0 8.7 (7.3) 100 91.3 (5.6)
Recent General Semantic Memory 33 96.5 (3.7) 40 .7 (2.2) 27 .7 (2.2) 0 2.1 (3.4) 100 97.9 (3.3)
Contemporary General Semantic Memory 73 93.2 (7.7) 7 0 13 1.3 (4.1) 7 5.5 (6.2) 93 94.5 (6.1)
Historical General Semantic Memory 53 73.2 (28.9) 13 .7 (2.2) 20 13.3 (27) 13 13.2 (10.3) 86 86.8 (4.5)
Semantic Plans 80 95.9 (6.5) 7 0 0 1.4 (3) 13 2 (6.3) 87 97.3 (6.4)
Episodic Plans 33 98 (6.3) 33 0 27 0 7 2 (6.3) 93 98 (6.3)
I dont know Semantic 66 99.3 (2.2) 33 .7 (2.2) 0 0 66 99.3 (2.2) 33 0
I dont know Episodic 7 100 93 0 0 0 7 100 93 0
c o r t e x 4 5 ( 2 0 0 9 ) 5 6 6 5 7 4 569
General Semantic Memory [t(9) 2.3, p <.05; estimated per-
centage of normal subjects falling below LMs scores 2%],
Semantic Plans [t(9) 2.5, p <.05; estimated percentage of nor-
mal subjects falling below LMs scores 1.5%], Episodic Plans
[t(9) 10.3, p <.001; estimated percentage of normal subjects
falling belowLMs scores .0%], and to I dont know Seman-
tic [t(9) 14.4, p <.001; estimated percentage of normal sub-
jects falling below LMs scores .0%], and I dont know
Episodic questions (Crawford and Garrthwaite, 2002; Crawford
and Howell, 1998). LM produced fewer correct responses than
NC to questions probing Linguistic Semantic Memory and His-
torical Semantic Memory, but this effect was not signicant.
The estimated percentage of normal subjects falling below
LMs scores was 19.7% for Linguistic Semantic Memory and
27.3% for Historical Semantic Memory.
LM produced signicantly fewer correct responses, i.e., I
dont know, to I dont know Episodic questions than to
any other type of question (all p <.0001; estimated percentage
of normal subjects exhibiting a difference more extreme than
LM.0, Crawford and Garrthwaite, 2002; Crawford and
Howell, 1998). Episodic Memory, Recent General Semantic
Memory and Episodic Plans were also signicantly affected
compared to Orientation in Time and Place and Historical
General Semantic Memory (all p <.0001; estimated percentage
of normal subjects exhibiting a difference more extreme than
LM.0), which in turn were impaired compared to Personal
Semantic Memory, Linguistic Semantic Memory, Contempo-
rary General Semantic Memory, Semantic Plans, and I dont
know Semantic questions (all p <.0001; estimated percent-
age of normal subjects exhibiting a difference more extreme
than LM.0, Crawford and Garrthwaite, 2002; Crawford and
Howell, 1998).
NC produced signicantly fewer correct responses to ques-
tions probing Linguistic Semantic Memory and Historical Gen-
eral Semantic Memory than to any other type of question (all
p <.05) except for questions probing Contemporary General
Semantic Memory.
To summarise, compared to NC, LM performed poorly
across the board on the Confabulation Battery, with his per-
formance being relatively better preserved in Personal Seman-
tic Memory, Linguistic Semantic Memory, Contemporary
Semantic Memory and Semantic Plans. He was particularly
impaired in answering I dont know Episodic questions
where he correctly answered I dont know only 1 out of 15
times.
3.2. Confabulations
NC produced virtually no confabulations. In fact, only one
subject produced one confabulation in Recent General Seman-
tic Memory and one in Historical General Semantic Memory.
