Neurocase (2002) Vol. 8, pp. 442–452

© Oxford University Press 2002

Contribution of the Left Dorsomedial Thalamus to
Recognition Memory: a Neuropsychological Case Study
N. M. J. Edelstyn, S. J. Ellis1, P. Jenkinson and A. Sawyer
Department of Psychology, University of Keele, Keele and 1Department of Neurology, Royal Infirmary, Stoke-on-Trent,
Staffordshire, UK

Abstract
This study reports a patient with a unilateral left thalamic lesion which was centred on the dorsomedial thalamic
nucleus. Cognitive neuropsychological assessment revealed a severe impairment in verbal memory and symptoms of
executive dysfunction, in the presence of relatively intact visual and facial recognition, working memory, praxis,
language and IQ. Verbal and visual recognition memory were investigated using the remember–know paradigm. The
results indicated a profound impairment in recollection-driven verbal recognition memory. These results are discussed
in the context of the role of the dorsomedial thalamic nucleus in recognition memory, and functional models of memory.

Introduction
Memory impairment is a well-documented sequela of thalamic damage. Traditionally, diencephalic amnesia has been
characterized by deficits in encoding and recollection (both
free recall and recognition), resulting in a severe anterograde
episodic memory impairment in the presence of preserved
IQ, motor learning, perceptual functions and short-term
memory (e.g. Von Cramon et al., 1985; Weiskrantz, 1985;
Graff-Radford et al., 1990; Aggleton and Saunders, 1997;
Squire and Zola, 1997). However, most of the evidence
linking thalamic pathology to memory deficits has been
reported in patients with alcoholic Wernicke-Korsakoff Syndrome (e.g. Victor et al., 1989). Many of these patients had
large lesions, which extended well beyond individual thalamic
nuclei, and often included the frontal lobes (Jacobson, 1989).
The lack of circumscribed neuropathology coupled with
evidence of widespread neuronal damage, which results
from prolonged exposure to alcohol (Okazaki, 1989), have
obscured which thalamic structures are critical for the development of diencephalic amnesia (Graff-Radford et al., 1990;
Mennemeier et al., 1992).
More recently, an attempt has been made to reformulate
the anatomy underlying anterograde amnesia and in so doing
it has been claimed that the distinction between temporal
lobe and diencephalic amnesia is of limited value (Aggleton
and Brown, 1999). The main argument for this is based on
the premise that diencephalic and medial temporal lobe
anterograde amnesia arise from damage to a common
‘extended hippocampal’ memory system, which is comprised
of reciprocal connections between the hippocampus, the

mamillary bodies and the anterior thalamic nucleus, via the
fornix (Aggleton and Shaw, 1996; Aggleton and Saunders,
1997; Aggleton and Brown, 1999). Aggleton and Brown
proposed that this extended hippocampal–diencephalic system was critical for the encoding and subsequent recall of
contextual information about a previous episode. Furthermore, they suggested that this memory system was not
necessary for efficient recognition. This proposal builds on
earlier suggestions that recognition memory arises out of
two functionally independent components, i.e. contextual
recollection and familiarity detection (Mandler, 1980; Jacoby
and Dallas, 1981). Aggleton and Brown claimed that contextual recollection was dependent on the extended hippocampal
system, and that a second non-hippocampal system, which
extended from the perirhinal cortex to the dorsomedial
thalamic (DMT) nucleus and the prefrontal lobes, subserved
the ability to evaluate the contextual memory strength, or
familiarity, of an item.
However, a number of authors have claimed that Aggleton
and Brown’s anatomical model is incompatible with empirical
evidence reported from behavioural studies of rats (Mair
et al., 1999) and monkeys (Parker, 1999) as well as patients
(Mayes et al., 1999; Van der Warf et al., 1999; Witter and
Van der Warf, 1999; Hanley et al., 2001). For instance, Witter
and Van der Warf (1999) described a patient who displayed
evidence of impaired familiarity-but intact recollection-driven
recognition in the context of a thalamic lesion which spared
the DMT. While not disputing the neuroanatomical separation
of recall from familiarity, the authors claimed that it was

Correspondence to: N. M. J. Edelstyn, Department of Psychology, University of Keele, Keele, Staffordshire ST5 5BG, UK. Tel: ⫹44 (0)1782 584318;
Fax: ⫹44 (0)1782 583387; e-mail: n.edelstyn@psy.keele.ac.uk

Contribution of the left DMT to recognition memory 443

projections from the perirhinal cortex to the intralaminar
nuclei, rather than the DMT, which subserved familiaritydriven recognition. From a different perspective, the case
reported by Hanley et al. (2001) also questioned the validity of
Aggleton and Brown’s claims. Their patient, ROB, exhibited a
severe deficit in free recall of verbal material following a
lesion which damaged the fornix and the anterior thalamic
nucleus. However, there was no evidence of impaired recollection-driven recognition memory. The authors acknowledged that damage to the extended hippocampal circuit had
a major impact on recall relative to recognition, but their
findings did not support the suggestion that the extended
hippocampal system was necessary for recognition memory
decisions which were based on contextual recollection.
Recognition memory has also been investigated using
signal detection models (Yonelinas, 1994; Donaldson, 1996)
Typically, such studies make use of Tulving’s (1985) remember–know paradigm, which requires subjects not only to
identify target stimuli from distracters during recognition
testing, but also to select the basis of their decisions in terms
of either explicit retrieval (i.e. remember) or feelings of
familiarity (i.e. know). Remember–know studies have led to
a number of dissociations between retrieval and familiarity
responses being identified. For instance, word frequency and
levels of processing have been shown to influence selectively
remember judgements (e.g. Gardiner and Java, 1990;
Rajaram, 1993); whereas conceptual fluency only affects
know responses (Rajaram and Geraci, 2000).
Reports of amnesic patients with widespread damage of
the medial temporal lobe using the remember–know paradigm
have yielded inconsistent results. Studies by Verfaellie and
Treadwell (1993), Kroll et al. (1996) and Schacter et al.
(1996) found that amnesics were more likely to use know
responses, based on feelings of familiarity, than remember
judgements. Whereas a later report by Knowlton and Squire
(1995) claimed that amnesics were less likely to make know
responses and concluded that familiarity responses were
disrupted.
Both Donaldson’s and Yonelinas’ signal detection models
allow for a dissociation between the two component processes
of recognition. However, they differ in terms of their theoretical predictions. Donaldson proposes that a single underlying
memory system based on a continuum of familiarity, subserves both know and remember judgements. It assumes that
subjects cannot directly determine whether an item is familiar
or not; rather individuals are only aware of an item’s level
of familiarity. Therefore, in order to label an item familiar
or unfamiliar, a criterion value of familiarity must be set,
such that items which exceed it are labelled familiar and
those which do not are labelled as unfamiliar. A second more
stringent criterion value of familiarity is used to determine
whether a target item accesses stored recollective information
or not. Items whose level of familiarity is sufficiently strong
to exceed this second criterion result in the retrieval of
contextualizing information and are labelled remember, while
those which fail to do so are labelled know. In contrast,

