Content uploaded by Faraneh Vargha-Khadem
Author content
All content in this area was uploaded by Faraneh Vargha-Khadem on Apr 12, 2014
Content may be subject to copyright.
Preserved Recognition in a Case of Developmental
Amnesia: Implications for the Acquisition of
Semantic Memory?
Alan Baddeley
1
, Faraneh Vargha-Khadem
2
, and Mortimer Mishkin
3
Abstract
& We report the performance on recognition memory tests of
Jon, who, despite amnesia from early childhood, has developed
normal levels of performance on tests of intelligence, language,
and general knowledge. Despite impaired recall, he performed
within the normal range on each of six recognition tests, but he
appears to lack the recollective phenomenological experience
normally associated with episodic memory. His recall of
previously unfamiliar newsreel events was impaired, but gained
substantially from repetition over a 2-day period. Our results
are consistent with the hypothesis that the recollective process
of episodic memory is not necessary either for recognition or
for the acquisition of semantic knowledge. &
INTRODUCTION
In the last few decades, there has been extensive
research into the cognitive neuropsychology of the
amnesic syndrome, ranging from Milner's (1966) early
research on patient HM, through the attempts to relate
the functional deficits of amnesic patients to laboratory
studies of normal subjects (Squire, Knowlton, & Musen,
1993; Baddeley & Warrington, 1970), up to the present
situation in which it is widely accepted that data from
study of the amnesic syndrome make a crucial contribu-
tion to our current concepts of normal human memory
(Schacter & Tulving, 1994). While there have been
numerous attempts to argue for different types of
amnesia usually related to differential lesions (Parkin,
Leng, & Hunkin, 1990; Huppert & Piercy, 1979), such
subdivisions have proved difficult to sustain, leading to a
somewhat negative, though by no means universal con-
sensus of a unitary amnesic syndrome (Squire, 1992;
Baddeley, 1997).
This modal view distinguishes between explicit or
declarative memory that is grossly impaired in amnesia,
and a range of implicit, procedural, or nondeclarative
memory systems that are intact (Baddeley, 1997;
Squire, 1992). These implicit memory systems have in
common the fact that they operate in the absence of
conscious recollection.
Explicit or declarative memory can itself be split into
two components: episodic memory, which represents
the capacity to store and recollect experienced events;
and semantic memory, a system that stores and retrieves
knowledge about the world (Tulving, 1972). The term
episodic memory is used generally to refer to the
capacity to recollect past experience, either by recalling
the event, or in the case of recognition, by recollecting
information associated with the learning experience
(Tulving, 1972). This phenomenological aspect of epi-
sodic memory is essential to Tulving's use of the term,
although others use the term somewhat more broadly.
There has in recent years been considerable interest in
the phenomenology of recognition, with considerable
evidence that subjects are successfully able to categorize
items they have recognized on the basis of whether they
have been remembered with associated recollective
experience, or whether subjects simply ``know'' that
the item had previously been presented (Gardiner &
Java, 1993).
Semantic memory is assumed to reflect accumulated
world knowledge such as the meaning of the word
``fish,'' the color and taste of a banana, the number of
yards in a mile, or the normal procedure for eating in a
restaurant. While amnesic patients usually retain seman-
tic knowledge acquired before the onset of amnesia,
they typically have great difficulty in adding further
information to their semantic memory systems, for
example, amnesic patients would generally be unable
to provide the names of current political figures such as
the U.S. President, and would be unaware of recent
world events and of the meaning of newly coined words
(Squire et al., 1993). The association of impaired episo-
dic memory with the failure to update semantic memory
is consistent with the view that semantic memory re-
1
University of Bristol,
2
University College London and Great
Ormond Street Hospital for Children, London,
3
National
Institute of Mental Health, Bethesda, MD
D 2001 Massachusetts Institute of Technology Journal of Cognitive Neuroscience 13:3, pp. 357±369
flects the residue of many episodes (Baddeley, 1997;
Squire, 1992), although this view is not universally held
(Mishkin, Vargha-Khadem, & Gadian, 1998; Tulving,
1972; Tulving & Markovitch, 1998).
Although this overall view has, for many years, been
able to account for the bulk of evidence, it has recently
been challenged by the description of what appears to
be a new amnesic syndrome. The patients concerned
appear to have developed amnesia to a greater or lesser
degree at, or within, a few years of birth, a situation
which should, according to the modal view, have led to
grossly impaired semantic memory. Nevertheless, three
such cases reported by Vargha-Khadem et al. (1997)
have demonstrated an apparently normal acquisition of
language, together with a remarkably well developed
knowledge of the world. The implication of such cases
for the concept of semantic memory is discussed else-
where (Mishkin et al., 1998; Squire & Zola, 1998; Tulving
& Markovitch, 1998); the purpose of the present study is
to investigate further the nature of the memory deficit,
and in particular, to provide a more robust test of the
suggestion by Vargha-Khadem et al. that this type of
amnesia may be associated with preserved recognition
memory despite clear evidence of impaired recall. Such
a dissociation is not typical of the amnesic syndrome:
Patients not only fail to successfully recollect prior
experiences, but they also fail to distinguish between
novel words, pictures or events, and items that have
already been presented (Squire et al., 1993).
We begin by recapitulating the case of Jon, one of the
three described by Vargha-Khadem et al., illustrating the
evidence for his amnesia, together with the comparative
preservation of his semantic memory. This is then
followed by a summary of his neuropathology as deter-
mined by quantitative magnetic resonance techniques.
We then go on to a brief account of earlier claims of
preserved recognition in patients who have acquired
their amnesic deficit as a result of brain injury during
adulthood. We discuss some of the methodological
problems facing any attempt to compare recall and
recognition, before describing a study in which we
compare Jon's recall and recognition for both visual
and verbal material. In order to study the generality of
our findings, we then compare Jon's performance to
that of normal control subjects across several other
recognition paradigms. We then discuss the implications
of our results for the distinction between recall and
recognition memory, and for the development of se-
mantic memory.
The tests used comprised both standardized mea-
sures for which appropriate norms are available, and
less widely used experimental tests. For the latter, the
question arises as to the appropriate control compar-
ison. When tested at age 19, Jon showed a slightly
unusual WAIS-R intelligence profile, with his perfor-
mance IQ (120) being somewhat greater than his verbal
IQ (108). Finding age- and education-matched subjects
with a similar profile, who were also able to absent
themselves from work for the several days necessary to
complete the full test battery, proved difficult. We were,
however, able to locate two such control subjects.
Their results were used, together with results from
the somewhat larger but less well matched groups
from the published literature, as a baseline for inter-
preting Jon's performance.
Case Description
Jon, who is now 23, was born prematurely at 26 weeks
of gestation, weighing less than 1 kg and suffering from
breathing problems (Gadian et al., 2000). For prolonged
periods during his first six weeks of life, he suffered
from severe apnea, requiring intubation and positive
pressure ventilation. He subsequently showed steady
improvement and normal development. At 3 years 10
months, he had an unconfirmed convulsive episode in
association with cold and cough. His memory problems
were first noted when he was about 5 years old, and
have since continued to be prominent. More specifi-
cally, Jon's parents report that he has difficulty in
reliably finding his way, tends to forget where objects
and belongings are usually kept or where they have just
been put, and makes many prospective memory errors,
finding it difficult to keep even regularly scheduled
appointments unless reminded. Jon also has difficulty
remembering details of everyday activities such as con-
versations or television programs watched. The severity
of the disability is such as to make it difficult for Jon to
cope independently, or to succeed in finding and keep-
ing a job.