A Chi-square test revealed a global signicant effect of the
type of questions on LMs confabulations [c
2
(10) 246,9,
p <.001]. LM did not confabulate in response to Linguistic
Semantic Memory questions. He produced signicantly more
confabulations to I dont know Episodic questions than to
any other type of question (Chi-square test, all p <.001). He
also produced signicantly more confabulations to Episodic
Memory, Recent General Semantic Memory, Episodic Plans
and I dont know Semantic questions than to questions
probing Personal Semantic Memory, Orientation in Time and
Place, Contemporary General Semantic Memory, Historical
General Semantic Memory and Semantic Plans (Chi-square
test, all p <.05).
To summarise, LM produced an impressive percentage of
confabulations in response to I dont know Episodic ques-
tions. He also produced more confabulations to questions in-
volving episodic recall and episodic plans compared to
questions involving recall from semantic memory.
3.3. Examples of LMs confabulation for each domain
3.3.1. Personal Semantic Memory
Q. What is your brothers job?
A. Re my is an artist. He works in variety shows. LM con-
fuses his brother with his son (Patrik) who actually works as
an artist in variety shows.
3.3.2. Episodic Memory
Q. Do you remember the last time a doctor examined you?
A. Well, he examined me without really examining me,
because my doctor lives near me, so I see him quite often.
The last time.not so long ago in any case.
3.3.3. Orientation in Time and Place
Q. What month is it?
A. End of December. In fact it was the end of April.
3.3.4. Recent General Semantic Memory
Q. What does the Mir station remind you of?
A. Its a metros station in the 15th arrondissement in
Paris.
3.3.5. Contemporary General Semantic Memory
Q. What happened in May 1968?
A. It was the beginning of the workmens revolution; the
government opposed Africa.
3.3.6. Historical General Semantic Memory
Q. What does the Dreyfus affaire remind you of?
A. Its the story of a poisoning and of a murder.
3.3.7. Semantic Plans
Q. What do you think will be the most important advance-
ments in the exploration of space in the next 10 years?
A. We will land on the Moon and see if somebody lived
there, if it is habitable.
3.3.8. Episodic Plans
Q. What are you going to do the day after tomorrow?
A. Ill go shopping, as usual, and then watch television.
3.3.9. I dont know Semantic
Q. Who won the Palme dOr at the Canne Festival in 1980?
A. Fernandel.
3.3.10. I dont know Episodic
Q. Do you remember what you had for dinner on Tuesday
two weeks ago?
R. Steak with French fries.
c o r t e x 4 5 ( 2 0 0 9 ) 5 6 6 5 7 4 570
Q. Do you remember what you were wearing on the rst
day of summer in 1979?
R. A short and a T-shirt.
Q. Do you remember what you did on March 13, 1985?
A. We spent the day at the Senart Forest (a place where
LM used to go often with his family).
Q. Do you remember what you did for your 27th
birthday?
R. Wait a minute.I came back from the army when I was
23, so the 27th birthday.I think.Yes! I had dinner at my fu-
ture wifes place.
4. Discussion
The patient we have described suffers from a relatively pure
amnesic syndrome, in the sense that the memory impairment
is not associated with other cognitive impairments likely to
interfere with memory performance. Although he performs
poorly on the Mini Mental State Examination, mainly due to
his disorientation, LM doesnt show a general deterioration
of intellectual and attentional abilities, and he scores nor-
mally on tests of short-term memory. His amnesia involves
memory for both personal and public events, as well as learn-
ing of new information as shown in his poor performance on
the Free and Cued Selective Reminding Test and onthe Logical
Memory subtest of the Wechsler Memory Scale. LM confabu-
lates both spontaneously and on direct questioning. His con-
fabulations are always plausible and never show the bizarre,
fantastic (Baddeley and Wilson, 1986) or semantically
anomalous (Dalla Barba, 1993b) content described in some
confabulating patients.