Yonelinas suggested that familiarity and retrieval are best
accounted for by a dual-process model, in which know
responses are familiarity driven, whereas remember judgements are dependent on a separate non-familiarity-based
mechanism operating in parallel (Yonelinas et al., 1996;
Dobbins et al., 2000).
Lesions of the thalamus not only give rise to memory
disorders, but can also lead to a disruption of prefrontal
processes [see review by Van der Warf et al. (2000)].
Extensive reciprocal projections exist between the DMT and
ipsilateral medial and orbital regions of the prefrontal cortex
(Bachevalier and Mishkin, 1986; see also Aggleton and
Brown, 1999). Thus, a lesion centred on the DMT is likely
to cause widespread dysfunction within prefrontal areas,
leading to disturbances of executive abilities, attention, initiation, inhibition and temporal organization of behaviour [see
review by Van der Warf et al. (2000)]. It is known from studies
of animals and patients that frontal/temporal interaction is
essential for the encoding of novel items (Parker et al., 1988)
and in key executive memory functions involved in strategic
memory search, monitoring and decision making (Shallice,
1980; Squire, 1981; Kapur et al., 1995; Markowitsch, 1995;
Parker and Gaffan, 1997, 1998; Parker et al., 1997; Rapcsak
et al., 1999; Ward et al., 1999). For instance, in normal
subjects, when exposed to an unfamiliar face which elicits
false feelings of familiarity because of certain physical
similarities to known faces, this uncertainty may be resolved
by attempts to retrieve more detailed contextual information.
Patients with damage to their prefrontal lobes may not
spontaneously engage in this effortful search, and therefore
produce high levels of false recognition, which is evident in
an elevated false alarm rate (Rapcsak et al., 1999).
Thus, memory problems may arise either as a direct
consequence of prefrontal damage, or secondarily as a result
of prefrontal cognitive deficits such as inattention. Memory
problems due to the latter rather than a disruption to memory
processes per se are evident in terms of an impairment in
the active recall from the memory store in the presence
of spared recognition or aided retrieval (Van der Warf
et al., 2000).
The current study sought to examine the effects of DMT
pathology on retrieval- and familiarity-driven recognition in
a single case study. The specific aims of the study were, first,
to characterize the patient’s memory deficits in terms of
familiarity- and retrieval-driven recognition using the remember–know paradigm. Second, by manipulating the level of
target familiarity within the recognition memory system, to
determine whether the observed profile was more readily
accounted for by the single or dual memory system signal
detection models.

Methods
Subjects
Patient. QX, a 55-year-old, right-handed male university
lecturer, woke to find that he was ‘talking rubbish’ and had

444 N. M. J. Edelstyn et al.

slurred speech. He noted that his right arm felt heavy, but
this returned to normal. For 2 days he was very confused.
QX developed disinhibited behaviour, reduced motivation,
memory problems and apraxia. For instance, he was unable
to return to his hobby of bell ringing because he could not
remember the complex sequences of actions involved. A
magnetic resonance imaging scan revealed a discrete lacuna
infarction involving the DMT of the left thalamus, which
extended superolaterally and involved the medial part of the
ventrolateral nucleus (see Fig. 1).
There was an incidental minor area of ischaemia in the
subcortical white matter of the left parietal lobe (Fig. 1b).
The patient’s initials have been changed to preserve
anonymity and informed consent was obtained.
Neuropsychological profile. A detailed neuropsychological
profile is presented in Table 1.
QX’s current IQ [Revised Wechsler Adult Intelligence
Scale (WAIS-R); Wechsler, 1981] was placed in the high
average range. This score indicates a slight decline in general
intellectual functioning from premorbid levels which placed
him in the superior range [National Adult Reading Test
(NART); Nelson, 1982]. There was a large discrepancy
between his high score on the verbal IQ subtests which placed
him in the superior range, and scoring on the performance IQ
subtests, which fell within the low average range. QX was
particularly poor on block design, object assembly and digit
symbol. Successful completion of block design and object
assembly requires orderliness and planning, and are notable
in that they do not lend themselves to verbalization (Lezak,
1995). Digit symbol is a non-specific indicator of brain
dysfunction, in that it involves the integration of many
different components, which include visuomotor processing,
visual search and shifting (Lezak, 1995).
QX displayed little difficulty in successfully completing
the facial and visual processing tasks. He was able (1) to
sort a pack of 50 famous faces drawn from a variety of
different ‘professions’ (including sport, politics, royalty, film
and television) from 50 unfamiliar faces and then (2) to name
the ‘famous’ faces. Similarly, he had no problems naming
32 animate (16 animals and 16 fruit and vegetables) and 32
inanimate (drawn from a variety of categories including work
tools, kitchen implements, vehicles, etc) line drawings. He
was able to mime the use (i.e. function) of inanimate objects
and to identify another object with which it would be
associated (association: e.g. a target item ‘cup’ would be
associated with a ‘saucer’). While for the animate exemplars
he was able to label their superordinate category (e.g. for
‘apple’ it would be ‘fruit’, or for ‘dog’ it would be ‘animal’).
Formal tests of language function [Psycholinguistic Assessments of Language Processing in Aphasia (PALPA); Kay
et al., 1992] failed to elicit the word-finding difficulties
reported during casual conversation. Similarly, there was no
evidence of buccofacial, ideomotor or constructional apraxia.
Despite the reported personality and motivational changes,
no deficits were evident on the tasks which tap executive