Table 1 shows the performance of Jon and the two
control subjects, Andrew and Max, on intelligence, scho-
lastic, and language performance tests, all carried out
between the ages of 19 and 22. Jon has above-average
intelligence with a full-scale IQ of 114, with subtest
scores ranging from scaled scores of 10 (the normal
average) on information, to 16 (two standard deviations
above the mean) on block design. His number skills as
measured on the Wechsler Objective Numerical Dimen-
sions (WOND) test are above average, while his reading
performance is within the normal range for basic reading
(55th percentile) and comprehension (42nd percentile)
coupled with poor spelling (14th percentile). Jon's
language is also within the normal range as measured
by the Token Test, which requires the following of
instructions of increasing grammatical complexity, and
his semantic judgements appear to be normal, as mea-
sured by the Pyramids and Palm Trees Test, a test of
semantic access from words and pictures (Howard &
Patterson, 1992). Jon's vocabulary development is also
within the normal range, with the WAIS-R vocabulary
subtest producing a scaled score of 14, while his com-
prehension subtest score was 11. Finally, his perfor-
mance on the Spot the Word Test (Baddeley, Emslie,
358 Journal of Cognitive Neuroscience Volume 13, Number 3
& Nimmo-Smith, 1993) was also within the normal
range. This is a lexical decision task in which pairs of
items are presented, one comprising a word and one a
nonword; Jon's score of 47 out of 60 places him at the
50th percentile. In conclusion, apart from poor spelling,
Jon has above-average intelligence and performs nor-
mally on scholastic and language tests, including eight
different tests of semantic or language processing.
In 1991, Jon was tested on the Wisconsin Card Sorting
Test, using the original Milner (1963) procedure. He
obtained five categories, making a total of 25 errors of
which 24 were perseverative. This indicates performance
within the normal range. His category fluency tested at
this time was also normal (20 animal names and 14 fruit
types). In 1997, he was tested on the FAS initial letter
fluency task, producing 50 words, which places him
between the 70th and 80th percentile. In the same year,
he was tested on the Behavioural Assessment of the
Dysexecutive Syndrome (BADS) (Wilson, Alderman, Bur-
gess, Emslie, & Evans, 1996), which comprises a number
of subtests sensitive to different aspects of executive
processing. He scored 21/24, an entirely satisfactory
performance. He was noted to be thinking ahead, plan-
ning for the future, but easily distracted, and with some
trouble with decision-making. In general, Jon's executive
capacities appear to be within the normal range.
Table 2 shows the performance of Jon, Andrew, and
Max on standardized memory tests. On the recently
developedExtendedRivermeadBehaviouralMemory
Test (Wilson et al., 1999), his profile score of 10 indicates
a level of performance that is well within the impaired
range (0±18). Jon's visual memory, as indicated by delayed
recall of the complex Rey Figure, was also dramatically
impaired with no subsequent scoreable reproduction.
The Children's Auditory Verbal Learning Test-2 (with
norms ranging from 6 to 18 years) involves presenting a
sequence of 16 words five times, after which a second
list is presented and tested (Delis, Kramer, Kaplan, &
Ober, 1987). The initial list is then presented again and
tested for recall and finally recognition. Jon does well on
Table 1. IQ, Academic Attainments and Language
Test Jon Andrew Max
WAIS-R Standard score Range Standard score Range Standard score Range
Full-scale IQ 114 High average 114 High average 101 Average
Verbal IQ 108 Average 107 Average 97 Average
Performance IQ 120 High 120 High 108 Average
Years of education 13 13 13
Academic attainments Standard score Percentile Standard score Percentile Standard score Percentile
WOND
Mathematics reasoning 103 58 118 88 118 88
Numerical operations 115 84 108 70 119 90
WORD
Basic reading 102 55 108 70 102 55
Spelling 84 14 111 77 102 55
Reading comprehension 97 42 102 55 86 18
Language Raw score Range Raw score Range Raw score Range
Token Test 60/62 Normal n/a
a
n/a 56/62 Normal
Pyramids and Palm Trees 49/52 Normal 52/52 Normal 51/52 Normal
Scaled score Range Scaled score Range Scaled score Range
WAIS-R vocabulary 12 Average 12 Average 9 Average
WAIS-R comprehension 10 Average 12 Average 10 Average
Spot the Word 10 Average 11 Average 9 Average
a
Not administered.
Baddeley, Vargha-Khadem, and Mishkin 359
the initial test (68th percentile) relying heavily on recal-
ling the last few words, the recency effect, which reflects
short-term memory and hence, tends to be preserved in
amnesia (Baddeley & Warrington, 1970). His subsequent
scores of 8, 7, 8, and 10 show little evidence of learning,
continuing to reflect a reliance on recency that is atypical
in subjects with normal memory (Greene, Baddeley, &
Hodges, 1996). Following the interfering list, a recall of
four correct items with seven intrusions places Jon
below the first percentile. In contrast, a final recognition
score of 29 out of 32 places him just outside the normal
cutoff of 30. A similar pattern of extremely poor word
recall was shown when he was tested some 5 years
earlier on a previous version of the test, whereas his
recognition scores of 29/32 and 30/32 on two different
occasions were both within the normal range for this
older version of the test.
The Warrington Recognition Memory Test involves
presenting 50 words and 50 faces, and in each case
requiring subsequent two-alternative, forced-choice re-
cognition (Warrington, 1984). In both cases, Jon scored
at or above the 25th percentile, less well than one might
expect from his general intelligence, but substantially
better than suggested by his impaired performance on
the various tests of recall described earlier. However, this
level of performance should be treated with caution
since Jon had performed this test on a number of
occasions in the past, albeit separated by intervals of
many months. Repetition might either have improved his
performance, as a result of his becoming more familiar
with the targets, or impaired it, since both targets and
distractors would have some degree of familiarity. It is
therefore clearly desirable to attempt to make a more
direct assessment of recognition memory, and, if possi-
ble, to compare it with a comparable test of recall.
Neuropathological findings
As reported previously (Gadian et al., 2000; Vargha-
Khadem et al., 1997), direct measurements performed
on Jon's magnetic resonance scans revealed that the
hippocampal volume in each hemisphere was approxi-
mately 50% less than the normal mean volume, a degree
of atrophy that was clearly visible on the scans. This brain
abnormality was also the only one visible, although
quantitative MR techniques suggested the presence of
more subtle pathology in other regions, including the
putamen bilaterally, the ventral part of the thalamus, and
the midbrain (Gadian et al., 2000). These additional
abnormalities, like the bilateral hippocampal atrophy,
are consistent with the known effects of hypoxia±ische-
mia, which is the presumed cause of the neuropathology
in Jon. It should be noted that none of these MR
techniques yielded evidence of bilateral pathology in
the medial temporal tissue outside the hippocampus.