On the Confabulation Battery, LM produced fewer correct
responses than NCs to all types of questions, although he per-
formed relatively better in Personal Semantic Memory, Lin-
guistic Semantic Memory, Contemporary Semantic Memory
and Semantic Plans than in other domains. This result is sim-
ilar to the pattern of performance of other amnesic patients
on the Confabulation Battery (Dalla Barba, 1993b; Dalla Barba
et al., 1997a, 1997b; Fotopoulou et al., 2004). In LM, like in other
patients, semantic memory and semantic plans are better pre-
served than episodic memory and episodic plans. However,
unlike other confabulators, LM produced only one correct I
dont know, response to I dont know Episodic questions.
To this type of question, 14 out of 15 times he produced
a confabulation.
LM confabulated in response to all types of questions, with
the exception of Linguistic Semantic Memory. However, he
produced signicantly more confabulations in Episodic Mem-
ory, Recent Semantic Memory and Episodic Plans compared to
the other types of questions. Like in other confabulators (Dalla
Barba, 1993a, 1993b; Dalla Barba et al., 1997a, 1997b), LM
showed confabulations with a pattern of selectivity, affecting
the episodic domain more than the semantic domain. He also
confabulated in Recent General Semantic Memory, whichmay
represent a grey area between semantic and episodic memory
(Shallice, 1988), and in I dont know Semantic questions. But
the most striking aspect of LMs confabulations is his ten-
dency to confabulate 3 times as much with episodic memory
questions that are usually answered I dont know by normal
subjects and which produce confabulations, but less orid
than LMs, in other confabulating patients (Dalla Barba,
1993a; Dalla Barba et al., 1998).
As far as total recall is concerned, i.e., the sumof correct re-
sponses, errors and confabulations, LM and NC showed a sim-
ilar pattern of performance. They both showed the tendency
to provide an answer, correct or not, rather than an I dont
know response to all types of questions with the exception
of I dont know questions where LM was always productive,
whereas NC always correctly answered I dont know.
From LMs performance on the Confabulation Battery
emerges a previously unreported phenomenon, which we re-
fer to as confabulatory hypermnesia, i.e., the tendency to provide
a confabulatory answer to questions for which the predicted
and most frequently observed response in normal subjects
and in confabulators is I dont know. It might be argued
that LMs specic pattern of performance reects a methodo-
logical bias, in that I dont know Semantic questions may be
more difcult or less likely to elicit confabulations than I
dont know Episodic questions. Confabulations in response
to I dont know questions, both semantic and episodic,
have been reported in some confabulating patients (Dalla
Barba, 1993b; Dalla Barba et al., 1998; Fotopoulou et al., 2004,
2007; Kopelman et al., 1997), whereas other patients do not
confabulate at all to these questions (Box et al., 1999; Dalla
Barba, 1993a; Dalla Barba et al., 1997a, 1997b). However,
none of the reported cases confabulated as much as LM to I
dont know Episodic questions, and in none of these cases
was there such a clear dissociation between confabulations
to I dont know Episodic and I dont know Semantic ques-
tions. Actually some patients confabulated slightly more to I
dont know Semantic than to I dont know Episodic ques-
tions (Fotopoulou et al., 2007). So, although some patients con-
fabulate to I dont know Episodic questions, none of them
show the specic pattern of performance described in LM.
This pattern of performance, that we call confabulatory hy-
permnesia, doesnt seem to simply reect differences in task
difculty. This leads us to address the problem of the origin
of confabulatory hypermnesia. By asking the following ques-
tions: Why does LM confabulate when answering episodic
memory questions for which the expected answer is I dont
know? and What is the difference between LM and confab-
ulating patients who do not confabulate so oridly with I
dont know questions?
