A COMPARISON OF EGOCENTRIC AND ALLOCENTRIC
SPATIAL MEMORY IN MEDIAL TEMPORAL LOBE
AND KORSAKOFF AMNESICS
J.S. Holdstock1, A.R. Mayes1, E. Cezayirli2, J.P. Aggleton3 and N. Roberts2
(1Department of Clinical Neurology, University of Sheffield; 2Magnetic Resonance
Imaging and Analysis Research Centre, University of Liverpool; 3School of Psychology,
Cardiff University)

ABSTRACT
Two patients with medial temporal lobe damage, seven Korsakoff amnesics and fourteen
healthy control subjects were tested on three conditions of a spatial memory test (‘short
delay’, ‘allocentric’ and ‘egocentric’). The task required subjects to recall the position of a
single spot of light presented on a board after various delays. The ‘short delay’ condition
tested memory over very short, unfilled intervals. The other two conditions used longer,
filled delays. The allocentric condition required subjects to move to a different place around
the board before recalling the position of the light. In the egocentric condition stimuli were
presented in darkness, which eliminated allocentric cues. The Korsakoff amnesics were
impaired at all delays of the short delay tasks, suggesting poor encoding. On the allocentric
and egocentric tasks the Korsakoff amnesics showed a comparable impairment in the two
conditions, which worsened with delay. This accelerated forgetting suggested that the
Korsakoff amnesics also had impaired memory for allocentric and egocentric information.
The patients with medial temporal lobe damage were unimpaired in the ‘short delay’
condition suggesting intact encoding and short-term memory of spatial information.
However, they were impaired in the allocentric condition and showed accelerated loss of
allocentric spatial information. In the egocentric condition, while the performance of one
patient was impaired, the performance of the other was as good as controls. This result
suggests that, in contrast to allocentric spatial memory, which is sensitive to medial temporal
lobe damage, an intact medial temporal lobe need not be necessary for successful
performance on an egocentric spatial memory task.
Key words: amnesia, egocentric, allocentric, spatial, memory

INTRODUCTION
Patients with anterograde amnesia suffer impairments of spatial memory. In
real life, amnesics can readily find themselves lost (Corkin, 1984), while
laboratory studies have revealed a failure to remember a variety of spatial stimuli
(Cave and Squire, 1991; Holdstock, Shaw and Aggleton, 1995; Joyce and
Robbins, 1991; MacAndrew and Jones, 1993; Mayes, Meudell and MacDonald,
1991; Pigott and Milner, 1993; Shoqeirat and Mayes, 1991; Smith and Milner,
1981, 1989; Warrington and Baddeley, 1974). Although all of these examples
assessed ‘spatial memory’ it may well have been that their varying tasks tapped
different cognitive processes. For example, Maguire, Burke, Phillips et al. (1996a)
cite studies which show a dissociation between topographical orientation and
other aspects of spatial memory. As a consequence, it is important to determine
Cortex, (1999) 35, 479-501

480

J.S. Holdstock and Others

whether all aspects of spatial memory are equally disrupted in amnesia. Further,
the nature of any spatial memory loss, e.g. disruption of encoding or storage,
needs to be investigated. These issues were addressed in the present study.
The best known distinction in the spatial memory literature is between
allocentric and egocentric memory (O’Keefe and Nadel, 1978; Semmes,
Weinstein, Ghent et al., 1963). The former refers to the identification of a place
by reference to the relative position of an array of external stimuli or landmarks.
This contrasts with egocentric memory, in which a location is identified by its
relative position to the observer. Single cell recording studies of rats (Muller,
Kubie and Ranke, 1987; O’Keefe, 1991; O’Keefe and Conway, 1978, O’Keefe
and Dostrovsky, 1971; O’Keefe and Nadel, 1978; O’Keefe and Speakman, 1987)
and monkeys (Feigenbaum and Rolls, 1991) have implicated the hippocampus in
spatial memory particularly when an allocentric spatial frame of reference is
used (but see Rolls and O’Mara, 1995). This has been supported by lesion
studies in rats of the hippocampus (Jarrard, 1993; Morris, Garrud, Rawlins et al.,
1982) and fornix (O’Keefe and Conway, 1980). In the study of human amnesia,
only a few studies have attempted to develop tests which selectively require just
allocentric or egocentric frames of reference. The majority of these studies have
investigated spatial memory in patients who have undergone unilateral temporal
lobectomy or who have unilateral hippocampal sclerosis rather than amnesia
(Abrahams, Pickering, Polkey et al., 1997; Feigenbaum, Polkey and Morris,
1996; Goldstein, Canavan and Polkey, 1989; Maguire, Burke, Phillips et al.,
1996; Morris, Pickering, Abrahams et al., 1996). The results of these studies
have implicated the temporal lobe, particularly that on the right, in allocentric
spatial memory. Further, the results of the study of Goldstein et al. (1989)
suggested that right and left unilateral temporal lobectomy, in which the anterior
5.5-6.5 cm of the temporal lobe was removed, did not impair egocentric spatial
memory. Consistent with these behavioural results, three positron emission
tomography (PET) studies have shown activation of the right hippocampus
whilst learning to navigate and orient oneself in an environment (topographical
learning) (Maguire, Burgess, Donnett et al., 1998; Maguire, Frackowiak and
Frith, 1996, 1997). These studies support the role of the hippocampus in
allocentric spatial memory in humans but as appropriate nonallocentric
comparison tasks have rarely been used, the question of whether the spatial
memory impairment found after temporal lobe damage is exclusively allocentric
has not been sufficiently investigated.
Recently, Holdstock, Mayes, Cezayirli et al. (submitted) studied the spatial
memory of a patient, YR, with relatively selective damage to the hippocampus
using a new procedure which distinguished allocentric from egocentric spatial
memory. YR was impaired both at recalling and recognising the position of a
previously illuminated light-emitting diode (LED) on a tabletop board after a
one minute delay when the use of an allocentric reference frame was
encouraged. However, YR was not significantly impaired on an egocentric
version of the recall task at the same delay. The results of this study indicate a
greater role for the hippocampus in allocentric than egocentric spatial memory
and are consistent with the published findings described above.
The parahippocampal gyrus has also been implicated in spatial memory. Deficits