In considering the unusual pattern of Jon's impaired and
spared memory abilities, it is important to keep in mind
this apparent selectivity of his ventral temporal damage,
as it has raised the interesting possibility that just as his
memory losses are attributable to the bilateral hippo-
Table 2. Memory Performance
Test Jon Andrew Max
Rivermead Behavioural Memory
TestÐExtended
Profile score Range Profile score Range Profile score Range
10 Impaired 33 Average 40 Good
CAVLT-2 Standard score Percentile Standard score Percentile Standard score Percentile
Trial 1 108 70 98 45 125 95
Trial 5 77 6 117 87 108 70
Post interference <60 <1 117 87 106 66
Delayed recall 63 1 118 88 118 88
Raw score Percentile Raw score Percentile Raw score Percentile
Recognition accuracy 29 <16 32 Normal 32 Normal
Intrusions 17 <16 0 Normal 0 Normal
Warrington Recognition Memory
Test
Raw score Percentile Raw score Percentile Raw score Percentile
Words 45 25 48 75 50 95
Faces 41 25 49 95 42 25
360 Journal of Cognitive Neuroscience Volume 13, Number 3
campal atrophy, his preserved memory abilities are due
at least in part to the preservation of the subjacent cortex
in the parahippocampal region (Mishkin, Suzuki, Gadian,
& Vargha-Khadem, 1997; Vargha-Khadem et al., 1997).
Neuropsychological Evidence for a
Recall±Recognition Dissociation
The claim that recall and recognition are differentially
affected by memory deficit has a long, if somewhat
checkered history. It has, for example, been suggested,
at least since Schonfield and Robertson (1966), that
aging has a greater impact on recall than on recogni-
tion, and for a while, it was suggested that the classic
amnesic syndrome might show such a dissociation
(Huppert & Piercy, 1979). Unfortunately, such early
studies are open to the objection that the recognition
tasks used were simply less demanding than the recall,
and it is now generally accepted that recognition
memory is not preserved in most patients suffering
from the amnesic syndrome (Aggleton & Brown, 1999;
Reed & Squire, 1997).
A number of atypical cases have, however, been
reported in which recognition does appear to have been
preserved. These include patients with bilateral fornix
damage (McMackin, Cockburn, Anslow, & Gaffan, 1995),
patients with damage in the region of the mammillary
bodies (Parkin & Hunkin, 1993; DuSoir, Kapur, Byrnes,
McKinstry, & Hoare, 1990), patients who sustained
anoxia (Volpe, Holtzman, & Hirst, 1986), cases with
alcohol-induced Wernicke's encephalopathy (Parkin,
Dunn, Lee, O'Hara, & Nussbaum, 1993), and a case with
anterior communicating artery aneurysm (Hanley, Da-
vies, Downes, & Mayes, 1994).
In a recent review of the amnesia literature, Aggleton
and Brown (1999) and Aggleton and Shaw (1996) have
argued for a preservation of recognition memory in
those cases of amnesia where the damage is limited to
the hippocampus or its diencephalic targets. Aggleton
and Brown argue that this is entirely consistent with the
literature from experimental lesions in animals, most of
which indicates that recognition memory is unimpaired
by lesions limited to the hippocampus, but is impaired
when extrahippocampal regions are damaged, such as
the perirhinal cortex and related regions. This view is
strongly opposed by Reed and Squire (1997), who
report a series of patients for whom the damage is
``limited primarily to the hippocampus or the hippo-
campal formation,'' and who show clear evidence of
impaired recognition. As Aggleton and Brown point out,
however, the crucial question is that of whether areas
beyond the hippocampus were indeed completely pre-
served; something that is difficult to establish, particu-
larly in view of the fact that brain areas that appear to be
structurally intact when imaged using MRI may show
reduced functional activity when studied using PET
(Aggleton & Brown, 1999). Jon's syndrome of amnesia
combined with apparent sparing of recognition memory
might thus be claimed to stem from a reduction in the
development of the hippocampus, together with appar-
ent sparing of other brain tissue associated with memory
function (Gadian et al., 2000; Mishkin et al., 1998;
Vargha-Khadem et al., 1997), hence, to be consistent
with the hypothesis proposed by Aggleton and Brown.
However, the principal function of the present study is
to explore the functional rather than the anatomical
characteristics of Jon's memory deficit and to discuss its
relevance to the apparent preservation of his semantic
memory development.
Problems in Comparing Recall and Recognition
As demonstrated by Mandler, Pearlstone, and Koopmans
(1969), it is clearly unsatisfactory to assume that tasks
based on recall and recognition of the same material will
be equivalent in difficulty, as measured by percentage of
correct responses. One way of attempting to circumvent
this is to attempt to adjust the difficulty of one task so
that it appears to give a score equivalent to that on the
other. Calev (1984), for example, adjusted the level of
recall and recognition by using categorized words for
recall and unrelated words for recognition. This of
course introduces potentially important differences be-
tween the two tasks. This procedure is also open to the
objection that it assumes that having found one point at
which the distribution of scores from the recall and
recognition methods give a similar percentage, then the
rest of the distribution of scores will also be equivalent.
This is not necessarily the case. Shallice (1988) makes a
similar point in challenging the claim by Hirst et al.
(1986) of having demonstrated a recall±recognition dis-
sociation in amnesic patients, pointing out the way in
which a difference in variance between two measures
could also give rise to a spurious apparent dissociation.
The absence of any differential effects of amnesia on
recall and recognition in most instances was further
supported by a thorough study by Haist, Shimamura,
& Squire (1992).
To avoid the earlier problems in comparing recall and
recognition, Baddeley, Emslie, and Nimmo-Smith (1994)
attempted to produce tests of visual and verbal recall
and recognition of equivalent difficulty by using scaled
scores. Such scores express a patient's performance on
each test in terms of Z scores based on the distribution
of normal population scores on that test. Using this
method, separate measures of visual and verbal mem-
ory, and of recall and recognition, may be obtained
together with measures of learning and forgetting. The
test has proved sensitive to a range of factors known to
influence episodic memory, including Alzheimer's Dis-
ease, normal aging, and schizophrenia (Baddeley, 1996);
the test is also able to differentiate between the effects
of right and left temporal lobectomy on visual and verbal
memory (Morris, Abrahams, Baddeley, & Polkey, 1995).
Baddeley, Vargha-Khadem, and Mishkin 361
Less dramatic visual±verbal memory differences were
also found in other groups, with the elderly tending to
be more impaired on names than on visual stimuli, while
the opposite pattern occurred for schizophrenic pa-
tients (Baddeley, 1996). On the other hand, there has
so far been no evidence reported for a marked recall±
recognition difference for either etiological groups or
individuals within the various groups. Since the subjects
assessed using the test have varied markedly in memory
ability, from elderly Alzheimer patients (Greene et al.,
1996) to young Ph.D. students (Baddeley, 1993), this
suggests that any recall±recognition discrepancy ob-
served is unlikely to be an artefact resulting from ceiling
or floor effects.
RECOGNITION TEST PERFORMANCE
Doors and People Test
We began by testing Jon on the Doors and People Test, a
relatively pure test of visual memory using recognition of
sets of photographs of doors, and recall by drawing of
four different versions of a cross (Baddeley et al., 1994).
These visual memory tests are then contrasted with
verbal recognition and recall based on people's names.
