NOTE

CONFABULATION IN APHASIA 1
Jennifer Sandson, Martin L. Albert and Michael P. Alexander
(Department of Neurology, Aphasia Research Center, Boston University
Medical School, and Boston Veterans Administration Medical Center, and
the Department of Neurology, Washington University Medical School)

Confabulation, the production of false information without intention to
deceive, has been said to result from impaired memory and decreased selfmonitoring ability (Stuss, Alexander, Lieberman, and Levine, 1978). Confabulations typically occur only in response to questions and generally consist of
information relevant to the patient's injury or personal history (Weinstein and
Lyerly, 1966). On the other hand, fantastic confabulations, assertions about
things which could not possibly have occurred, are less common and may be
spontaneous, i.e. may be produced without the prompting of a specific question.
The neuropsychologic deficit most often associated with either type of a confabulation is perseveration of response set within the area of impaired cognitive
function (Shapiro, Alexander, Gardner and Mercer, 1981). Weinstein (1981) has
noted the similarities in behavior in patients with semantic jargon aphasia and
patients with confabulation, particularly the uncritical acceptance of their own
deviant utterances.
We report the case of an aphasic patient with two distinct forms of confabulation: (1) typical confabulations about personal events from the recent past,
apparently triggered by gaps in memory; and (2) atypical (aphasic) confabulations in response to questions about word meanings or attributes, apparently
triggered by gaps in semantic representation.

CASE REPORT

The subject was a 66 year old right-handed retired factory worker who was
found in generalized status epilepticus on the floor. Skull X-ray revealed a left
temporal linear skull fracture; CT scan demonstrated a left temporal intracerebral hematoma (Figure 1). The electroencephalogram was abnormal due to
bilateral slow posterior rhythms. Past history was significant for alcohol abuse.
He was an amateur boxer in the remote past. There was no evidence of confabulation or of memory disturbance prior to the current illness.
1
This paper was presented in part at the 22nd annual meeting of the Acadamy of Aphasia, Los
Angeles, California, October 29, 1984.

Cortex (1986) 22, 621-626

622

Jennifer Sandson, Martin L. Albert and Michael P. Alexander

Fig. I - CT scan showing an area of attenuation in the left temporal lobe corresponding to
the middle temporal gyrus.

When examined six weeks post-onset, the subject was fully oriented to person,
place and time. Speech was well articulated and generally informative, although
there was marked word finding difficulty. Evaluation with the Boston Diagnostic
Aphasia Examination (Goodglass and Kaplan, 1983) revealed fluent output with
press of speech, circumlocutions, perseverations, and literal and verbal paraphasias. Auditory comprehension and repetition were only mildly impaired. Word
discrimination was average for aphasic patients and 19/20 pictures of common
animals were correctly identified on a multiple choice test. Further evaluation
revealed instances of failure to recognize common words as familiar, and strikingly bizarre responses to questions of common knowledge such as "Do dogs
fly?". No concern over or awareness of these problems was ever acknowledged,
and the patient actively denied language deficits. Language was essentially
unchanged when re-evaluated 6 months post-onset.

TABLE I

WAIS-R Scores

Prorated Verbal IQ = 79
Sub test
Information
Comprehension
Digit Span
Vocabulary

Peer Score
7
7
7
6

Performance IQ = 89
Sub test
Picture Completion
Block Design
Picture Arrangement
Object Assembly
Digit Symbol

Peer Score
7
9
9

10
6

623

Confabulation in aphasia

Neuropsychological evaluation was characterized by mildly impaired performance on verbal tests measuring vocabulary, general information, and social
judgement, and low average to average performance on visuospatial tasks. Ability
to abstract similarities between related concepts, however, was severely impaired
as was comprehension of common one word metaphors such as feeling "blue"
WAIS-R results are summarized in Table I.
Memory was moderately impaired, with zero out of five hidden objects
recalled after five minutes. Recall of the details of the Wechsler Memory Scale
(WMS) narrative passages was below normal with signs of confusion and confabulation. Immediate recall of the WMS figures was good, but was impaired by
intrusions after a delay. Delayed recall of the Rey Osterreith Complex Figure was
simplified and contained reduplication of details.
Experimental Test Procedures