Fig. 1. Magnetic resonance imaging scans of QX showing the extent of the
left thalamic damage. (a) T2-weighted magnetic resonance scan demonstrating
the lacuna infarction in the inferior medial part of the left thalamus. (b) T2weighted magnetic resonance scan demonstrating the lacuna infarction in the
superior lateral part of the left thalamus and a minor area of ischaemic change
in the parietal white matter. (c) T2-weighted coronal magnetic resonance scan
demonstrating the ischaemic lacuna infarct in the left thalamus involving
inferior medial structures and more lateral structures superiorly.

Contribution of the left DMT to recognition memory 445
Table 1. Neuropsychological profile of QX
Neuropsychological test

QX

Percentile/range

NART
Full-scale IQ

124

Superior

WAIS-R
Information
Digit symbol
Vocabulary
Arithmetic
Comprehension
Similarities
Picture completion
Picture arrangement
Block design
Object assembly
Digit symbol
Verbal IQ
Performance IQ
Full-scale IQ

13a
12
12
14
16
12
10
12
8
8
5
124
88
119

84th percentile
75th percentile
75th percentile
91st percentile
98th percentile
75th percentile
50th percentile
75th percentile
25th percentile
25th percentile
5th percentile
Superior
Low average
High average

Stroop
Word
Colour

112/112b
99/112

100th percentile
52nd percentile

WCST
Categories
Failure to maintain set
Perseverative errors
Fluency (FAS)

6b
0
6
35c

⬎16th percentile
⬎16th percentile
55th percentile
32

Facial processing
Un/familiar face sorting
Familiar face naming

50/50b
49/50

NAD
NAD

Visual processing
Function (inanimate)
Association (inanimate)
Naming (animate and inanimate)

64/64b
64/64
64/64

NAD
NAD
NAD

WMS
Auditory delayed recognition
Auditory delayed
Visual delayed
Auditory immediate
Visual immediate
General memory
Immediate memory
Working memory

6a
9
17
12
12
32
24
18

9th percentile
1st percentile
27th percentile
6th percentile
5th percentile
5th percentile
3rd percentile
32nd percentile

Rey Complex Figure
Copy
Immediate recall
Delayed recall

30/36b
9.5/36
11.5/36

NAD
⬍1st percentile
⬍1st percentile

20/21b
8/9

16–21
6–9

19/21
9/9

17–21
7–9

14/21
6/9
53/63
23/27

19–21
7–9
54–63
19–27

AMI
Childhood
Personal semantic
Autobiographical incidents
Early adult life
Personal semantic
Autobiographical incidents
Recent life
Personal semantic
Autobiographical incidents
Personal semantic
Autobiographical incidents

NART, National Adult Reading Test; WAIS-R, Wechsler Adult Intelligence
Scale; WMS, Wechsler Memory Scale; AMI, Autobiographical Memory
Inventory; WCST, Wisconsin Card Sorting Test; NAD, no abnormalities
detected.
aScaled score.
bRaw score.
cMean score.

functions [Stroop, 1989; Wisconsin Card Sorting Test
(WCST), Heaton et al., 1989; and verbal fluency].
Assessment of memory revealed a complex pattern of
preserved and deficient abilities [Autobiographical Memory
Inventory (AMI), Kopelman et al., 1990; Rey Complex
Figure, Myers and Myers, 1995; Wechsler Memory Scales
(WMS), Wechsler, 1997]. Working memory (WAIS-R, WMS)
and autobiographical memory (AMI) for childhood and early
adult life were intact. However, more recent memories for
incidents and personal semantic information showed signs of
impairment. Assessment of immediate and delayed visual
and auditory memory (Rey Complex Figure, WMS) showed
differential rates of impairment. Whereas auditory memory
was impaired on all assessments, delayed visual memory
appeared intact on the WMS but deficient on the Rey
Complex Figure. This discrepancy may be accounted for in
terms of task demands: whereas the WMS delayed visual
memory subtest does not involve a visuomotor/visuoconstructive component, the Rey Complex Figure does (refer to
previous outline of WAIS-R performance IQ subtests).
Controls. The performance of two groups of normal
controls are reported: a larger cohort of 29 young psychology
undergraduate and A level student volunteers (NC[y]) aged
between 16 and 29 years (mean 19.5 years) and a smaller
group (NC[o]) of five older age-matched (range 41–49 years,
mean 45.6 years) university employees. All subjects were
right-handed, had English as their first language, had normal/
corrected visual acuity and no reported dyslexia, neurological
or psychiatric history.