The results are shown in Figure 1, where performance
on the four subtests is expressed in terms of scaled
scores; a score of 10 represents the population mean for
subjects of Jon's age, with each scaled point represent-
ing one-third of a standard deviation. For both visual and
verbal recall memory, Jon scores below the 5th percen-
tile, while scoring between the 50th and 75th percentile
for recognition. Neither of the control subjects show a
clear recall±recognition discrepancy. The various scores
can be combined to estimate the likelihood of finding a
visual±verbal or recall±recognition discrepancy of this
magnitude within the population. In the case of the
visual±verbal distinction, Jon has a scaled score of 10,
Andrew a scaled score of 9, and Max a score of 9; hence,
none of the three show a clear difference in perfor-
mance between pictorial and verbal material. In the case
of the recall±recognition distinction, however, the dis-
crepancy gives a scaled recall±recognition difference
score for Jon of below 3, which represents the bottom
of the scale on this particular test measure. In contrast,
Andrew and Max both have scaled scores of 9, placing
them within one-third of a standard deviation of the
population norm.
The results of the Doors and People Test indicate a
marked discrepancy in Jon between clearly impaired
recall and recognition performance that is entirely with-
in the normal range. However, in discussing the claims
for a recall±recognition discrepancy in patients with
damage limited to the hippocampus, Reed and Squire
(1997) criticize Aggleton and Shaw for relying too heavily
on a limited range of tests, and on relatively short tests
with immediate recognition. For that reason, we ex-
tended our investigation to cover three further tests,
one of which uses auditory presentation and rapid
paced recognition (Andrade, Munglani, Jones, & Badde-
ley, 1994; Andrade, 1996; Shepard & Teghtsoonian,
1961), while the others are tests developed by Squire
and his collaborators with the explicit purpose of invol-
ving a reasonably long sequence of words, together with
an interpolated delay (Hamann & Squire, 1997). None of
these tests have standardized norms, but in all cases,
data from normal control subjects are available, as are
data from Andrew and Max.
Continuous Recognition Test
Shepard and Teghtsoonian (1961) developed a test in
which a sequence of words is presented, and the subject
is required to perform a running recognition task,
deciding when each word is presented, whether it has
been shown previously. Repetitions can occur immedi-
ately, or after 1, 2, 4, 8, or 16 intervening words.
Despite the rapidly paced presentation, Jon detected
all repetitions with the exception of one at a delay of
four, and two at an eight-word delay. Jon's score of
97.3% is slightly higher than the scores of Andrew
(86.5%) or Max (83.0%) and exceeds the mean scores
of 92.9% and 94.4% reported for normal subjects (an-
esthetists) in two studies by Andrade et al. (1994) and
Andrade (1996). However, Jon made an average of 0.5
false alarms per list compared with 0.31 by Andrew,
none by Max, and an average of 0.20 reported by
Andrade et al., although Jon's score was within the
range observed in that study. Because of the nature of
the test, the probability of a repetition increases as the
list progresses, making a bias-free estimate of memory
problematical. The next two tests avoid this problem in
two different ways, one by using two-alternative forced
choice, where a single response is always required, and
the other by holding the probability of a target constant
at 0.5 and noting both detection and false alarm rates.
Figure 1. The performance of Jon and two control subjects on the
Doors and People Test of Visual and Verbal Recall and Recognition. Jon
scores at the average level on both recognition tests, but is impaired
on the two recall measures.
362 Journal of Cognitive Neuroscience Volume 13, Number 3
Delayed Recognition Tests
The next recognition tests comprised the yes±no recog-
nition and two-alternative forced-choice recognition
measures used by Hamann and Squire (1997), both
tested after a 5-min filled delay.
Figure 2a shows Jon's performance on forced-choice
recognition, together with data from Andrew and Max
and from Hamann and Squire, illustrating the overall
level of performance from their amnesic and control
groups, based on six separate tests given over 2 days.
Jon's performance is very close to that of the two age-
matched control subjects and of Hamann and Squire's
control subjects, and substantially above that of their
amnesic patients.
Figure 2b shows the equivalent data for yes±no re-
cognition performance. In the case of forced-choice
recognition, performance is measured by simple percen-
tage correct. In the case of yes±no recognition, hits and
false alarms are combined to create d
0
, a measure of
discrimination accuracy; Jon's average detection prob-
ability is .813, while his average false alarm rate is .201
giving him a d
0
score of 1.76. As is clear from Figure 2b,
Jon's performance broadly resembles the average of
Hamann and Squire's control subjects (mean d
0
=
1.89) and falls between that of Andrew (d
0
= 1.19) and
Max (d
0
= 2.12). Performance on this task was highly
variable, with Andrew and one of Hamann and Squire's
control subjects scoring below the level of one of their
amnesic patients. This suggests that subjects may have
difficulty in setting an appropriate criterion, since per-
formance under forced-choice conditions, which re-
move the problem of criterion, does not show this
variability.
To this point, we have reported Jon's scores on a
range of recognition tests, both visual and verbal. On all
of these, Jon has performed at, or even somewhat
above, the average level to be expected from a normal
subject. Method of testing has included both forced-
choice and yes±no recognition, tested both immediately
and after a 5-min filled delay. Finally, recognition has
been tested under both unpaced and rapidly paced
conditions. We therefore feel reasonably confident in
claiming that despite clear evidence of his amnesia, Jon
has a well preserved recognition memory. As Reed and
Squire (1997) point out, such a dissociation is certainly
not typical of the amnesic syndrome. Jon is therefore
atypical in two respects; first, in his recall±recognition
disparity, and second, in his spared capacity to incre-
ment semantic memory despite his episodic memory
deficit, which again is not characteristic of the amnesic
syndrome (Squire et al., 1993).
At this point, we encounter an apparent paradox.
Jon's semantic memory is not limited to his ability to
recognize facts about the world. He is quite capable of
recalling the information necessary for defining words,
can recall the names of historic figures, and discuss at
considerable length issues such as current politics. All
these suggest that it is not the capacity to recall per se
that is impaired, but rather the ability to use episodic
memory in order to facilitate the recall of newly acquired
material such as lists of unrelated words presented once.
The fact that Jon has acquired an extensive amount of
semantic knowledge suggests that learning, whether
tested by recall or recognition, should be possible under
conditions where naturalistic material is presented on
multiple occasions. While it is clearly impractical to
simulate adequately a lifetime of rich experience, we
went some way in this direction by presenting Jon with a
series of videos based on newsreels of events from the
years 1937 and 1957. One of these, 1937, was presented
four times distributed over a 2-day period, while the
other was presented only once. Both recall and recogni-
tion were tested immediately after the final presentation
in each case, and were retested the following morning
after an interval of approximately 18 hr.
The results shown in Figure 3 indicate first of all that
after a single presentation, Jon's immediate recognition
performance is comparable to that of the control sub-
jects, and even his overnight recognition is only slightly
below theirs. His immediate recall score, by contrast, is
clearly much lower than theirs. However, after four
presentations, Jon's immediate recall score has risen
substantially, and indeed, slightly exceeds that of An-
drew, while all three show a modest increase in recogni-
tion scores. It is clearly the case, therefore, that Jon's
recall deficit can be minimized given sufficient practice.
The overnight recall data are somewhat more difficult
to interpret, since there is evidence to indicate that
successfully performing a recognition test can enhance
subsequent recall. This probably occurs because the
presentation of recognition alternatives may provide a
reminder, which in turn acts as a relearning trial. For
example, one of the 1937 questions concerned a Zep-
pelin disaster, asking for the name of the craft involved,
and subsequently providing four alternatives, the Bis-
Figure 2. Delayed verbal recognition memory as measured by two-
alternative forced-choice (a) and yes±no (b) recognition. On both
measures, Jon resembles the normal control subjects rather than the
amnesic patients of Hamann & Squire (1997).