Spatial memory in amnesia

481

in topographical memory have been reported following medial temporal lobe
lesions in humans (De Renzi, 1985; Habib and Sirigu, 1987; Landis, Cummings,
Benson et al., 1986; Maguire, Burke, Phillips et al., 1996) including selective
lesions of the right parahippocampal cortex (Bohbot, Kalina, Stepankova et al.,
1997). In addition, a recent study (Bohbot, Kalina, Stepankova et al., 1998) found
that a lesion to the right parahippocampal cortex impaired performance on a task
which was analogous to the Morris Water Maze when memory was tested after a 30
minute delay. In contrast, a lesion to the right hippocampus did not impair
performance on this task (Bohbot et al., 1998). The results of the study provided
further support for the view that the parahippocampal cortex plays a critical role in
some types of spatial memory. Further evidence for the role of the parahippocampal
gyrus in spatial memory has come from functional magnetic resonance imaging
(Aguirre, Detre, Alsop et al., 1996; Aguirre, Zarahn and D’Esposito, 1998) and
PET studies (Maguire, Frackowiak and Frith, 1996) which have shown bilateral
activation of the parahippocampal gyrus during topographical memory tasks. More
specifically, Maguire, Frith, Burgess et al. (1998) suggested that the role of the
parahippocampal gyrus is to encode object position e.g. the position of landmarks
(but see Epstein and Kanwisher, 1998). Two other PET studies, in which the right
parahippocampal gyrus was activated during retrieval of the location of an array of
pictures presented on a computer screen (Milner, Johnsrude and Crane, 1997;
Owen, Milner, Petrides et al., 1996), indicate that the memory processing of the
parahippocampal gyrus may not necessarily be restricted to the allocentric frame
used in topographical memory. In these studies, either allocentric or egocentric
frameworks could have been used to remember picture locations.
The present study investigated the contribution of regions adjacent to the
hippocampus to egocentric and allocentric spatial memory. As it is very unusual
to find patients with selective damage to the parahippocampal gyrus, we studied
two patients who were shown, by magnetic resonance imaging (MRI), to have
bilateral medial temporal lobe damage which included both the hippocampus and
the parahippocampal gyrus. In the light of the recent imaging data concerning
spatial memory and the parahippocampal gyrus, we were interested to see whether
having damage to the parahippocampal gyrus as well as the hippocampus, would
accentuate the allocentric spatial memory deficit found after selective
hippocampal lesions. Of equal interest was whether additional parahippocampal
damage would affect egocentric spatial memory for single light positions. As
discussed above, recent PET studies suggest that some aspects of egocentric
spatial memory may be affected by parahippocampal gyrus damage (Milner et al.,
1997; Owen et al., 1996) and we wanted to see whether egocentric memory for
the information presented in our task (the position of the most recently presented
LED) was dependent on the parahippocampal gyrus. Given the widespread view
that damage to medial temporal lobe and diencephalic structures within Papez
circuit result in equivalent memory deficits (Aggleton and Saunders, 1997; Delay
and Brion, 1969; Parker and Gaffan, 1997; although see Parkin, 1992; Hunkin,
Parkin and Longmore, 1994, for an alternative view), it is also of interest to
investigate the role of diencephalic structures in allocentric and egocentric spatial
memory. Therefore, a group of a patients with diencephalic damage due to
Korsakoff’s syndrome were also tested. Korsakoff amnesic patients are interesting