It has been argued that confabulation reects poor source
monitoring. Source monitoring refers to the set of processes
involved in making a decision about the origin of different in-
formation (see Metcalf et al., 2007, for a review). A related con-
cept is reality monitoring, which refers to the ability to
discriminate between internally and externally generated in-
formation (e.g., imagination vs veridical perception). Accord-
ing to this view, confabulation is the result of mistaking
imagined events for truly experienced events (Johnson,
1991). This kind of explanation, however, does not account
for the fact that confabulation, as in the present case and in
other studies (Dalla Barba, 1993a, 1993b; Dalla Barba et al.,
1990, 1997a, 1997b; Delbecq-Derouesne et al., 1990), is absent
or signicantly less frequent in tasks that require retrieval of
non-episodic information. In fact, if a disruption of reality
monitoring were responsible for confabulation, confusion
c o r t e x 4 5 ( 2 0 0 9 ) 5 6 6 5 7 4 571
between truly experienced and imagined events and informa-
tion should not be limited to episodic retrieval but should ex-
tend across all memory domains. Moreover, if we consider the
examples of LMs confabulations in I dont know Episodic
questions, there is no evidence that he is mistaking imagined
events for really experienced events. Most of his confabula-
tions to I dont know Episodic questions reect the retrieval
of personal habits or semantic information rather than re-
trieval of imagined events: the Senart Forest is a place where
he used to go often with his family, steak and French fries
are a very common meal in France, in summer many people
wear shorts and T-shirts.
The retrieval of personal habits and semantic information
in LMs confabulations could be traced back to a decit of stra-
tegic retrieval (Moscovitch, 1989, 1995; Moscovitch and Melo,
1997). In the presence of such a decit, only associative re-
trieval is possible and this might be reected by the type of
confabulations produced by LM. A decit of strategic retrieval
applies equally across all domains of memory. In other words,
confabulation should equally affect episodic and semantic
memory if both episodic and semantic memory tasks require
the involvement of strategic retrieval processes. However, this
hypothesis has difculties in accounting for cases in which
confabulations are conned to the domain of episodic mem-
ory (Dalla Barba, 1993b; Dalla Barba et al., 1990, 1997a,
1997b), also when both episodic and semantic memory tasks
require the involvement of strategic retrieval processes (Dalla
Barba et al., 1997a, 1997b, 1999). As far as LMs case is con-
cerned, a strategic retrieval decit account predicts that if I
dont know Semantic and I dont know Episodic questions
make equal demands on strategic retrieval processes then
confabulations should occur both for I dont know Semantic
and I dont know Episodic questions, which is not what was
actually observed. As discussed above, there is no reason to
consider one type of I dont know question more difcult
or more likely to elicit confabulation than the other type. Since
there is no specic available measure of the amount of strate-
gic retrieval involved in a task, it is impossible to know
whether I dont know Episodic questions are more demand-
ing in terms of strategic retrieval processes than I dont
know Semantic questions. However, there is clear evidence
that LM was engaged in an active search when answering
both I dont know Episodic questions (see his answer to
the question Do you remember what you did for your 27th
birthday?) and I dont know Semantic questions, as clearly
emerges from the following example:
Q. What is the rst name of Jimmy Carters wife?
R. Her rst name?.(search); no; no though I should know
it because she is well known;
Tothis typeof questions, normal subjects dont evenattempt
to searchand immediately produce an I dont know response,
even when they are explicitly asked to make a plausible guess.
We have argued (Dalla Barba, 2000, 2001; Dalla Barba et al.,
1997a, 1997b, 1999) that confabulation can be interpreted
within the framework of the MCTT (Dalla Barba, 2002), The
MCTT, that is schematised in the model presented in Fig. 1,
holds that (1) consciousness is not a unitary and generic di-
mension that passively receives what comes from outside,
but is the set of distinct and original modes for addressing
the object. Among the modes of consciousness there is
knowing consciousness KC and temporal consciousness TC. (2)
KC describes the mode of addressing the object in order to
know it. (3) TC describes the mode of consciousness that tem-
poralises its object according to the subordinate structures of
temporality, the past, the present and the future. (4) Con-
sciousness must necessarily always be consciousness of
something. Every type of consciousness is consciousness
of anobject. The object of consciousness represents a determi-
nation and a indetermination, a Uniqueness and a Multiplicity.