Anterograde memory experiments
Procedure. Recognition memory for words and faces was
assessed separately, using the same procedure. A standard
recognition memory format was used, which consisted of a
study phase, during which time the target stimuli were
presented, and a recognition phase, when target stimuli were
discriminated from novel distracters.
A total of four recognition memory tasks were administered: two facial recognition memory (FRM) tasks and two
verbal recognition memory (VRM) tasks. Both FRM tasks
and VRM tasks were matched in all respects apart from the
duration of target exposure during the study phase. In one
version, the facial and verbal targets were exposed for
3000 ms and in the other version it was 6000 ms.
Visual stimuli were presented in a program written in
Microsoft Visual C⫹⫹ using MFC on a personal computer
running Windows 95. The images presented in the FRM
tasks consisted of 100 black and white photographs of
unfamiliar male faces (images extended 8 cm horizontally
and 9 cm vertically, and viewed at a visual angle of 9°). The
words used in the VRM tasks were a mixture of 50 highand low-frequency unfamiliar adjectives (mean 137, range
2–2035). All stimuli were viewed one at a time in the centre
of the screen.

446 N. M. J. Edelstyn et al.

Fig. 2. Hit and false alarm rates in experiment 1.

Fifty target items were presented during the study phase.
To help concentration, the subjects were instructed to state
whether the item (face or word) was pleasant or not. During
the recognition phase, which immediately followed completion of the study phase, the studied targets were mixed with
50 unfamiliar distracters. The subject’s task was to identify
the target items from the distracters. This process was
achieved by pressing one of two hand-held buttons: pressing
the button held in the right hand indicated a ‘yes’ response
for a positive recognition (i.e. an old item) while pressing
the left-hand button indicated a ‘no’ response to an unrecognized item (i.e. a new item).
The subjects had a 3000 ms window in which to register
their choice before the next stimulus item appeared. The
participants were instructed to respond as quickly and as
accurately as possible. The subjects were familiarized with
the experimental set-up using a practice test prior to starting
the experimental tasks. The presentation order of the recognition memory tasks (verbal versus facial; 3000 versus 6000 ms)
was counterbalanced across subjects.
Correct identification of an old item was defined as a hit,
while false recognition of a distracter was termed a false
alarm. Following each positive identification, irrespective of
whether it was a target hit or a false alarm, the subjects had
to decide the basis of their decision in terms of whether the
item elicited (a) feelings of familiarity (i.e. a know response)
or (b) a specific retrieval of the item having been presented

in the study phase (i.e. a remember judgement). A third
category, (c) guess, which was based on a superordinate
category decision, was included during a pilot study but
dropped for the reported tasks because the subjects failed to
make use of it. The selection of remember–know choices
was not time constrained and was achieved using the handheld buttons.

Results
Experiment 1. Short duration (3000 ms)
recognition memory tasks
Hit rates, defined as the conditional probability of responding
yes to an old item, and false alarm rates, defined as the
conditional probability of responding yes to a new item, were
calculated [number of hits (or false alarms) ⫹ 0.5/number
of targets (or distracters) ⫹ 1] for both recognition memory
tasks and are presented in Table 2 and Fig. 2.
In this and subsequent analyses, QX’s performance was
considered to be significantly impaired if his score was 2.0
standard deviations or more away from either of the two
normal control group means. The two control groups differed
in terms of their means and standard deviations, such that,
on occasion, QX’s score fell outside the normal range of one
group but not the other. Such cases are highlighted and may
be interpreted in terms of the differences in numbers and

Contribution of the left DMT to recognition memory 447
Table 2. Data summary table for QX and both normal control groups for experiments 1 and 2
Normal controls [old]

Experiment 1: 3000 ms exposure
Hit rate/words
False alarm rate/words
Hit rate/faces
False alarm rate/faces
Remember judgements/words
Know judgements/words
Remember judgements/faces
Know judgements/faces
Discrimination accuracy/words
Discrimination accuracy/faces
Response bias/words
Response bias/faces
Experiment 2: 6000 ms exposure
Hit rate/words
False alarm rate/words
Hit rate/faces
False alarm rate/faces
Remember judgements/words
Know judgements/words
Remember judgements/faces
Know judgements/faces
Discrimination accuracy/words
Discrimination accuracy/faces
Response bias/words
Response bias/faces

Normal controls [young]

QX

Mean

1 standard deviation Normal rangea

Mean

1 standard deviation Normal rangea

0.52
0.28
0.32
0.09
0
26
3
13
0.7
0.74
0.41
0.91

0.9
0.1
0.83
0.19
39
7
28
14
0.93
0.89
–0.2
–0.07

0.06
0.05
0.06
0.06
6.6
5.4
8.4
9.7
0.03
0.03
0.33
0.29

0.78–1
0–0.20
0.71–0.95
0.07–0.31
26–50
0–17
11–45
0–33
0.87–0.99
0.83–0.95
–0.86–0.46
–0.65–0.51

0.6
0.09
0.6
0.16
24
11
16
14
0.88
0.82
0.57
0.58

0.17
0.08
0.14
0.11
10.7
6.6
8.9
7.1
0.07
0.06
0.37
0.35

0.26–0.94
0–0.25
0.32–0.88
0–0.38
2–45
0–24
0–34
0–28
0.74–1
0.7–0.94
–0.17–1
–0.12–1

0.48
0.28
0.95
0.07
1
23
29
19
0.82
0.98
0.87
–0.17

0.9
0.08
0.9
0.21
34
12
30
15
0.95
0.91
0.11
–0.35

0.05
0.02
0.06
0.12
9.8
8.7
3.4
3.7
0.02
0.04
0.19
0.04

0.80–1
0.04–0.12
0.78–1
0–0.45
14–50
0–29
23–37
8–22
0.91–0.99
0.83–0.99
–0.03–0.41
–0.43– –0.27

0.88
0.07
0.74
0.17
35
7
23
15
0.95
0.87
0.41
–0.31

0.11
0.12
0.11
0.14
11.9
5.4
6.2
4.7
0.04
0.04
0.59
0.55

0.66–1
0–0.31
0.52–0.96
0–0.48
11–50
0–19
11–35
6–24
0.87–1
0.79–0.95
–0.67–1
–1–0.79