482

J.S. Holdstock and Others

because they all display bilateral pathology in the mamillary bodies (Victor et al.,
1971) which appear to be important for allocentric spatial memory (Neave, Nagle
and Aggleton, 1997; Sutherland and Rodriguez, 1989). It is, of course, the case
that people with Korsakoff’s syndrome typically display pathology in a range of
sites in addition to the mamillary bodies which may cause more widespread
spatial memory deficits. The performance of these two sets of patients was
compared with a group of healthy control subjects.
MATERIALS AND METHODS
Subjects
Two subjects with medial temporal lobe damage (RS and NM), seven subjects suffering
from alcoholic Korsakoff’s syndrome and fourteen healthy control subjects (CON) were
tested in this study. The Korsakoff group (two female, five male) had a mean age of 61
years (range 54-68).
Amnesic subjects were assessed with the WAIS-R and NART-R (Nelson and Willison,
1991) to obtain estimates of present and premorbid IQ, respectively. Memory was assessed
with the revised version of the Wechsler Memory Scale (WMS-R) and the Recognition
Memory Test (RMT) (Warrington, 1984). The scores of the individual patients and the mean
scores of the group on these tests are shown in Table I. Measures of frontal lobe function
were obtained for the Korsakoff patients using the Wisconsin Card Sorting Test (WCST)
(Heaton, 1981), verbal fluency (Benton, 1968), design fluency (Jones-Gottman and Milner,
1977) and cognitive estimates (Shallice and Evans, 1978). The results of the patients on these
tests are shown in Table II. On the WCST the performance of the group was a little low but
not abnormal for this age group. Also shown in Table II are scores for the picture arrangement
(Mcfie and Thompson, 1972) and block design (Benton, 1968; Goodglass and Kaplan, 1979)
subtests of the WAIS-R which are also thought to tax frontal lobe functions.
Patient NM is a 46 year old man who became amnesic after suffering meningitis in
1970. Patient RS is a 49 year old man who became amnesic as a result of herpes simplex
encephalitis in September of 1984. A magnetic resonance imaging (MRI) scan was obtained
for both patients using a 3D T1-weighted radio-frequency spoiled gradient echo (SPGR)
TABLE I

Details of the Amnesic Subjects in Terms of Aetiology, Age, NART-R Full-scale IQ, WAIS-R Full-scale,
Verbal and Performance IQ, Three Index Scores of the WMS-R and Number Correct on the
Recognition Memory Test (RMT)
NART-R

WAIS-R
VIQ

WMS-R

RMT

Subject

Aet.

Age

FIQ

FIQ

PIQ

GM

ATT/C

DR

W

HKK
RST
SM
BP
BJ
HK
TC
Mean
S.D.

K
K
K
K
K
K
K

64
58
68
58
66
54
58
61
5.1

90
102
91
94
91
118
98
97.7
9.9

85
82
91
105
104
105
96
98
95
95
94
99
101
103
97
101
100
88
80
80
80
94.7 94.4 93.6
9.1
9.8
8.1

62
51
72
72
64
74
50
63.6
10

90
81
107
111
101
87
60
91
17.5

55
50
61
51
64
51
58
55.7
5.5

28 29
30 38
30 38
38 38
28 29
24 33
30 24
29.7 32.7
4.2 5.6

NM
RS

M
HSE

46
49

84
111

84
107

77
74

88
92

50
55

26
33

82
110

87
104

F

32
33

FIQ = full-scale IQ, VIQ = verbal IQ, PIQ = performance IQ, GM = general memory, ATT/C = attention/concentration,
DR = delayed recall, W = words, F = faces. Aetiology key: K = alcoholic Korsakoff syndrome, HSE = post herpes
simplex encephalitis, M = post meningitis.

Spatial memory in amnesia

483

TABLE II

Performance of the Amnesic Subjects on Tests of Frontal Lobe Function
WCST
No.
Cat.

WCST
No.
P.R.

VF
(FAS)

DF
ppr

DF
no

HKK
RST
SM
BP
BJ
HK
TC

2
5
2
6
3
0
2

28
30
57
20
66
85
23

31
26
34
16
32
41
23

6.7
28.6
37.5
0.0
15.8
16.7
61.1

5
5
1
4
11
10
0

NM
RS

6
6

20
23

14
26

WAIS-R WAIS-R
picture
block
arrang.
design
6
5
6
9
7
6
2

7
9
9
8
8
9
6

Cog.
Est.
6
9
12
3
6
4
2
11
4

The scores are number of categories (No. cat.) and number of perseverative responses (No. P.R.) on the WCST,
corrected score on the verbal fluency test (VF), percent perseverative response on the design fluency test (DF ppr),
novel output on the design fluency test (DF no), scaled score on the picture arrangement subtest of the WAIS-R
(WAIS-R picture arrang.), scaled score on the block design subtest of the WAIS-R, raw score on the Cognitive
Estimates Test (Shallice and Evans, 1978) (Cog. Est.). Scores for patients NM and RS are given for WCST, verbal
fluency and cognitive estimates.