Lets consider this point closer. This pen on the desk is both
a pen and the pen. In the rst case it is an undetermined pen,
something that belongs to the category of pens, an object
that I recognize and use appropriately because I recognize it.
On the contrary, in the second case the pen is a determined
object, is exactly this pen in front of me, the pen I bought yes-
terday and that I will be using tomorrow. So, the pen reveals
a uniqueness and a multiplicity, the multiplicity of being
a pen and not a different object of the object in front of me,
the uniqueness of being precisely this penand not another pen.
Within the framework of the MCTT, we have proposed that
confabulation in episodic memory and in personal future plan-
ning tasks, as in LMs case and in other patients (Dalla Barba
et al., 1997a, 1997b, 1999), canbeinterpretedastheresult of acon-
dition in which TC remains but, in order to remember the per-
sonal past and plan the personal future, addresses Multiplicity
as Uniqueness. This condition is schematised in Fig. 2. In other
Fig. 1 A hypothesis concerning the relation between
memory, consciousness and temporality. NS [nervous
system; X and Y[patterns of organisation and activation
of the nervous system; TC[temporal consciousness;
KC[knowing consciousness; U[uniqueness; and
M[multiplicity.
Fig. 2 Confabulation: TC addresses M as U.
c o r t e x 4 5 ( 2 0 0 9 ) 5 6 6 5 7 4 572
terms, thisisaconditioninwhichTCisstill workingbut isnolon-
ger able to accomplish its usual task, i.e., to distinguish between
Multiplicity andUniqueness inorder toset upa personal tempo-
ral workspace. What TC does instead in this condition is to ad-
dress Multiplicity of the object as Uniqueness, so that the
result is that not only unique personal episodes, but also habits
or personal semantic information are considered in a personal
temporal framework. When asked what they did the previous
dayor what theyaregoingtodothefollowingday, confabulating
patients of this typetypicallyanswer reportingas memories and
plans what they usually do intheir daily life. Although admitted
to the hospital, they will say, for example, that the previous day
theywent out shoppingandthat thefollowingdaywill bevisiting
somefriends, actsthat presumablywerepart of their routinelife.
This kind of interpretation, however, still doesnt explain
why, unlike other confabulators, LM confabulates to I dont
know Episodic questions. The analysis of total recall shows
that, with the exception of I dont know questions, LM and
NC have the same performance, a mean of 94% and 95% of re-
sponses, respectively. This indicates that both LMand NC have
answers to almost all the questions, although in LMthe answer
is often confabulatory, especially inthe episodic domain. It also
indicates that, according to the MCTT, in both NC and LM, KC
and TC are working, although the latter is working in an abnor-
mal way in LM. However, while in NC, KC and TC respect the
boundary of I dont know questions, in LM this boundary is
overstepped. What is the nature of this boundary? Why NC
and most confabulators promptly respond I dont know to
I dont know questions, whereas LM confabulates?
One possibility is that LMmay have always beena cognitive
outlier, in the sense that, even premorbidly, he may have
used unusual cognitive strategies and that his confabulatory
hypermnesia is a continuation of the use of such strategies.
However, no evidence in this sense emerged fromquestioning
the patients wife. So, although this possibility cannot be ex-
cluded, there is no specic evidence favouring a premorbid
personality account of LMs disorder.
LMs confabulatory hypermnesia doesnt seem to be corre-
lated to any specic pattern of brain damage. The MRI brain-
scan showed a mild diffuse cortical atrophy, which is an
unremarkable nding. LMs cognitive prole was also unre-
markable. In particular, he didnt show an especially severe
impairment of executive functions, compared to other con-
fabulators. Accordingly, LMs confabulatory hypermnesia
cannot be traced back to any specic pattern of brain damage
or of cognitive impairment. Froma merely descriptive point of
view, and making reference to the MCTT, LM, compared to
normal subjects and to other confabulators, shows an ex-
panded TC, which overows the limits of time (Do you re-
member what you did on March 13, 1985?) and of details
(Do you remember what you were wearing on the rst day
of summer in 1979?) that are usually respected in normal
subjects and in confabulating patients.