QX’s score is in bold type when it falls outside the normal range.
aMean ⫾ 2 standard deviations.

educational background between the two cohorts. QX displayed evidence of both VRM and FRM impairment, which
was apparent in terms of a significantly elevated false alarm
rate for words and a reduced hit rate for faces.
A breakdown of hit rate in terms of remember–know
judgements for the VRM and FRM tasks is presented in
Table 2 and Fig. 3. Both normal control groups recorded
higher numbers of remember responses relative to know
responses. QX, however, displayed significant impairment in
terms of his paucity of remember judgements, which were
compensated for by a significantly elevated number of know
responses. The number of remember judgements was also
significantly reduced in the FRM task when his performance
was compared with the NC[o] group only.
Subsequent analyses converted hit rates and false alarm
rates into the non-parametric signal detection indices of A⬘
(Snodgrass and Corwin, 1988) and B⬙D (Donaldson, 1992)
which provided measures of discrimination accuracy and
response bias, respectively (see Table 2). B⬙D values of 0
indicate a neutral response bias, i.e. when in an uncertain
state a yes or no response is registered with equal probability,
whereas positive (B⬙D ⫽ 0.1–1) and negative
(B⬙D ⫽ –0.1 to –1) values indicate conservative and liberal
bias, respectively. A conservative bias is defined by the
increased probability of no responses when in an uncertain
state and a liberal bias is defined by the increased probability
of yes judgements.

QX’s discrimination accuracy on the VRM task was
significantly impaired relative to both normal control groups’
normal ranges, and on the FRM task compared with the
NC[o] set. Both normal control sets varied in terms of their
response bias. Whereas the NC[o] group tended to exhibit a
slightly liberal response bias on both recognition memory
tasks, the NC[y] group displayed a more conservative one.
QX also exhibited a conservative bias, which fell outside the
NC[o] normal range for the FRM task only.

Experiment 2. Long duration (6000 ms) recognition
memory tasks
Extending target exposure during the study phase facilitated
VRM for the NC[y] group; this was evident in an increased
hit rate and a reduced false alarm rate (see Table 2, Fig. 4).
The NC[o] group showed little change from their already
high level of performance on the shorter target exposure
version. QX’s VRM, however, remained significantly
impaired. Both his hit rate and his false alarm rate fell outside
the normal range of both control groups. In contrast, a strong
facilitatory effect was evident during the facial version, with
both his hit rate and his false alarm rate falling well within
the normal range of both control groups.
A breakdown of hit rate in terms of remember–know
judgements for both recognition memory tasks is presented
in Table 2 and Fig. 5.

448 N. M. J. Edelstyn et al.

Fig. 3. Remember and know judgements in experiment 1.

Fig. 4. Hit and false alarm rates in experiment 2.

Contribution of the left DMT to recognition memory 449

Fig. 5. Remember and know judgements in experiment 2.

The number of remember judgements recorded during the
VRM task remained significantly impaired relative to the
normal ranges of both control groups. The paucity of remember judgements was again compensated for by a significantly
elevated number of know responses, which fell outside the
NC[y] normal range only. In contrast, a large increase in
recorded remember judgements was evident in the FRM task,
such that the numbers of recorded remember and know
responses fell within the normal range of both normal
control groups.
QX’s discrimination accuracy during the VRM task
remained significantly impaired relative to both normal control groups. However, a slight improvement was noted in the
longer compared with the shorter exposure duration VRM
task. This is attributed to the decrease in his false alarm rate.
A dramatic improvement in discrimination accuracy was
evident in the FRM version. This result is underpinned by
an increased hit rate and a reduced false alarm rate. Finally,
a slight change in response bias was noted. Both sets of
normal controls displayed a conservative response bias on
the VRM task and a liberal one on the FRM task. QX
followed this trend. However, his stringently conservative
response evident on the VRM task fell outside the NC[o]
normal range. Measures of discrimination accuracy and
response bias are presented in Table 2.

Discussion
The reported results are broadly consistent with other studies
showing that amnesia is associated with a disproportionate
disruption of retrieval- compared with familiarity-driven
recognition memory (e.g. Verfaellie and Treadwell, 1993;
Kroll et al., 1996). The paucity of QX’s recollection judgements was most evident on the two VRM tasks, where the
absence of a demonstrable remember response rate was
compensated for by an increased know response rate. The
performance profiles on the FRM tasks were more variable.
The shorter exposure version was characterized by a low
remember rate but no evidence of a compensatory increase
in know rate, whereas the longer exposure version was
marked by a significant increase in remember judgements to
the extent that they were more numerous than know responses.
The significant increase in the contribution of remember
judgements to FRM on the longer exposure version suggests
that recollection- and familiarity-driven recognition memory
can be independently manipulated. However, it remains
uncertain whether the nature of the dissociation between
remember and know judgements on the VRM tasks reflects
qualitatively distinct memory processes (Yonelinas, 1994;
Yonelinas et al., 1996; Dobbins et al., 2000) or a memory
system based on trace strength (Donaldson, 1996).
Currently in the literature there is an ongoing debate as to