sequence (TE = 9 ms, TR = 34 ms, flip angle = 45 degrees, matrix size = 256 × 192, 2
NEX, field of view = 20 cm, acquisition time = 27 minutes and 52 seconds) available on a
1.5 T SIGNA whole-body magnetic imaging system (General Electric, Milwaukee, WI,
U.S.A.). Each image referred to a contiguous section of tissue, 1.6 mm thick. Images for
both patients are shown in Figure 1.
Detailed radiological inspection of NM’s scans showed partial bilateral damage to the
head, body and tail of the hippocampus accompanied by partial damage to the amygdala
and to the parahippocampal, perirhinal and entorhinal cortices. In addition, there was slight
atrophy of the superior frontal and parietal lobes, mamillary bodies and cerebellum. RS’s
scan showed the hippocampus to be shrunken throughout its length with more evident
volume loss at the head. There was no apparent damage to the amygdala. The
parahippocampal, perirhinal and entorhinal cortices appeared to be thin on the left and there
was possible damage to the perirhinal cortex on the right. The scan also revealed slight
gyral and cerebellar atrophy.
In order to obtain quantitative measures of pathology, the volumes of the whole
hemisphere, lateral ventricle, third ventricle, caudate, temporal lobe, hippocampus,
parahippocampal gyrus, amygdala and parietal lobe were estimated on the right and the left
using stereology (Bartzokis, Mintz, Marx et al., 1993) available in Analyze (Mayo
Foundation, Minnesota, U.S.A.). The area of the parahippocampal gyrus, the gyrus adjacent
and inferior to the hippocampus, which includes the parahippocampal, the entorhinal and
the perirhinal cortices plus white matter was measured on the same sections as the
hippocampus. In addition to these measures, volumes of grey and white matter were
obtained for the prefrontal cortex, which was defined as that part of the frontal lobe which
is rostral to the anterior-most point of the corpus callosum when the images are aligned
along the AC-PC line. For grey/white matter segmentation, images were spatially coregistered with a reference template using a 12 point affine transformation in the Talairachspace and were segmented into grey and white matter compartments using a modified
clustering algorithm.
For each subject, the area of a single midsagittal T1 weighted image from that subject’s
scan was used to correct the structure volumes for inter-subject variability in pre-morbid
brain volume. The section used was that on which the pineal gland and the whole aquaduct
of Sylvius could be seen. The slice also passed between the mamillary bodies so that a
minimal amount of these structures was visible. The section area was delineated by the
internal table of the skull, the tentorium cerebelli, and an imaginary line joining the most
anterior point of the tentorium cerebelli and the most inferior point of the frontal lobe. The

484

J.S. Holdstock and Others

Fig. 1 – Coronal sections from the patients RS, NM, and an age- and sex-matched control
subject. Images on the top and bottom rows show sections through the frontal and parietal lobes,
respectively. Images on the second and third row pass through the head and body of the
hippocampus, respectively. Heads of the hippocampi of the patients are significantly smaller than
controls on both sides. These images also show the parahippocampal gyri.

section area was calculated by counting the number of pixels within the area of interest.
The individual estimated brain structure volumes were then corrected by dividing them by
the section area. The estimated structure volumes for the patients were compared with those
of a group of eight healthy male controls matched for age.
The estimated volumes of NM’s and RS’s hippocampi were comparable and
significantly smaller than their controls (see Table III). However, more detailed investigation
of the hippocampus revealed differences in RS’s and NM’s pattern of pathology. In this

Spatial memory in amnesia

485

TABLE III

Volume Measures, before and after Correcting for Sagittal Section Area (see text), of the Hemispheres,
Lateral Ventricles, Third Ventricles, Caudate, Temporal Lobe, Hippocampus, Parahippocampal Gyrus,
Amygdala and Parietal Lobe for Patients RS and NM, the Mean Volume and Standard Deviation, in
brackets, of these Structures for a Group of Age and Sex Matched Controls and the Number of
Standard Deviations that RS’s and NM’s Corrected Volumes were from the Control Mean for each
Structure. (Negative Standard Deviations indicate a Smaller Volume than Controls)
Con
raw
R hem

515.7
(41.9)
L hem
500.8
(40)
R wm
232.7
(27.55)
L wm
213.8
(24.07)
R vent
7.47
(3.25)
L vent
7.98
(3.41)
3rd vent
4.78
(1.58)
R caud
3.79
(0.77)
L caud
3.85
(0.92)
R temp
77.06
(4.45)
L temp
71.49
(7.49)
R hipp
2.5
(0.44)
L hipp
2.37
(0.41)
R parahipp
2.66
(0.54)
L parahipp
2.49
(0.58)
R amyg
1.92
(0.29)
L amyg
1.8
(0.32)
R Par
8.75
(0.62)
L Par
8.53
(0.58)
R Prefron
25.65
(3.8)
L Prefron
26.67
(4.38)

RS

NM

corr

raw

corr

sds

raw

corr

sds

4.63
(0.139)
4.496
(0.106)
2.088
(0.184)
1.918
(0.133)
0.068
(0.029)
0.071
(0.028)
0.043
(0.014)
0.034
(0.007)
0.035
(0.008)
0.693
(0.039)
0.643
(0.065)
0.022
(0.003)
0.021
(0.003)
0.024
(0.004)
0.022
(0.004)
0.017
(0.002)
0.016
(0.002)
0.079
(0.003)
0.077
(0.004)
0.231
(0.038)
0.241
(0.041)