5. Conclusions
We have described a previously unreported phenomenon,
confabulatory hypermnesia, arguing that it reects an expan-
sion of TC beyond its normal limits. Although we believe that
our interpretation best accounts for the LMs confabulatory
hypermnesia, other interpretations of this phenomenon can-
not be completely dismissed since a clear double dissociation
between I dont know Semantic and Episodic questions has
not been yet reported. Be that as it may, this leaves us with
a number of open questions: What are the cognitive and neu-
ral correlates of confabulatory hypermnesia? How are the
boundaries of TC normally determined? What is the cause of
their breakdown? What would be the critical test or set of ev-
idence that would conrm or disprove this hypothesis? An-
swers to these questions would shed light on confabulatory
hypermnesia and on the organisation of TC in general. In
the absence of further experimental evidence these remain
unanswered, indeed I dont know questions.
r e f e r e n c e s
Alexander MP and Freedman M. Amnesia after anterior
communicating artery aneurysm rupture. Neurology, 34:
752757, 1984.
Baddeley A and Wilson B. Amnesia, autobiographical memory
and confabulation. In Rubin DC (Ed), Autobiographical Memory.
Cambridge, UK: Cambridge University Press, 1986: 225252.
Benson DF, Djenderedjian A, Miller BL, Pachana NA, Chang L,
Itty L, et al. Neural basis of confabulation. Neurology, 46:
12391243, 1996.
BerlyneN. Confabulation. BritishJournal of Psychiatry, 120: 3139, 1972.
Bonhoeffer K. Der korsakowsche symptomenkoplex in seinen
beziehungen zu den verschiedenen krankheitsformen.
Allgemeine Zeitung Psychiatrie, 61: 744752, 1904.
Box O, Laing H, and Kopelman MD. The evolution of spontaneous
confabulation, delusional misidentication and a related
delusion in a case of severe head injury. Neurocase, 5: 251262,
1999.
Burgess PW and Shallice T. Confabulation and the control of
recollection. Memory, 4: 359411, 1996.
Buschke H. Cued recall in amnesia. Journal of Clinical
Neuropsychology, 6: 433440, 1984.
Cermak LS, Uhly B, and Reale L. Encoding specicity inthe alcoholic
Korsakoff patient. Brain and Cognition, 11: 119127, 1980.
Crawford J and Garrthwaite P. Investigation of the single case in
neuropsychology: condence limits on the abnormality of test
scores and test scores differences. Neuropsychologia, 40:
11961208, 2002.
Crawford J and Howell D. Comparing an individuals test score
against norms derived from small samples. The Clinical
Neuropsychologist, 12: 482486, 1998.
Dalla Barba G. Confabulation: knowledge and recollective
experience. Cognitive Neuropsychology, 10: 120, 1993a.
Dalla Barba G. Different patterns of confabulation. Cortex, 29:
567581, 1993b.
Dalla Barba G. Memory, consciousness and temporality: what is
retrieved and who exactly is controlling the retrieval? In
Tulving E (Ed), Memory and Consciousness and the Brain: The
Tallin Conference. Philadelphia: Psychology Press, 2000: 138155.
Dalla Barba G. Beyond the memory trace paradox and the fallacy
of the homunculus: a hypothesis concerning the relation
between memory, consciousness and temporality. Journal of
Consciousness Studies, 8: 5178, 2001.
Dalla Barba G. Memory, Consciousness and Temporality. Boston:
Kluwer Academic Publishers, 2002.
Dalla Barba G, Boisse M-F, Bartolomeo P, and Bachoud-Le vi A-C.