Three sets of experimental tests were administered. These included:
Tests to elicit confabulation (adapted from Mercer, Wapner, Gardner and Benson,
1977)
The patient was asked to discuss events from his remote and immediate past
and to answer questions requiring the use of environmental cues. Questions
involving estimation (adapted from Shallice and Evans, 1978), difficult factual
material, and information about the attributes of common objects and animals
were also posed.
Tests to examine lexical access and semantic representation
First, the patient was asked to sort a deck of cards co~taining written stimuli
into two piles: words or non-words. Failure to differentiate real words from
nonsense words would indicate a disruption of or an inability to access the
lexicon. Second, a verbal description was requested for each of the same stimuli
presented orally. This task was primarily intended to encourage self monitoring
but also provided a means for exploring possible disruptions in the representation
of word meanings. Third, the patient was requested to sort basic level printed
words (e.g. pea) into five superordinate category piles (e.g. vegetable) to look for
disruptions in semantic representation.
1

Tests to elicit perseveration
Perseveration of response set was evaluated with a non-verbal test requiring
alternation of symbols and multiple reversals, and a visual-verbal test requiring
conceptual shifting. Other forms of perseveration were evaluated on a variety of
additional verbal and non-verbal tests described elsewhere (Sandson and Albert,
1984).
RESULTS

Confabulation

Questions about remote personal history (e.g. military service and job history), questions requiring estimation, questions dependent on environmental cues,
and difficult factual questions all failed to elicit confabulation. Responses con-

624

Jennifer Sam/son, Martin L. Albert and Michael P. Alexander

sisted of correct answers, reasonable guesses, and admissions of ignorance.
Typical confabulations were obtained in response to questions about how the
patient first became ill, the course of his hospital stay, and his role when he was
interviewed before a large group in a clinical case conference. For the last, he
claimed that he was at a debate about philosophy and religion.
Atypical confabulations, on the other hand, were obtained in response to
questions about the attributes of common objects, in particular of animals. He
confidently stated, for example, that the stars in the American flag are red, that
dogs can fly, and that elephants are taller than giraffes. A giraffe was, on three
different occasions, defined as a smaller version of some other animal (Table II).
All confabulatory responses, though sometimes fantastic, were appropriate to the
semantic category being discussed. Animals, for example, were never associated
with the attributes of inanimate objects. Atypical confabulatory responses were
not obtained from a control patient who had Korsakoff's disease when we
administered the tests of animal description to him. This control patient did
produce typical confabulations.

TABLE II

Excerpts from Test of Animal Description

Item 3
Examiner: Describe a giraffe for me.
Subject: Giraffe? Boy, that's an animal I've never seen that much, only in certain countries
where they have a lot of, uh, giraffes. Uh, an animal that uh, uh, like a field, uh, where
there's grass.
Examiner: What does it look like?
Subject: Well, he's a smaller animal. He's uh like something that would fly and uh in
dangerous in that way.
Item 6
Examiner: How about a tiger? What would you tell me about a tiger?
Subject: A tiger you'd find nine times out of ten in a foreign country where they have more
tigers than just about any other animal and like, like Africa; (yes, O.K.) and that I would
say you'd find more, although we got' em in the United States injhe southern part. I'd say
the southern part of the United States with farms that tried to g{ow 'em or something. Jeez,
whatever they've done uh, they've made a mistake I would say.
Item 12
Examiner: The last one I'm going to ask you is about an elephant. What's an elephant
like?
Subject: Huh. An elephant, oh brother. He's, he could, uh, well, I don't know if they're
actually that much trouble, but they haye 'em on certain farms or around certain farms.
Where there's a lot of trees 'n every... cause they're another animal, God, they live in trees,
they live in pretty good uh, acreage. Uh, live near uh where there's water.

Confabulation in aphasia

625

Semantic Representation

10/30 written non-words and 15/60 spoken non-words were misclassified on
the lexical decision task; that is, the non-words were incorrectly identified as real
words. Further, confabulated definitions were produced for some of the misclassified non-words (Table III) both spontaneously and when meanings were
requested (e.g. barg - "the situation or way you're doing something, either in
writing or in speaking"). All but one real word, "fig" presented in the auditory
modality, were correctly classified. 5/33 written words were sorted into the wrong
superordinate category on the categorization test.
TABLE III

Excerpts of Definitions for Printed Nonsense Words
Spood: Well, a spood, its a part of, its part of a, ooh, I say, urn, a dramatic thing, something
about doing with, uh, pictures or clothes or like, uh, (unintelligible), spumatic, something
spumatic. What's that mean? That, that's like men who build a house, even, or build a
platform. Women who go to a certain store to buy certain clothes. It's not the perfect word
but I know it, that's what it means.
Cratem: ... Cratman, it's, it must be a certain individual, uh, way of doing something.... A
woman wants to make a coat or a jacket. She could buy the right material. Right? And the
right quarter in order to make it right. Cratman, and this might be part of the individual or
what she's buying or the certain way she's doing it.