450 N. M. J. Edelstyn et al.

whether the raw number of know responses is an accurate
measure of familiarity (Gardiner et al., 1996; RichardsonKlavehn et al., 1996) or not (Jacoby, 1991; Yonelinas and
Jacoby, 1996; Khoe et al., 2000). Yonelinas and Jacoby
(1996) claim that the number of recorded know responses is
mathematically constrained by the proportion of remember
judgements in the remember–know procedure. They argue
that the contribution of familiarity to recognition memory
performance should be calculated from the proportion of
know responses, given that the subjects are allowed to make
such a response [i.e. familiarity ⫽ know/(1 – remember)].
This is an important debate, but has little impact on QX’s
VRM performance given his near complete reliance on know
responses.
There are also questions regarding the extent to which the
recollective component of recognition memory, as measured
by remember judgements, maps on to the contextual–recollection processes operating during free recall of verbal material
(Hanley et al., 2001). Hanley et al. reported a patient who
exhibited a severe impairment of free recall of verbal material
in the presence of intact use of remember and know judgements of tests of recognition memory. Drawing on the neural
network model of the hippocampus proposed by Hasselmo
and Wyble (1997), the authors suggested that during free
recall, context is used as a cue to retrieve an associated item,
whereas during recollective-driven recognition the item is
used as a cue to retrieve context. This case raises important
questions as to whether important differences exist between
the contextual–recollection processes used during free recall
and recollective-driven recognition, which in turn has implications for Aggleton and Brown’s (1999) proposals regarding
a common extended hippocampal system subserving free
recall as well as retrieval-driven recognition. QX displayed
clear evidence of contextual retrieval on verbal tests of
both free recall as well as recollection-driven recognition.
According to Hanley et al.’s interpretation of Hasselmo and
Wyble’s neural network model, this implies disruption of
both context- as well as item-cue retrieval. However, performance on the tests of visual memory was more variable, with
clear deficiencies evident on immediate and delayed retrieval
of the Rey Complex Figure and the recollection-driven FRM
(short exposure duration), yet spared delayed visual memory
(WMS) and recollection-driven FRM (long exposure duration). His ability to use recollection-driven recognition in
both FRM tasks, albeit to a lesser extent in the short
duration version, suggests that QX does experience conscious
recollective memory of the context in which target faces
were presented at encoding. And unlike VRM, his FRM
decisions are based on contextual retrieval and do not rely
disproportionately on familiarity-driven recognition. The fact
that QX’s lesion is restricted to the left DMT suggests that
this recollection-driven recognition for visual material is
being supported by his right episodic memory system.
Evidence regarding the neuroanatomical regions and pathways implicated in familiarity- and recollection-driven recognition memory remains equivocal. In their anatomical model,

Aggleton and Brown (1999) proposed that processes involved
in familiarity-driven recognition memory depend on a DMT–
perirhinal–prefrontal network whereas recollection-driven
recognition memory is supported by a hippocampal–diencephalic system. Thus, damage restricted to the DMT
should result in a selective deficit of familiarity not recollection. However, in the reported case, extensive left DMT
damage was associated with deficient recollection- not familiarity-driven recognition memory. Van der Warf et al. (1999)
also questioned DMT involvement in familiarity-driven
recognition memory. Their patient exhibited evidence of
impaired familiarity but spared recollection in the context of
a thalamic lesion that spared the DMT, anterior thalamic
nucleus and the mamillothalamic tract but involved the region
of the right lateral internal medullary lamina (IML). The
authors concluded that the IML and/or nuclei within, rather
than the DMT, is involved in familiarity-driven recognition
memory.
In addition to the reported verbal memory problems, QX
also displayed evidence of a mild impairment of processes
ascribed to the prefrontal cortex. For instance, QX’s behaviour
had become increasingly apathetic and disinhibited, and he
performed poorly on tasks which required high levels of
orderliness and planning (WAIS-R performance subtests).
Surprisingly, however, high levels of preservation were evident on traditional tests of executive processing (WCST and
Stroop). In addition, there were also signs of executive
memory dysfunction on the recognition memory tasks. This
was evident in QX’s extremely high false alarm rates on
both VRM tasks, and to a lesser extent, his abnormal response
bias exhibited on the short duration FRM and long duration
VRM tasks.
It has been claimed that the precise pattern of ‘prefrontal’
deficits depends on the focus of damage within the thalamus,
as specific thalamic nuclei have reciprocal projections with
different parts of the prefrontal cortex. Thus, QX’s behavioural symptoms are suggestive of dysfunction in the medialorbital prefrontal cortex (Duffy and Campbell, 1994); while
the presence of mild executive dysfunction suggests increased
levels of preserved dorsolateral prefrontal processing (Fuster,
1989). The presence of behavioural and cognitive evidence
of prefrontal dysfunction following DMT damage is broadly
consistent with Van der Warf et al.’s review of the neuropsychology of thalamic infarction, in which the authors tentatively
concluded that destruction of the midline nuclei (IML or the
DMT) may lead to executive dysfunction. Interestingly, the
authors go on to suggest that lesions of only one of these nuclei
may be insufficient to produce a pronounced dysexecutive
syndrome, which may account for the mild prefrontal signs
and symptoms reported for QX.
Taken together, these findings suggest that QX’s verbal
memory problems may be linked to recollection-based problems, which have their origins in impaired retrieval that
involves using context as well as items as cues and abnormal
strategic memory search strategies. Furthermore, when considering Aggleton and Brown’s neuroanatomical model, their

Contribution of the left DMT to recognition memory 451

prediction of impaired familiarity-driven recognition memory
following DMT damage is not supported. In sum, rather than
advocating a key role for the DMT in familiarity, it may be
that the DMT acts in an integrative role with respect to the
prefrontal cortex (Parker, 1999).