Confabulation following rupture of posterior communicating
artery. Cortex, 33: 563570, 1997a.
c o r t e x 4 5 ( 2 0 0 9 ) 5 6 6 5 7 4 573
Dalla Barba G, Cappelletti YJ, Signorini M, and Denes G.
Confabulation: remembering another past, planning
another future. Neurocase, 3: 425436, 1997b.
Dalla Barba G, Cipolotti L, and Denes G. Autobiographical memory
loss and confabulation in Korsakoffs syndrome: a case report.
Cortex, 26: 525534, 1990.
Dalla Barba G, Mantovan MC, Cappelletti YJ, and Denes G.
Temporal gradient in confabulation. Cortex, 34: 417426,
1998.
Dalla Barba G, Mantovan MC, Traykov L, Rieu D, Laurent A, and
Devouche E. The functional locus of intrusions: encoding or
retrieval? Journal of Clinical and Experimental Neuropsychology,
24: 633641, 2002.
Dalla Barba G, Nedjam Z, and Dubois B. Confabulation, executive
functions and source memory in Alzheimers disease.
Cognitive Neuropsychology, 16: 385398, 1999.
Davies A. The inuence of age on trial making test performance.
Journal of Clinical Psychology, 24: 9698, 1968.
Delbecq-Derouesne J, Beauvois MF, and Shallice T. Preserved
recall versus impaired recognition. Brain, 113: 10451074, 1990.
Fischer RS, Alexander MP, DEsposito M, and Otto R.
Neuropsychological and neuroanatomical correlates of
confabulation. Journal of Clinical and Experimental
Neuropsychology, 17: 2028, 1995.
Folstein MF, Folstein SE, and McHugh PR. Mini mental state:
a practical method for grading the cognitive state of patients
for the clinician. Journal of Psychiatric Research, 12: 189198,
1975.
Fotopoulou A, Conway M, and Solms M. Confabulation: motivated
reality monitoring. Neuropsychologia, 45: 21802190, 2007.
Fotopoulou A, Solms M, and Turbull O. Wishful reality distortions
in confabulation: a case report. Neuropsychologia, 42: 727744,
2004.
Gilboa A, Alain C, Stuss DT, Melo B, and Moscovitch M.
Mechanism of spontaneous confabulations: a strategic
retrieval account. Brain, 129: 13191414, 2006.
Irle E, Wowra B, Kunert HJ, and Kunze S. Memory disturbances
following anterior communicating artery rupture. Annals of
Neurology, 31: 473480, 1992.
Johnson MK. Reality monitoring: evidence from confabulation in
organic brain disease patients. In Prigatano GP, and
Schacter DL (Eds), Awareness of Decit After Brain Injury. New
YorkOxford: Oxford University Press, 1991: 176197.
Johnson MK. Source monitoring and memory distortion.
Philosophical Transactions of The Royal Society of London B Series B:
Biological Science, 29: 17331745, 1997.
Johnson MK, OConnor M, and Cantor J. Confabulation, memory
decits, and frontal dysfunction. Brain and Cognition, 34:
189206, 1997.
Kapur N and Coughlan AK. Confabulation after frontal lobe
dysfunction. Journal of Neurology, Neurosurgery and Psychiatry,
43: 461463, 1980.
Kern R, Van Grop W, Cummings J, Brown W, and Osato S.
Confabulation in Alzheimers disease. Brain and Cognition, 19:
172182, 1992.
Kopelman MD. Two types of confabulation. Journal of Neurology,
Neurosurgery and Psychiatry, 50: 14821487, 1987.
Kopelman MD, Guinan EM, and Lewis PDR. Delusional memory,
confabulation, and frontal lobe dysfunction: a case study in de
Cle rambaults syndrome. Neurocase, 1: 7177, 1995.
Kopelman MD, Ng N, and Van Den Brouke O. Confabulation
extending across episodic, personal and general semantic
memory. Cognitive Neuropsychology, 14: 683712, 1997.