Perseveration

There was no perseveration of response set on the task requiring alternation
and reversal of symbols. There were verbal recurrences on the Visual/Verbal test
but these perseverations did not seem to interfere with conceptual shifting.
Verbal perseverations were prominent on tests of naming and description.
When he was re-tested six months post-onset, 2/14 written non-words were
misclassified. Responses on a test of animal description remained grossly confabulatory. At the time of re-testing, the subject was living independently in his
own home.
DISCUSSION

Two distinctive forms of confabulation occurred in this patient in the absence
of a confusional state or severe memory disturbance. Detailed investigation of the
nature of his confabulatory responses suggests mechanisms which produce the
two separate types of confabulation. (1) The typical confabulations occur only in
the context of small gaps in recent memory and only when self-monitoring fails
and memory deficits are denied. (2) The aphasic confabulations occur only in the
context of impaired or missing semantic representation (e.g. reduced knowledge
about word and category boundaries, loss of meaning for some words), but also
only with defective self-monitoring and denial of language deficit.

626

Jennifer Samlson, Martin L Albert and Michael P. Alexander

Two additional factors, word meaning alienation (e.g. Yamadori and Albert,
1973) and perseveration, may also play a role in the aphasic confabulations of this
patient. Confabulatory responses were always noted on the infrequent occasions
when the patient indicated that he had no recognition of a common stimulus. The
patient asserted, for example, that the word "bubblegum" was totally unfamiliar
to him, although he then stated that it was probably blue.
The aphasic confabulations of this patient often contained elements from,
previous responses. The statement that elephants live in trees followed the di.scussion of several animals where trees were appropriately mentioned. Perseveration may, in fact, also play a role in explaining why typical confabulations often
contain elements from the patient's recent experiences.
ABSTRACT

We describe a brain-damaged patient who produced two distinctive forms of
confabulation, typical and atypical (aphasic), both of which were associated with
perseveration, poor self-monitoring, and denial of deficit. We propose that the
typical confabulations are triggered by gaps in memory for the period surrounding the onset of his· illness, while the aphasic (fantastic) confabulations are
triggered by gaps in semantic representation. These observations suggest that
confabulations may comprise a portion of the language deficit in patients with
other varieties of fluent aphasia.
Acknowledgments. This research was supported in part by the Veterans
Administration Medical Research Service, NIH Grant No. NS06209, and the
Seidel Fund for the Study of Dementia.
REFERENCES

GOODGLASS, H., and KAPLAN, E. The Assessment of Aphasia and Related Disorders.
Philadelphia: Lea and Febiger, 1983.
MERCER, B., WAPNER, W., GARDNER, H., and BENBON, D.F. A study of confabulation.
Archives of Neurology, 34: 429-433, 1977.
.
SANDSON, J., and ALBERT, M.L. Varieties of perseveration. Neuropsychologia, 22: 715-732,
1984.
SHALLICE, T., and EVANS, M. The involvement of the frontal lobes in cognitive estimation.
Cortex, 14: 294-303, 1978.
SHAPIRO, B., ALEXANDER, M., GARDNER, H., and MERCER, B. Mechanisms of confabulation. Neurology, 31: 1070-1076, 1981.
Sruss, D., ALEXANDER, M., LIEBERMAN, A., and LEVINE, H. An extraordinary form of
confabulation. Neurology, 28: 1166-1172, 1978.
WEINSTEIN, E. Behavioral aspects of jargonaphasia. In J. Brown (Ed.), Jargon Aphasia.
New York: Academic Press, 1981, pp. 139-148.
WEINSTEIN, E., and LYERLY, 0. Confabulation following brain injury. Archives of General
Psychiatry, 18: 348-354, 1966.
YAMADORI, A., and ALBERT, M.L. Word category aphasia. Cortex, 9: 83-89, 1973.
Jennifer Sandson, Department of Neurology, Boston Veterans Administration Medical Center, 150 South Huntington Ave., Boston, MA 02130.