References
Aggleton JP, Brown MW. Episodic memory, amnesia, and the hippocampal–
anterior thalamic axis. Behavioral and Brain Sciences 1999; 22: 425–89.
Aggleton JP, Saunders RC. The relationship between temporal lobe and
diencephalic structures implicated in anterograde amnesia. Memory 1997;
5: 49–71.
Aggleton JP, Shaw C. Amnesia and recognition memory: A reanalysis of
psychometric data. Neuropsychologia 1996; 34: 51–62.
Bachevalier J, Mishkin M. Visual recognition impairment follows ventromedial
but not dorsolateral prefrontal lesions in monkeys. Behavioral Brain
Research 1986; 20: 249–61.
Dobbins IG, Khoe W, Yonelinas AP, Kroll NEA. Predicting individual false
alarm rates and signal detection theory: A role for remembering. Memory
and Cognition 2000; 28: 1347–56.
Donaldson W. Measuring recognition memory. Journal of Experimental
Psychology: General 1992; 121: 275–7.
Donaldson W. The role of decision processes in remembering and knowing.
Memory and Cognition 1996; 24: 523–33.
Duffy JD, Campbell JJ. The regional prefrontal syndromes: a theoretical and
clinical review. Journal of Neuropsychiatry and Clinical Neurosciences
1994; 6: 379–87.
Fuster JM. The prefrontal cortex: Anatomy, physiology and neuropsychology
of the frontal lobe. New York: Raven Press, 1989.
Gardiner JM, Java RI. Recollective experience in word and nonword
recognition. Memory and Cognition 1990; 18: 23–30.
Gardiner JM, Java RI, Richardson-Klavehn A. How level of processing
really influences awareness in recognition memory. Canadian Journal of
Experimental Psychology 1996; 50: 114–22.
Graff-Radford NR, Tranel D, Van Hoesen GW, Brandt JP. Diencephalic
amnesia. Brain 1990; 113: 1–25.
Hanley JR, Davies ADM, Downes JJ et al. Remembering and knowing in a
patient with preserved recognition and impaired recall. Neuropsychologia
2001; 39: 1003–10.
Hasselmo ME, Wyble BP. Free recall and recognition in a network model of
the hippocampus: simulating effects of Scopolamine on human memory
functions. Behavioural Brain Research 1997; 82: 169–77.
Heaton RK, Chelune GJ, Talley JL et al. Wisconsin Card Sorting Test. Odessa,
FL: Psychological Assessment Resources, 1989.
Jacobson RR. Alcoholism in Korsakoff’s syndrome and the frontal lobes.
Behavioral Neurology 1989; 2: 25–38.
Jacoby LL. A process-dissociation framework: separating automatic from
intentional uses of memory. Journal of Memory and Language 1991; 30:
513–41.
Jacoby LL, Dallas M. On the relationship between autobiographical memory
and perceptual learning. Journal of Experimental Psychology: General 1981;
3: 306–40.
Kapur N, Scholey K, Moore F et al. The mamillary bodies revisited: Their
role in human memory functioning. In: Neuropsychological explanations
in memory and cognition: Essays in honour of Nelson Butters. New York:
Plenum Press, 1995.
Kay J, Lesser R, Coltheart M. Psycholinguistic Assessments of Language
Processing in Aphasia. Hove: LEA, 1992.
Khoe W, Kroll NEA, Yonelinas AP et al. The contribution of recollection and
familiarity to yes–no and forced-choice recognition tests in healthy subjects
and amnesics. Neuropsychologia 2000; 38: 1333–41.
Knowlton BJ, Squire LR. Remembering and knowing: Two different
expressions of declarative memory. Journal of Experimental Psychology
1995; 21: 699–710.
Kopelman M, Wilson B, Baddeley A. The Autobiographical Memory Interview.
Bury St Edmunds, Suffolk, UK: Thames Valley Test Co, 1990.
Kroll NEA, Knight RT, Metcalfe J et al. Cohesion failure as a source of
memory illusions. Journal of Memory and Language 1996; 35: 176–96.
Lezak MD. Neuropsychological assessment, 3rd edn. Oxford: Oxford
University Press, 1995.

Mair RG, Burk JA, Porter MC, Ley JE. Thalamic amnesia and the
hippocampus: Unresolved questions and an alternative candidate. Behavioral
and Brain Sciences 1999; 22: 458–9.
Mandler G. Recognizing: the judgment of previous occurrence. Psychological
Review 1980; 87: 252–71.
Markowitsch HJ. Which brain regions are critically involved in the recollection
of old episodic memory. Brain Research Reviews 1995; 21: 117–27.
Mayes AR, van Eijk R, Gooding PA et al. What are the functional deficits
produced by hippocampal and perirhinal cortex lesions? Behavioral and
Brain Sciences 1999; 22: 460–1.
Mennemeier M, Fennell E, Valenstein E, Heilman KM. Contributions of the
left intralaminar and medial nuclei to memory. Comparisons and report of
a case. Archives of Neurology 1992; 49: 1050–8.
Myers JE, Myers KR. Rey Complex Figure Test. Odessa, FL: Psychological
Assessment Resources, 1995.
Nelson HE. National Adult Reading Test. Windsor: NFER-Nelson, 1982.
Okazaki H. Fundamentals of neuropathology, 2nd edn. Tokyo: Igaku-shoin,
1989.
Parker A. Memory systems, frontal cortex, and the hippocampal axis.
Behavioral and Brain Sciences 1999; 22: 464–5.
Parker A, Gaffan D. The effects of anterior thalamic and cingulate cortex
lesions on ‘object-in-place’ memory in monkeys. Neuropsychologia 1997;
35: 1093–102.
Parker A, Gaffan D. Memory after frontal/temporal disconnection in monkeys:
Conditional and non-conditional tasks, unilateral and bilateral frontal lesions.
Neuropsychologia 1998; 36: 259–71.
Parker A, Eacott MJ, Gaffan D. The recognition memory deficit caused by
mediodorsal thalamic lesion in nonhuman primates. European Journal of
Neuroscience 1997; 9: 2423–31.
Parker A, Wilding E, Akerman C. Novelty and memory encoding: Von
Resteroff effects in normal human and non-human primate recognition
memory, and the role of frontal/perirhinal interaction. Journal of Cognitive
Neuroscience 1998; 10: 691–703.
Rajaram S. Remembering and knowing: Two means of access to the personal
past. Memory and Cognition 1993; 21: 89–102.
Rajaram S, Geraci L. Conceptual fluency selectively influences knowing.
Journal of Experimental Psychology: Learning, Memory and Cognition
2000; 26: 1070–4.
Rapscak SZ, Reminger SL, Glisky ML et al. Neuropsychological mechanisms
of false recognition following frontal lobe damage. Cognitive
Neuropsychology 1999; 16: 267–92.
Richardson-Klavehn A, Gardiner JM, Java RI. Memory: task dissociations,
process dissociations and dissociations of consciousness. In: Underwood
G, editor. Implicit cognition. Oxford: Oxford University Press 1996; 85–158.
Schacter DL, Verfaellie M, Pradere D. Neuropsychology of memory illusions:
False recall and recognition in amnesic patients. Journal of Memory and
Language 1996; 35: 319–34.
Shallice T. From neuropsychology to mental structure. Cambridge: Cambridge
University Press, 1980.
Snodgrass JG, Corwin J. Pragmatics of measuring recognition memory.
Applications to dementia and amnesia. Journal of Experimental Psychology:
General 1988; 117: 34–50.
Squire LR. Two forms of human amnesia: An analysis of forgetting. Journal
of Neuroscience 1981; 1: 635–40.
Squire LR, Zola SM. Amnesia, memory and brain systems. Philosophical
Transactions of the Royal Society of London Series B 1997; 352: 1663–73.
Stroop JR. Stroop Test. Odessa, FL: Psychological Assessment Resources,
1989.
Tulving E. Memory and consciousness. Canadian Psychologist 1985; 26: 1–12.
Van der Warf YD, Weerts JGE, Jolles J et al. Neuropsychological correlates
of a right unilateral lacunar thalamic infarction. Journal of Neurology,
Neurosurgery and Psychiatry 1999; 66: 36–42.
Van der Warf Y, Witter M, Uylings HBM, Jolles J. Neuropsychology of
infarctions in the thalamus: a review. Neuropsychologia 2000; 38: 613–27.
Verfaellie M, Treadwell JR. Studies of recognition memory in amnesia.
Neuropsychology 1993; 7: 5–13.
Victor M, Adams RD, Collins GH. The Wernicke–Korsakoff syndrome and
related neurologic disorders due to alcoholism and malnutrition, 2nd edn.
Philadelphia: Davis, 1989.
Von Cramon DY, Hebel N, Schuri U. A contribution to the anatomical basis
of thalamic amnesia. Brain 1985; 108: 993–1008.
Ward J, Parkin AJ, Powell G et al. False recognition of unfamiliar people:
‘Seeing film stars everywhere’. Cognitive Neuropsychology 1999; 16:
239–315.