457.9

3.821

– 5.82*

557.5

4.34

– 2.07*

424.2

3.539

– 9.02*

525

4.089

– 3.84*

186.9

1.560

– 2.87*

259.2

2.018

– 0.38

173.5

1.448

– 3.54*

221.8

1.728

– 1.4

18.57

0.155

2.99*

11.25

0.088

0.69

28.8

0.240

5.99*

15.42

0.120

1.73

7.36

0.061

1.3

4.01

0.031

– 0.86

3.06

0.025

– 1.29

3.73

0.029

– 0.76

3.51

0.029

– 0.65

4.37

0.034

– 0.07

58.35

0.487

– 5.34*

84.37

0.657

– 0.94

53.71

0.448

– 3*

79.66

0.620

– 0.35

1.27

0.011

– 4.03*

1.46

0.011

– 3.76*

0.98

0.008

– 5*

1.16

0.009

– 4.69*

2.85

0.024

0

2.08

0.016

– 2*

2.46

0.021

– 0.41

1.62

0.013

– 2.2*

1.6

0.013

– 1.86

0.97

0.008

– 4.57*

1.68

0.014

– 0.88

0.62

0.005

– 4.71*

8.09

0.068

– 3.36*

6.79

0.053

– 7.79*

8.57

0.072

– 1.33

6.79

0.053

– 6.1*

26.59

0.222

– 0.26

30.37

0.237

0.14

27.40

0.229

– 0.29

27.34

0.213

– 0.68

* Indicates greater than 1.96 SDs below the control mean.
Con = control mean (sd in brackets); raw = estimated volume of structure; corr = volume structure after correcting for
sagital section volume (see text); sds = number of standard deviations that the patient’s corrected volume was from
the corrected control volume; R hem = Right hemisphere; L hem = Left hemisphere; R vent = Right lateral ventricle;
L vent = Left lateral ventricle; 3rd vent = Third ventricle; R caud = Right caudate; L caud = Left caudate; R temp =
Right temporal lobe; L temp = Left temporal lobe; R hipp = Right hippocampus; L hipp = Left hippocampus; R
parahipp = Right parahippocampal gyrus; L parahipp = Left parahippocampal gyrus; R amyg = Right amygdala; L
amyg = Left amygdala; R Par = Right parietal lobe; L Par = Left parietal lobe; R Prefron = Right prefrontal lobe grey
matter; L Prefron = Left prefrontal lobe grey matter.

486

J.S. Holdstock and Others

analysis, for NM, RS and their controls, the sectional area of the hippocampus was plotted
for 15 slices sampled from the entire length of the structure. As shown in Figure 2, RS had
a smaller hippocampus anteriorly than NM on both the right and the left whereas NM had
a slightly smaller hippocampus than RS posteriorly. The volume measures of the
parahippocampal gyrus and amygdala confirmed that NM had suffered more extensive
damage to medial temporal lobe structures beyond the hippocampus than RS.
The estimated volumes of the whole hemispheres suggested that RS had greater cortical
atrophy than NM (see Table III). The measures of whole hemisphere white matter volume
showed that RS also had significantly less white matter than the controls whereas this was
not the case for NM.
Separate volume measures of the prefrontal, temporal and parietal lobes revealed that
neither patient showed significant atrophy of the frontal lobe. In the temporal lobe, atrophy
was greater bilaterally in RS than NM but the reverse was true for the parietal lobe, where
NM had much greater atrophy bilaterally than RS (see Table III). This latter finding is
particularly relevant given the proposed role of the parietal cortex in egocentric spatial
processing (e.g. Anderson, 1997; Berthoz, 1997; Karnath, 1997; McDaniel, Via, Smith et
al., 1995; Save and Moghaddam, 1996). RS also had enlarged lateral and third ventricles
whereas NM’s ventricle volumes did not differ significantly from the control mean volume
(see Table III). Enlarged third ventricle volume may indicate pathology of diencephalic
structures such as the thalamus.
Psychometric test scores for patients NM and RS are shown in Tables I and II. In
addition, both patients have been tested on the Doors and People test battery (Baddeley,
Emslie and Nimmo-Smith, 1994). NM showed poor performance on all subtests of the
battery. His performance on the recall subtests, “People” and “Shapes”, and the nonverbal
recognition test, “Doors”, was at less than the first percentile and was at the first percentile
level for the verbal recognition, “Names”, subtest. RS’s performance on the recall subtests,
“People” and “Shapes”, was at less than the first percentile, which was slightly worse than
his performance on the two recognition tests, “Doors” and “Names”, where he was between
the 10th and 25th percentile level and at the 25th percentile, respectively.
The control (CON) subjects (7 male and 7 female) had a mean age of 56 years (range
47-73). Psychometric testing of these subjects was limited to the NART-R and the mean
predicted FSIQ was 103 (range 86-122). A t-test showed that the NART-R FSIQ scores did
not differ significantly from those of the Korsakoff patients (t = 1.05, d.f. = 19, p > 0.1)
nor did they differ from the WAIS-R FSIQ scores of the Korsakoff patients (t = 1.68,
d.f. = 19, p > 0.05). The WAIS-R FSIQ scores of patients NM and RS were within one
standard deviation of the mean NART-R FSIQ score of the controls.
Apparatus
The study required subjects to view a single light on a large uniform board which
consisted of a featureless sheet of translucent Perspex which had a matt finish, so that
reflections were not visible in its surface. The sheet was 60.5 cm wide, 91 cm long, and
rounded at both ends. Care was taken to ensure that neither the Perspex nor its surrounding
edges provided distinctive cues, i.e. it was free of blemishes or joins. Embedded under the
Perspex board were twenty-five 20 milliamp standard red LEDs arranged in a random
manner. When unlit, the LEDs were not visible from above the test board. When lit, the
LEDs were sufficiently dim to avoid causing afterimages.
Design and Procedure
For all task conditions the lightboard containing the LEDs was positioned on a stand in
the centre of a room away from obvious local cues. The test apparatus was portable and a
variety of rooms were used for the study. The subject was instructed to attend to the
lightboard, an LED was lit for approximately two seconds and, after a delay, the subject had
to mark the location of that LED. Each of the three task conditions examined consisted of a
single session of 45 trials in which three different retention durations (15 trials for each
duration) were presented in a mixed sequence. As there were 25 LEDs on the board and 45