Baddeley, Vargha-Khadem, and Mishkin 363
mark, the Mecklenburg, the Deutschland, and the Hin-
denberg. Having been reminded of a correct answer that
could not be recalled, the subject may then retain the
information. The most important feature of these re-
sults, however, is the fact that Jon is able to retain the
information he has acquired across the overnight delay
of 18 hr, whether tested by recall or recognition; long-
term retention of newly acquired knowledge is, of
course, a necessary feature for the incrementation of
semantic memory.
While variables that influence recall typically also
affect recognition, the question of whether they reflect
equivalent processes has remained controversial, at least
since Mandler (1980) proposed a two-component inter-
pretation of recognition. He proposed that a target word
might be subsequently recognized on the basis of either
the recollection of having encountered that word, e.g.,
because it reminded one of the earlier learning experi-
ence, or alternatively, simply on the basis of a feeling of
familiarity. This line of discussion was further developed
by Tulving (1985), who defines episodic memory in
terms of the subjective experience of recollection. He
proposed that requiring subjects to judge whether they
truly ``remember'' a word, or simply ``know'' that it was
previously presented may prove to be a productive way
of investigating episodic memory. This has indeed
proved to be the case, with a whole range of variables
including age, type of processing, and degree of atten-
tion influencing ``remember'' judgements while having
little influence on ``know'' responses (Baddeley, 1997;
Gardiner & Java, 1993).
A possible interpretation of our results is to assume
that while Jon retains the capacity to acquire new
information as reflected by a greater sense of familiarity
and general accessibility, he lacks the capacity to recol-
lect the learning experience, leading to the absence of
associated sensory and contextual information that en-
ables subjects to say they ``remember'' an event, rather
than simply ``know'' that it happened.
We attempted to test this interpretation, spending a
considerable time trying to explain to Jon what it means
to ``remember'' something, as opposed to simply know-
ing that it had been presented earlier. We explained that
the ``remember'' judgement is to be given whenever a
recognized word reminds the subject of the learning
experience, e.g., the word ``dachsund'' might be remem-
bered because of recalling that when it was presented it
reminded the subject of a friend's pet seen last week, or
perhaps because it seemed to be associated with an
earlier item such as ``sausage.'' We tried to get Jon to
recollect events such as his journey to the Institute that
morning. He gave a relatively detailed account but was
unable to provide any specific detail that might differ-
entiate this from previous journeys. He indicated, for
instance, that he had attempted to run up the final flight
of stairs in one rush, and had got just over halfway.
However, when asked how he knew this, he replied ``I
always get about halfway.'' We nevertheless decided to
proceed, using material and testing procedures taken
from Parkin and Walter (1992), in which words are
presented for learning and then tested by yes±no re-
cognition, with subjects categorizing the ``yes'' re-
sponses as being ``remembered'' or ``known.'' This
method has the safeguard of subsequently checking a
subsample of ``remember'' judgements and requiring
the subject to give the reason for his judgement. In
terms of overall performance, Jon had a hit rate of .917
and a false alarm rate of .194, giving him a d
0
score of
2.25. This is slightly lower than the d
0
obtained by the
young control subjects (mean = 2.70, SD = 0.7), but
clearly higher than that obtained by the normal elderly
subjects (d
0
= 1.90, SD = 0.8). At first sight, Jon
appeared to be using the ``remember'' judgement even
more readily than the young, and substantially more
often than the elderly subjects. However, in each case,
when subsequently asked why he gave the ``remember''
judgement, he provided the same account, namely that
he looked at the word, tried to imagine it printed on the
stimulus card, and if this resulted in an immediate and
clear image, gave a ``remember'' response. Such a
method is much closer to the use of perceptual fluency,
a method commonly associated with implicit rather than
recollective memory (Whittlesea, 1993). The fact that we
were unable to train Jon to use the ``remember/know''
Figure 3. Recall and recognition of events from unfamiliar news
videos viewed on one or four occasions and tested immediately or after
an overnight delay.
364 Journal of Cognitive Neuroscience Volume 13, Number 3
distinction appropriately is entirely consistent with the
hypothesis that he lacks the capacity to recollect the
contextual detail necessary for an appropriate ``remem-
ber'' response. However, a conclusion that relies on a
failure to teach the subject to use a relatively subtle
concept clearly does not provide adequately strong
evidence concerning the hypothesis. Further investiga-
tion is ongoing (see below), based on the question of
whether Jon exhibits the electrophysiological pattern
that has been shown to be associated with ``remember''
judgements (Rugg, Schloerscheidt, & Mark, 1998).
DISCUSSION
Despite Jon's impaired memory when measured on
tests that rely on recall, his recognition memory was
found to be normal across a range of tests involving
different materials, speeds of presentation, and recogni-
tion paradigms. The single exception was his slightly
lowered score for overnight recognition of items on the
newsreel presented only once. On the basis of this, we
would claim that Jon's recognition memory perfor-
mance falls within the normal range. More specifically,
his recall and recognition scores on the Doors and
People Test explicitly designed to allow a recall±recog-
nition contrast, demonstrate very clearly impaired recall
and preserved recognition. Such a dissociation is not
attributable to the differential sensitivity of the two
measures, since it is not shown by Alzheimer's disease
patients (Greene et al., 1996), by schizophrenic patients
or normal elderly subjects (Baddeley, 1996), or by a
group of patients recently reported by Manns and
Squire (1999), selected as having lesions that are located
principally in the hippocampus.
With the single exception noted above, Jon's recogni-
tion performance is consistently within the normal
range; can we therefore conclude that his performance
is unimpaired? This would be unwise; Jon has above-
average intelligence, is highly motivated, and is an
experienced and sophisticated subject. He does, for
example, attempt to optimize his guessing by keeping
count of the proportion of ``yes'' responses he makes. It
is thus possible that, in the absence of neurological
damage, he would have performed above average. The
important feature of Jon's case, therefore, is the clear
dissociation between recall and recognition rather than
the seemingly complete preservation of recognition
performance. It is important to bear this in mind, both
when seeking to replicate our findings and in consider-
ing possible theoretical explanations.
If we accept that the dissociation is indeed genuine,
what are its implications for the recall±recognition dis-
tinction and for understanding the specific association of
symptoms found in Jon, namely impaired recall memory
coupled with preserved language, semantic memory,
and intelligence? It is initially tempting to argue that
the preservation of recognition in the presence of clearly
impaired recall suggests that they reflect two separate
memory systems. This would be premature for two
reasons. First, our results are based on a single dissocia-
tion rather than the much more powerful double dis-
sociation, for which we would need a case in which
recall is preserved and recognition impaired. Second, it
is important to bear in mind that recall and recognition
are experimental methods of tapping memory, not
themselves direct measures of underlying memory sys-
tems, although they may differ in their relative reliance
on underlying systems. Finally, however, Jon clearly is
capable of recalling a good deal of information of both a
semantic and personal semantic nature, suggesting that
recall per se is not necessarily defective.