452 N. M. J. Edelstyn et al.
Wechsler D. Wechsler Adult Intelligence Scales–Revised. Cleveland, OH:
Psychological Corporation, 1981.
Wechsler D. Wechsler Memory Scale–III. Psychological Corporation. Texas:
Harcourt Brace, 1997.
Weiskrantz L. On issues and theories of the human amnesic syndrome. In:
Weinberger NM, McGaugh JL, Lynch G, editors. Memory systems of the
human brain: animal and human cognitive processes. New York: Guilford
1985; 381–415.
Witter MP, Van der Warf YD. The medial dorsal nucleus of the thalamus is
not part of a hippocampal–thalamic memory system. Behavioral and Brain
Sciences 1999; 22: 407–8.
Yonelinas AP. Receiver-operating characteristics in recognition memory:
Evidence for a dual-process model. Journal of Experimental Psychology:
Learning, Memory and Cognition 1994; 20: 1341–54.
Yonelinas AP, Jacoby LL. Noncriterial recollection: familarity as automatic,
irrelevant recollection. Consciousness and Cognition 1996; 5: 131–41
Yonelinas AP, Dobbins I, Szymanski MD et al. Signal detection, threshold
and dual-process models of recognition memory: ROCs and conscious
recollection. Consciousness and Cognition 1996; 5: 418–41.

Received on 3 August, 2001; resubmitted on 12 April, 2002; accepted
on 8 May, 2002

Contribution of the left dorsomedial
thalamus to recognition memory:
a neuropsychological case study
N. M. J. Edelstyn, S. J. Ellis, P. Jenkinson and
A. Sawyer
Abstract
This study reports a patient with a unilateral left thalamic lesion which was
centred on the dorsomedial thalamic nucleus. Cognitive neuropsychological
assessment revealed a severe impairment in verbal memory and symptoms of
executive dysfunction, in the presence of relatively intact visual and facial
recognition, working memory, praxis, language and IQ. Verbal and visual
recognition memory were investigated using the remember–know
paradigm. The results indicated a profound impairment in recollection-driven
verbal recognition memory. These results are discussed in the context of the
role of the dorsomedial thalamic nucleus in recognition memory, and functional
models of memory.

Journal
Neurocase 2002; 8: 442–52

Neurocase Reference Number:
O278

Primary diagnosis of interest
Severe impairment in verbal memory and symptoms of executive dysfunction

Author’s designation of case
QX

Key theoretical issue
d Relationship between the (left) dorsomedial thalamus and the component

processes of recognition memory
Key words: dorsomedial thalamus; recognition memory; neuropsychology;
recollection; familiarity

Scan, EEG and related measures
Magnetic resonance imaging

Standardized assessment
Revised Wechsler Adult Intelligence Scale (WAIS-R), National Adult Reading
Test (NART), Psycholinguistic Assessments of Language Processing in
Aphasia (PALPA), Stroop, Wisconsin Card Sorting Test (WCST), verbal
fluency, Wechsler Memory Scales (WMS), Autobiographical Memory
Inventory (AMI), Rey Complex Figure

Other assessment
Unfamiliar/famous face sorting, familiar face identification, in/animate naming
(including function and association assessments for inanimate items), two
versions of a recognition memory test for words and faces (3000 and 6000 ms
target exposure)

Lesion location
d Left dorsomedial thalamus

Lesion type
A discrete lacuna infarction

Language
English