Spatial memory in amnesia

487

Fig. 2 – Right and left hippocampal profiles of patients RS and NM plotted against eight ageand sex-matched control subjects. The shaded area is formed by the lines representing two standard
deviations above and below the mean section area of the eight controls.

488

J.S. Holdstock and Others

trials per session, 20 LEDs were presented twice within a session. The repeated LEDs were
varied between tasks and between subjects. For two of the three conditions (‘allocentric’ and
‘egocentric’) the retention interval was filled as the subjects were required to read a passage
of prose. This distractor task was intended to preclude the use of a verbal mediating strategy
to help solve the task and to ensure that it was not possible to fixate on the target region
through the delay period. A particular advantage of the memory task was that it was
demanding even at very short retention delays, which made it possible to avoid ceiling effects.
(1) ‘Short Delay’
In the ‘short delay’ condition the subject was asked to place a counter (22 mm diameter
with a 2 mm hole in the centre) in the exact position of the light immediately after it
disappeared (0 s) or after an unfilled delay of 3 s or 8 s during which the subject was
instructed to look away from the board. The subjects did not move during the test procedure.
Each subject was given one of two test sequences, which differed in the order in which the
LED lights were presented and the order of the retention delays. Alternate subjects were
given the same sequence. Patients RS and NM were given different sequences to each other.
(2) Allocentric
The allocentric condition used a manipulation similar to that used by Morris et al. (1996)
and Abrahams et al. (1996). Presentation of the LED sample was followed by a filled delay
of 5 s, 20 s or 60 s. Approximately three seconds before the end of the delay, subjects
walked to a location indicated by the experimenter which was either the other side of the
board or one of the two ends and then, in this new position, placed the counter (described
above) as close as they could to the position of the previously illuminated LED. As a
consequence of the subject’s changed position, the target retained its spatial relationship to
all exterior cues except for the subject. Subjects then returned to their original start position
before seeing the next target LED.
(3) Egocentric
In the egocentric condition testing took place in a blacked out room with the room lights
switched off. This manipulation ensured that no allocentric cues were available to contribute
to performance. At the end of the filled delay, during which subjects read using a torch, the
subject was asked to position a light pointer (Loewe Opta Luxitron 40) to indicate the location
of the previous LED. The light pointer was attached to the arm of an angle-poised lamp (with
the shade and bulb removed). This arrangement produced a stable circle of light on the board
that was approximately the same size as the counter used in the other conditions. The subject
was encouraged to stand in the same place throughout each trial. In order to reduce dark
adaptation and hence to help ensure that room cues and the edge of the board were not visible
to the subject whilst the target light was presented, the room lights were turned on between
each trial. Although this meant that subjects were not prevented from forming a mental image
of the room it ensured that visible allocentric cues were not available during target LED
presentation; therefore the position of the LED could not be encoded with reference to these
cues, so an egocentric frame of reference had to be used. Debriefing of control subjects
confirmed that they had used an egocentric frame of reference.
In all three conditions the position of the centre of the counter (or circle of light in the
egocentric task) was marked on a removable Perspex sheet. Each control subject completed
just one allocentric and one egocentric sequence. A number of trial sequences were constructed
which differed in the order in which target positions and delays were tested and each sequence
was used an equal number of times in each condition. The trial sequences which were presented
to NM in the allocentric and egocentric conditions were presented in the opposite conditions
for RS. All Korsakoff patients were tested on the ‘short delay’ condition but it was not possible
to test patient TC on the allocentric and egocentric conditions.
Finally, a subset of subjects were tested on a control task to determine if the amnesic

Spatial memory in amnesia

489

subjects were able to place the counter with the same degree of precision as the control
subjects when there was no memory load. Six control subjects, six Korsakoff subjects,
patient NM and RS were tested on this ‘simultaneous condition’, each receiving 15 trials
which corresponded to 15 different positions. For this, a single LED was illuminated and
the subject asked to position the counter as accurately as possible over the light. The light
remained visible while the subject positioned the counter. The position of the centre of
the counter was recorded on the sheet of Perspex in the same way as for the memory
tasks.