It seems more productive to consider interpreting our
findings in terms of the processes underlying recall and
recognition. It is now widely accepted that human long-
term memory is not unitary, reflecting at least two
underlying processes or systems. One involves the con-
scious recognition and recall of previously experienced
events or facts, termed by Squire and Zola (1998) as
declarative memory. This system, in which the subject is
aware of the fact that he or she is remembering, may be
contrasted with a range of nondeclarative or implicit
memory processes in which prior learning is reflected
indirectly through enhanced performance. Nondeclara-
tive memory may, for instance, occur in the acquisition
of skills, classical conditioning, and a range of perceptual
priming effects. In general, amnesic patients show pre-
served performance on these implicit memory tasks
together with impaired performance on explicit or
declarative measures (Squire et al., 1993).
It is now generally accepted that recognition memory
itself reflects at least two underlying processes, one
recollective, and based on the capacity to associate the
recognized item with some aspect of its original pre-
sentation (``remembered'' items), while the other is
more automatic (``known'' items), possibly based on a
general feeling of familiarity (Tulving, 1985; Mandler,
1980). As discussed earlier, one method of accessing the
recollective component is through asking the subject to
decide whether the item they have recognized is ``re-
membered,'' or whether it is simply ``known'' to have
been presented (Gardiner & Java, 1993). A second
approach is based on requiring the subject to make a
judgement based on list membership, with instructions
and interfering lists so arranged that under certain
circumstances, a knowledge of list membership will
help, while in others, it will hinder performanceÐthe
so-called process dissociation method developed by
Jacoby (1994). There is some disagreement as to
whether the two methods are equivalent (see Baddeley,
1997, chapter 20 for a discussion). However, recent
work using electrophysiological methods has claimed
that the two approaches are both ways of accessing
contextual or source memory, and ultimately rely on
equivalent underlying processes (Rugg et al., 1998).
Baddeley, Vargha-Khadem, and Mishkin 365
Our failure to teach Jon to make the remember/
know distinction suggests that Jon may lack the capa-
city to utilize this recollective source of evidence. Yet
he appears to have no difficulty in judging whether or
not an item was presented before, presumably on the
basis of the presence or absence of a feeling of
familiarity. Evidence consistent with these notions was
obtained in an electrophysiological study that was
recently carried out on Jon (Du
È
zel, Vargha-Khadem,
Heinze, & Mishkin, 1999). This preliminary report
indicated that during recognition performance, Jon
exhibits the electrophysiological index that has been
associated with stimulus familiarity, but he appears to
lack the one that normally accompanies episodic re-
collection of the stimulus item. The latter process may
normally aid in recognition, and so its possible absence
or impairment in Jon could well be reducing his
recognition scores below those he would otherwise
have been able to achieve. It might account, for
example, for his relatively low score on overnight
recognition of the newsreel items presented only once,
the single instance in which he may have shown a
recognition deficit. Yet the absence or impairment of
episodic recollection clearly did not prevent Jon from
attaining scores that fell well within the normal range
on every other measure of recognition memory with
which he was assessed here.
The recall involved in well-established semantic mem-
ory, which may be conceived as decontextualized, like-
wise does not require, and, indeed, may not greatly
benefit from, the recollective process. By contrast, epi-
sodic recollection is clearly of major importance for the
recall of newly presented material. Thus, an impairment
in episodic recollection could explain Jon's poor perfor-
mance on standard recall tests, which are typically
assumed to rely heavily on the capacity for recollection.
Similarly, inasmuch as episodic recollection presumably
facilitates and enhances the recall of newly acquired
semantic information, an impairment in recollection
could also be the reason that Jon had difficulty in
recalling the items of information contained in the
newsreel that was shown to him only once. However,
on the newsreel that was presented four times over a 2-
day period, a procedure that decontextualized the learn-
ing experience at least with respect to time, and thereby
presumably decreased the importance of episodic re-
collection for recall, Jon's recall scores relative to those
of the control subjects improved substantially. There is
therefore good reason to suppose that just such repeti-
tion, even when limited in frequency, is what has
enabled Jon to acquire, and recall, his very large store
of semantic information.
We should end by considering the generality of the
findings we have obtained in Jon. It seems likely that
the preservation of recognition relative to recall of new
material may potentially occur more widely among
developmental cases (Vargha-Khadem et al., 1997). It
is also reported in a few cases of acquired amnesia
(Aggleton & Brown, 1999), although such cases are
highly atypical (Squire et al., 1993) and the claims of
selective preservation can be criticized on the grounds
of uncertain comparability between the recall and
recognition measures (Manns & Squire, 1999; Hamann
& Squire, 1997). However, Holdstock et al. (2000)
recently reported the case of a patient with acquired
hippocampal damage who, like Jon, shows preserved
recognition relative to recall performance on the Doors
and People Test.
Jon's capacity to acquire semantic memory is also
atypical of acquired amnesic patients (Gabrieli, Cohen,
& Corkin, 1988; Rozin, 1976). Although there does
appear to be limited evidence for preserved semantic
acquisition in some patients, the extent of such learning
is far from normal (Verfaellie, Koseff, & Alexander, 2000;
Kitchener, Hodges, & McCarthy, 1998). It is therefore
conceivable that Jon's preserved abilities are based, not
only on the selectivity of his hippocampal damage, but
also on the developmental nature of his deficit reflecting
the greater plasticity of the infant brain, together with
the possible development of a range of alternative
learning strategies as suggested by Manns and Squire
(1999). This question will not be settled until it becomes
clear whether the pattern of preserved recognition in
association with spared acquisition of semantic memory
is limited to cases of developmental amnesia, or whether
this pattern can also occur in cases where the amnesic
deficit is acquired during adulthood.
METHODS
The following tests were administered. They required
a number of test sessions distributed over a period
of weeks.
The Doors and People Test
The visual recognition test involves presenting colored
photographs of 12 doors, one at a time, each accom-
panied by an appropriate but ultimately unhelpful label,
for example ``church door'' or ``barn door'' (Baddeley
et al., 1994). Recognition is tested immediately by pre-
senting 12 successive arrays of four doors, in each case
requiring the subject to identify the door already pre-
sented. By ensuring that the items within each array all
have the same label (e.g., church doors), the role of
verbal labeling can be minimized. Two different sets of
12 doors are presented and tested in succession. The
visual recall task involves presenting and requiring the
subject to copy four crosses, each distinguished by the
features comprising the overall shape and the presence
or absence of elaboration at the crux and at the end of
the arms of the cross. Subjects are shown the four
crosses on three successive trials, each immediately
followed by a recall test. A final recall test follows after
366 Journal of Cognitive Neuroscience Volume 13, Number 3
a filled delay of approximately 10 min. Verbal recogni-
tion is tested by presenting 12 names such as John
Wilkins, which are subsequently tested by recognition
from four-item sets (e.g., John Wilby, John Wilkie, John
Wilkins, John Willis). Again, two sets of 12 items are
presented and tested immediately after each set. Finally,
verbal recall is tested by showing the subject pictures of
four people: a minister of the church, a newspaper boy,
a doctor, and a postman, and in each case, providing the
first and second names. Again, there are three presenta-
tions, each followed by recall tests of the four names,
cued by their occupations, then followed by delayed
recall after a 10-min filled interval.
Continuous Recognition Test
In a version of the test developed to study memory
under light anesthesia, each of 16 lists comprised 23
words, with two words repeated immediately, and one
each repeating at delays of 1, 2, 4, 8, and 16 inter-
polated words (Andrade et al., 1994; Andrade, 1996). In
order to increase the level of difficulty, the distractor
words used were semantically associated with the
target items. The words were read out at a rate of 2
sec per word, with the subject responding verbally
whenever he detected a repetition.