Korsakoff SS. Etude me dico-psychologique sur une forme des
maladies de la me moire. Revue Philosophique, 28: 501530, 1889.
Luria AR. The Neuropsychology of Memory. New York: John Wiley,
1976.
Mercer B, Wapner W, Gardner H, and Benson F. A study of
confabulation. Archives of Neurology, 34: 429433, 1977.
Metcalf K, Langdon R, and Coltheart M. Models of confabulation:
a critical review and a new framework. Cognitive
Neuropsychology, 24: 2347, 2007.
Moscovitch M. A neuropsychological approach to memory and
perception in normal and pathological aging. In Craik FIM and
Trehub S (Eds), Aging and Cognitive Processes. New York:
Plenum Press, 1982.
Moscovitch M. Confabulation and the frontal systems: Strategic
versus associative retrieval in neuropsychological theories of
memory. In Roedinger HL and Craik FI (Eds), Varieties of
Memory and Consciousness: Essay in Honor of Endel Tulving.
Hillsdale NJ: Lawrence Erlbaum., 1989: 133160.
Moscovitch M. Confabulation. In Schacter DL (Ed), Memory
Distortion. Cambridge: Harvard University Press, 1995: 226251.
Moscovitch M and Melo B. Strategic retrieval and the frontal
lobes: evidence from confabulation and amnesia.
Neuropsychologia, 35: 10171034, 1997.
Nedjam Z, Dalla Barba G, and Pillon B. Confabulation in a patient
with frontotemporal dementia and in a patient with
Alzheimers disease. Cortex, 36: 561577, 2000.
Nedjam Z, Devouche E, and Dalla Barba G. Confabulation, but not
executive dysfunction discriminate ad from frontotemporal
dementia. European Journal of Neurology, 11: 728733, 2004.
Papagno C and Muggia S. Confabulation: dissociation between
everyday life and neuropsychological performance. Neurocase,
2: 111118, 1996.
Sandson J, Albert ML, and Alexander M. Confabulation in aphasia.
Cortex, 22: 621626, 1986.
Schacter DL. Memory, amnesia and frontal lobe dysfunction.
Psychology, 17: 2136, 1987.
Schacter DL, Harbluk JL, and McLachlan DR. Retrieval without
recollection: an experimental analysis of source amnesia.
Journal of Verbal Learning and Verbal Behavior, 23: 593611, 1984.
Schnider A, Gutbrod K, and Schroth G. Memory without context:
amnesia with confabulations after infarction of the right
capsular genu. Journal of Neurology, Neurosurgery and Psychiatry,
61: 186193, 1996a.
Schnider A, von Da niken C, and Gutbrod K. The mechanisms of
spontaneous and provoked confabulations. Brain, 119:
13651375, 1996b.
Shallice T. From Neuropsychology to Mental Structure. Cambridge:
Cambridge University Press, 1988.
Stuss DT, Alexander MP, Lieberman A, and Levine H. An
extraordinary form of confabulation. Neurology, 28: 11661172,
1978.
Talland GA. Confabulation in the WernickeKorsakoff syndrome.
Journal of Nervous and Mental Disease, 132: 361381, 1961.
Wechsler D. Wechsler Adult Intelligence Scale Revised Manual. New
York: Psychological Corporation, 1981.
Wechsler D. Wechsler Memory Scale III. New York: Psychological
Corporation, 1997.
Weinstein E and Lyerly O. Confabulation following brain injury.
Archives of General Psychiatry, 18: 348354, 1968.
Wyke M and Warrington E. An experimental analysis of
confabulation in a case of Korsakoffs syndrome using
a tachistoscopic method. Journal of Neurology, Neurosurgery and
Psychiatry, 23: 327333, 1960.
c o r t e x 4 5 ( 2 0 0 9 ) 5 6 6 5 7 4 574