RESULTS
The results for all conditions of the spatial task were based on the distance of
the recalled target location from the centre of the actual target LED and, for each
condition, the mean spatial error was calculated for each delay. The scores of the
two medial temporal lobe amnesics were considered to be impaired if they were
more than 1.96 SDs worse than the control mean, giving a type 1 error probability
of 0.05, 2 tailed. All simple effect analyses used the method of Keppel (1973).
Simultaneous Condition
The mean error distance was 2.4 mm for the six control subjects and an
identical mean error was found for the six Korsakoff patients. Patient RS
produced a mean error of 2.9 mm and NM produced a mean error of 3.3 mm,
which were unimpaired.
‘Short Delay’ Condition
RS and NM
Both RS and NM were unimpaired in the ‘short delay’ condition. Table IV
shows their mean error scores for each delay of this task which were all within
1.96 SDs of the control mean.
TABLE IV

Mean Error Distance, with Standard Deviation in Parentheses, for the Korsakoff and Control Groups
and Mean Score for Patient RS and Patient NM for the Three Delays (0 s, 3 s, 8 s) of the ‘Short
Delay’ Condition of the Spatial Memory Task
Delays
Subject type
Korsakoff group
Control group
NM
RS

0s

3s

8s

2.1 (0.7)
1.5 (0.5)
1.1
0.9

3.0 (1.1)
2.4 (0.4)
2.3
1.8

3.6 (1.2)
2.8 (0.8)
3.9
1.7

490

J.S. Holdstock and Others

Fig. 3 – Graph showing the mean error score for the Korsakoff amnesics (N = 7) and the
controls (N = 14) for the three delays (0 s, 3 s, 8 s) of the ‘short delay’ condition of the spatial
memory task.

Korsakoff Group
An ANOVA with a between subject factor of group and within subject factor
of delay compared the performance of the control subjects and Korsakoff
patients at the three delays (0 s, 3 s, 8 s) of the ‘short delay’ condition (Figure
3). The Korsakoff patients were significantly worse in this condition than the
controls (F = 6.76; d.f. = 1, 19; p < 0.05). There was no significant group by
delay interaction (F = .24; d.f. = 2, 38; p > 0.05), however, an analysis of simple
effects showed that the only significant difference between groups was at the
0 s delay (p < 0.05) and not at the 3 s and 8 s delays (p > 0.05). Finally, there
was a significant effect of delay (F = 34.0; d.f. = 2, 38; p < 0.001) indicating
that, for both groups, accuracy decreased as delay length increased.
Allocentric and Egocentric Conditions
RS and NM
Both patients were impaired in the allocentric condition at the longer delays
(see Table V). Neither RS nor NM were impaired at the 5 s delay. In contrast,
RS was clearly impaired at 20 s and there was a trend for him to also be
impaired at 60 s. Similarly, NM was impaired at the 60 s delay but unlike RS
he was not impaired at the 20 s delay.

Spatial memory in amnesia

491

TABLE V

Mean Error Distance, with Standard Deviation in Parentheses, for the Korsakoff and Control Groups
and Mean Score for Patient RS and Patient NM for the Three Delays (5 s, 20 s, 60 s) of the
Allocentric Condition and the Egocentric Condition of the Spatial Memory Task
Allocentric task delay

Egocentric task delay

Subject type
Korsakoff group
Control group
NM
RS

5s

20 s

60 s

5s

20 s

60 s

11.3 (6.0)
5.9 (1.6)
7.6
6.8

14.0 (6.2)
6.7 (1.6)
6.11
11.5

17.2 (6.4)
8.3 (2.4)
13.4
12.1

10.0 (4.1)
7.1 (1.7)
4.3
9.7

12.2 (5.0)
7.8 (2.7)
8.6
14.0

19.2 (5.1)
9.7 (2.2)
9.4
16.1

In the egocentric condition the two patients showed conflicting patterns of
performance. As shown in Table V, RS was impaired in this condition at the
20 s and 60 s delays wheras NM was not impaired at any delay; his performance
was actually better than the control mean at two delays.
In order to investigate further the rate of forgetting in each condition we
calculated for the controls, NM and RS the difference in mean error distance
between successive delays. In the allocentric condition accelerated forgetting
(i.e. a difference score 1.96 SDs greater than the control mean difference) was
shown by NM between delays of 20 and 60 seconds only (difference score was
2.8 SDs larger than the control mean difference), whereas RS showed
accelerated forgetting between the first two delays only (difference score was
3.75 SDs larger than the control mean difference). In the egocentric condition,
the difference scores of both patients were within 1.96 standard deviations of the
mean control difference for both pairs of delays. However, the difference
between error distance at 60 and 5 seconds was significantly larger than the
control difference for RS (2.9 SDs larger), but not NM.
Korsakoff Group
An ANOVA was performed with a between subject factor of group and
within subject factors of condition and delay in order to compare the
performance of the control subjects and the Korsakoff patients on the allocentric
and egocentric conditions at delays of 5 s, 20 s and 60 s (Figure 4).
The analysis revealed a highly significant effect of group (F = 22.2; d.f. = 1,
18; p < 0.001) but this did not interact with condition (F = 1.72; d.f. = 1, 18;
p > 0.05). This indicates that the Korsakoff patients were similarly impaired in
both conditions. This significant difference from the control subjects was
confirmed in separate analyses of scores from the allocentric condition (F = 6.09;
d.f. = 1, 18; p < 0.05) and the egocentric condition (F = 16.42; d.f. = 1, 18;
p < 0.05). The main effect of condition was not significant (F = .54; d.f. = 1, 18;
p > 0.05), showing that the overall scores in the allocentric and egocentric
conditions were similar. Simple effects analysis of the nonsignificant group-bycondition interaction showed that, for each subject group, there was no
significant difference between performance in the two conditions (both p > 0.05).
There was, however, a significant main effect of delay (F = 62.1; d.f. = 2, 36;
p < 0.05) which interacted significantly with group (F = 15.3; d.f. = 2, 36;