Delayed Recognition Tests
For each of these tests, 24 target words were first
presented for study on a computer visual display at a
rate of 3 sec per item (Hamann & Squire, 1997). Each list
was preceded and followed by three subsequently un-
tested buffer words in order to minimize the effects of
primacy and recency. At study, the subject read each
word aloud, and after 5 min filled by conversation,
recognition was tested. In the case of two-alternative
forced-choice recognition, the subject was tested by
being shown 24 word pairs, each comprising one item
from the study list and one new item. He was asked to
say aloud the word that he had been shown previously.
In the case of the yes±no recognition test, a total of 24
old (studied) words and 24 new (unstudied) words were
presented one at a time in mixed order; in each case, the
subject was asked to say whether he had been shown
that word previously or not. A total of six test lists were
given for each of the two conditions, beginning with
three forced-choice lists, followed by three yes±no lists,
followed by the remaining three forced-choice lists, and
finally, the last three yes±no lists. The appropriate
recognition instructions were given before each test list.
Newsreel Recognition and Recall Task
The test was based on presenting video clips based on
excerpts from the Pathe Movie newsreels from 1937 and
1957, respectively. In each case, an initial section of
approximately 35 min was presented. Testing took place
over three successive days. The 1937 newsreel was pre-
sented on the morning and early afternoon of the first day,
and the morning and late afternoon of the second. The
1957 newsreel was presented at the end of the afternoon
of the first day. The 1957 newsreel was tested immediately
after this single presentation and at the beginning of the
next morning session, approximately 18 hr later. The 1937
newsreel was tested after its fourth presentation and
retested on the morning of the third day, again after an
18-hr delay. The two newsreel tests each involved a total
of 45 questions, each comprising a recall test, followed by
a four-alternative forced-choice recognition test. Ques-
tions ranged from specific detail such as the names of
people shown and numbers of people killed in a parti-
cular accident, to more general issues mentioned, such as
the state of the economy at a particular time and the cause
of particular accidents. Virtually all the information ap-
peared to be novel for all subjects.
Acknowledgments
We thank Jon, his parents, and his friends Andrew and Max, for
their unfailing cooperation. We are grateful to Jackie Andrade,
Larry Squire, and Alan Parkin for generously providing details
of materials and procedures, and to Kate Watkins, Ingrid
Johnsrude, and Louise Parry for their technical assistance. Alan
Baddeley acknowledges the support of grant G9423916 from
the Medical Research Council.
Reprint requests should be sent to: Alan Baddeley, Department
of Experimental Psychology, University of Bristol, 8 Woodland
Road, Bristol BS8 1TN UK. Tel.: +44-117-928-8541; Fax: +44-
117-928-8562; e-mail: alan.baddeley@bristol.ac.uk.
REFERENCES
Aggleton, J. P., & Brown, M. W. (1999). Episodic memory,
amnesia, and the hippocampal±anterior thalamic axis. Be-
havioral and Brain Sciences, 2, 425±490.
Aggleton, J. P., & Shaw, C. (1996). Amnesia and recognition
memory: A reanalysis of psychometric data. Neuropsycho-
logia, 34, 51±62.
Andrade, J. (1996). Investigations of hypesthesia: Using
anaesthetics to explore relationships between conscious-
ness, learning and memory. Consciousness and Cognition,
5, 562±580.
Andrade, J., Munglani, R., Jones, J. G., & Baddeley, A. D. (1994).
Cognitive performance during anaesthesia. Consciousness
and Cognition, 3, 148±165.
Baddeley, A. D. (1993). Short-term phonological memory and
long-term learning: A single case study. European Journal of
Cognitive Psychology, 5, 129±148.
Baddeley, A. D. (1996). Applying the psychology of memory to
clinical problems. In D. Herrmann, C. McEvoy, C. Hertzog,
P. Hertel, & M. K. Johnson (Eds.), Basic and applied
memory research: I (pp. 195±220). Mahwah, NJ: Erlbaum.
Baddeley, A. D. (1997). Human memory: Theory and Practice.
(Revised Ed.). Hove, UK: Psychology Press.
Baddeley, A. D., Emslie, H., & Nimmo-Smith, I. (1993). The
Spot-the-Word Test: A robust estimate of verbal intelligence
based on lexical decision. British Journal of Clinical Psy-
chology, 32, 55±65.
Baddeley, Vargha-Khadem, and Mishkin 367
Baddeley, A. D., Emslie, H., & Nimmo-Smith, I. (1994). The
Doors and People Test. Bury St. Edmunds, UK: Thames
Valley Test Company.
Baddeley, A. D., & Warrington, E. K. (1970). Amnesia and the
distinction between long- and short-term memory. Journal
of Verbal Learning and Verbal Behavior, 9, 176±189.
Calev, A. (1984). Recall and recognition in chronic nonde-
mented schizophrenics. Journal of Abnormal Psychology,
93, 172±177.
Delis, D. C., Kramer, J. H., Kaplan, E., & Ober, B. A. (1987).
California Verbal Learning Test: Form II. San Antonio, TX:
Psychological Corporation.
DuSoir, H., Kapur, N., Byrnes, D. P., McKinstry, S., & Hoare,
R. D. (1990). The role of diencephalic pathology in human
memory disorder. Brain, 113, 1695±1706.
Dy
È
zel, E., Vargha-Khadem, F., Heinze, H. J., & Mishkin, M.
(1999). ERP evidence for recognition without episodic re-
collection in a patient with early hippocampal pathology.
Society for Neurosciences Abstracts, 25, 259.11.
Gabrieli, J. D. E., Cohen, N. J., & Corkin, S. (1988). The im-
paired learning of semantic knowledge following bilateral
medial temporal-lobe resection. Brain and Cognition, 7,
157±177.
Gadian, D. G., Aicardi, J., Watkins, K. E., Porter, D. A., Mishkin,
M., & Vargha-Khadem, F. (2000). Developmental amnesia
associated with early hypoxic±ischaemic injury. Brain, 123,
499±507
Gardiner, J. M., & Java, R. I. (1993). Recognising and remem-
bering. In A. F. Collins, S. E. Gathercole, M. A. Conway, & P.
E. Morris (Eds.), Theories of memory (pp. 163±188). Hove,
UK: Erlbaum.
Greene, J. D. W., Baddeley, A. D., & Hodges, J. R. (1996).
Analysis of the episodic memory deficit in early Alzheimer's
Disease: Evidence from the Doors and People Test. Neu-
ropsychologia, 34, 537±551.
Haist, F., Shimamura, A. P., & Squire, L. R. (1992). On the
relationship between recall and recogition memory. Journal
of Experimental Psychology: Learning , Memory, and Cog-
nition, 18, 691±702.
Hamann, S. B., & Squire, L. R. (1997). Intact perceptual mem-
ory in the absence of conscious memory. Behavioural
Neuroscience, 111, 850±854.
Hanley, J. R., Davies, A. D. M., Downes, J. J., & Mayes, A. R.
(1994). Impaired recall of verbal material following rupture
and repair of an anterior communicating artery aneurysm.
Cognitive Neuropsychology, 11, 543±578.
Hirst, W., Johnson, M. K., Kim, J. K., Phelps, E. A., Risse, G., &
Volpe, B. T. (1986). Recognition and recall in amnesics.
Journal of Experimental Psychology, 12, 445±451.
Holdstock, J. S., Mayes, A. R., Cezayirli, E., Isaac, C. L., Aggleton,
J. P., & Roberts, N. (2000). A comparison of egocentric and
allocentric spatial memory in a patient with selective hip-
pocampal damage. Neuropsychologia, 38, 410±425.
Howard, D., & Patterson, K. (1992). Pyramid and palm trees:
A test of semantic access from pictures and from words.
Bury St. Edmunds, UK: Thames Valley Test Company.
Huppert, F. A., & Piercy, M. (1979). Normal and abnormal
forgetting in amnesia: Effect of locus of lesion. Cortex, 15,
385±390.
Jacoby, L. L. (1994). Measuring recollection: Strategic versus
automatic influences of associative context. In C. Umilta &
M. Moscovitch (Eds.), Attention and performance: XV.
Conscious and non-conscious information processing (pp.
661±680). Cambridge: MIT Press.
Kitchener, E. G., Hodges, J. R., & McCarthy, R. (1998).
Acquisition of post-morbid vocabulary and semantic
facts in the absence of episodic memory. Brain, 121,
1313±1327.
Mandler, G. (1980). Recognizing: The judgement of previous
occurrence. Psychological Review, 87, 252±271.
Mandler, G., Pearlstone, Z., & Koopmans, H. S. (1969). Effect of
organization and semantic similarity on recall and recogni-
tion. Journal of Verbal Learning and Verbal Behavior, 8,
410±423.
Manns, J. R., & Squire, L. R. (1999). Impaired recognition
memory on the Doors and People Test after damage
limited to the hippocampal region. Hippoccampus, 9,
495±499.
McMackin, D., Cockburn, J., Anslow, P., & Gaffan, D. (1995).
Correlation of fornix damage with memory impairment in six
cases of colloid cyst removal. Acta Neurochirugica, 135,
12±18.
Milner, B. (1963). Effect of different brain lesions on card
sorting. Archives of Neurology, 9, 90±100.
Milner, B. (1966). Amnesia following operation on the tem-
poral lobes. In C. W. M. Whitty & O. L. Zangwill (Eds.),
Amnesia (pp. 109±133). London: Butterworth.
Mishkin, M., Suzuki, W., Gadian, D. G., & Vargha-Khadem, F.
(1997). Hierarchical organization of cognitive memory. Phi-
losophical Transactions of the Royal Society of London,
Series B, 352, 1461±1467.
Mishkin, M., Vargha-Khadem, F., & Gadian, D. G. (1998). Am-
nesia and the organization of the hippocampal system.
Hippocampus, 8, 212±216.
Morris, R. G., Abrahams, S., Baddeley, A. D., & Polkey, C. E.
(1995). Doors and People: Visual and verbal memory fol-
lowing unilateral temporal lobectomy. Neuropsychology, 9,
464±469.
Parkin, A. J., Dunn, J. C., Lee, C., O'Hara, P. F., & Nussbaum, L.
(1993). Neuropsychological sequeli of Wernicke's ence-
phalopathy in a twenty-year-old woman: Selective impair-
ment of a frontal memory system. Brain and Cognition,
21, 1±19.
Parkin, A. J., & Hunkin, N. M. (1993). Impaired temporal
context memory on anterograde but not retrograde
tests in the absence of frontal pathology. Cortex, 29,
267±280.
Parkin, A. J., Leng, N. R. C., & Hunkin, N. M. (1990). Differential
sensitivity to context in diencephalic and temporal lobe
amnesia. Cortex, 26, 373±380.
Parkin, A. J., & Walter, B. M. (1992). Recollective experience,
normal aging and frontal dysfunction. Psychology and Aging,
7, 290±298.
Reed, J. M., & Squire, L. R. (1997). Impaired recognition
memory in patients with lesions limited to the hippo-
campal formation. Behavioral Neuroscience, 111,
667±675.
Rozin, M. (1976). The psychobiological approach to human
memory. In M. R. Rosenzweig & E. L. Bennet (Eds.), Neural
mechanisms of learning and memory (pp. 3±46). Cam-
bridge: MIT Press.
Rugg, M. D., Schloerscheidt, A. M., & Mark, R. E. (1998). An
electrophysiological comparison of two indices of recollec-
tion. Journal of Memory and Language, 39, 47±69.
Schacter, D. L., & Tulving, E. (1994). Memory systems. Cam-
bridge: MIT Press.
Schonfield, D., & Robertson, B. A. (1966). Memory storage and
aging. Canadian Journal of Psychology, 20, 228±231.
Shallice, T. (1988). From neuropsychology to mental struc-
ture. Cambridge: Cambridge University Press.
Shepard, R. N., & Teghtsoonian, M. (1961). Retention of in-
formation under conditions approaching a steady state.
Journal of Experimental Psychology, 62, 302±309.
Squire, L. R. (1992). Declarative and non-declarative memory:
Multiple brain systems supporting learning and memory.
Journal of Cognitive Neuroscience, 4, 232±243.
368 Journal of Cognitive Neuroscience Volume 13, Number 3
Squire, L. R., Knowlton, B., & Musen, G. (1993). The structure
and organisation of memory. Annual Review of Psychology,
44, 453±495.
Squire, L. R., & Zola, S. M. (1998). Episodic memory, semantic
memory and amnesia. Hippocampus, 8, 205±211.
Tulving, E. (1972). Episodic and semantic memory. In E. Tul-
ving & W. Donaldson (Eds.), Organization of memory
(pp. 381±403). New York: Academic Press.
Tulving, E. (1985). Memory and consciousness. Canadian
Psychologist, 26, 1±12.
Tulving, E., & Markovitch, H. J. (1998). Episodic and declarative
memory: The role of the hippocampus. Hippocampus, 8,
198±204.
Vargha-Khadem, F., Gadian, D. G., Watkins, K. E., Connelly, A.,
Van Paesschen, W., & Mishkin, M. (1997). Differential effects
of early hippocampal pathology on episodic and semantic
memory. Science, 277, 376±380.
Verfaellie, M., Koseff, P., & Alexander, M. P. (2000). Acquisition
of novel semantic information in amnesia: Effects of lesion
location. Neuropsychologia, 38, 484±492.
Volpe, B. T., Holtzman, J. D., & Hirst, W. (1986). Further
characterization of patients with amnesia after cardiac arrest:
Preserved recognition memory. Neurology, 36, 408±411.
Warrington, E. K. (1984). Recognition memory test. Windsor,
England: Nelson.
Whittlesea, B. W. A. (1993). Illusions of familiarity. Journal of
Experimental Psychology: Learning, Memory, and Cogni-
tion, 19, 1235±1253.
Wilson, B. A., Alderman, N., Burgess, P., Emslie, H., & Evans,
J. J. (1996) Behavioural assessment of the dysexecutive
syndrome. Bury St. Edmunds, UK: Thames Valley Test
Company.
Wilson, B. A., Clare, L., Cockburn, J. M., Baddeley, A. D., Tate,
R. M., & Watson, P. (1999). The Rivermead Behavioural
Memory TestÐExtended Version. Bury St. Edmunds, UK:
Thames Valley Test Company.
Baddeley, Vargha-Khadem, and Mishkin 369