492

J.S. Holdstock and Others

Fig. 4 – Graph showing the mean error score for the Korsakoff amnesics (N = 6) and the
controls (N = 14) for the three delays (5 s, 20 s, 60 s) of the allocentric and egocentric conditions of
the spatial memory task.

p < 0.001). An analysis of simple effects revealed a significant effect of delay for
both subject groups (both p < 0.05), but the decline in performance over delay
was greater in the Korsakoff than control group (see Figure 4). There was also a
significant interaction between condition and delay (F = 3.44; d.f. = 2, 36;
p < 0.05). Simple effects indicated that the interaction was due to performance on
the egocentric and allocentric conditions diverging after the 20 s delay. This
condition-by-delay effect was particularly evident in the Korsakoff data but the
group-by-condition-by-delay interaction failed to reach significance (F = 2.5;
d.f. = 2, 36; p = 0.096). To control for the effects of large outlying scores that
may have resulted from a failure to consistently attend to stimuli, the analyses
were repeated using median scores. Identical results were obtained.
Test Correlations
Spearman rank correlations tested whether there was a relationship between
performance in the egocentric and allocentric conditions, and whether
performance on the spatial task correlated with any of the other psychometric
measures.
The first set of correlations showed that the overall performance of the

Spatial memory in amnesia

493

Korsakoff subjects on the egocentric and allocentric tasks was not significantly
correlated (rho = 0.49; p > 0.05). The scores of the control subjects did, however,
reveal a significant correlation between the two spatial tasks (rho = 0.70,
p < .05). The second set of correlations tested for a relationship in the Korsakoff
patient data between the mean score from the three delays on the ‘short delay’
condition of the spatial memory task, the allocentric condition or the egocentric
condition and scores on the tests of frontal lobe function which are listed in
Table II. No correlations were significant (all ps > 0.05). An additional set of
tests investigated correlations between the mean score of the Korsakoff patients
over the three delays of the ‘short delay’ spatial memory task and digit span,
visual span and the attention/concentration index of the WMS-R. No correlations
reached statistical significance (all ps > 0.05). A final set of correlations tested
for a relationship between performance on the allocentric and egocentric tasks
and scores on the WMS-R general memory index, WMS-R delayed memory
index and total score of the WRMT. None of these correlations reached statistical
significance (all ps > 0.05).
DISCUSSION
The results suggest that the Korsakoff patients have a possible deficit in both
attention and spatial memory. The Korsakoff patients were impaired at
remembering spatial positions even at the zero second delay and it was confirmed
that this deficit was not due to difficulty placing the counter where intended.
These results appear to suggest that the Korsakoff patients could not encode
spatial information as well as controls either due to impaired encoding
mechanisms or to an attentional deficit. The literature concerning encoding by
Korsakoff amnesics is conflicting so it has not been clearly established whether
this patient group has widespread cognitive deficits that affect information
encoding. Mayes et al. (1993) and Warrington and Baddeley (1974) argue that
encoding is intact in Korsakoff amnesics whereas others, such as Cave and
Squire (1992), have found an impairment of spatial short-term memory at delays
as short as zero seconds. The apparent discrepancy between the results of these
studies may not be surprising given that a variety of tasks have been used which
may tap slightly different aspects of spatial memory. Thus the encoding success
in these patients may depend on the precise nature of the test materials and
conditions. The study which has used the most similar task to ours found no
evidence of an impairment in a group of amnesics, including patients with
Korsakoff’s syndrome, in the ability to recall the location of a dot on a lightboard
after short delays (Warrington and Baddeley, 1974). Differences with the present
task include the use of a considerably smaller test board than ours (31 × 31 cm)
and a shorter sample presentation (one second rather than two second). The
impaired performance for our patients was therefore not consistent with these
findings. Cave and Squire (1992) proposed that Korsakoff amnesics show an
excessive degree of distractibility, caused by frontal lobe damage, which may
often cause them to fail to notice the stimulus so that they show impairments
even at short delays. Our findings were not consistent with this proposal. The

494

J.S. Holdstock and Others

analysis of median scores suggested that the results obtained in the present study
could not be explained by sporadic lapses of attention and a correlational analysis
gave no indication of a relationship between spatial short-term memory and
frontal lobe function, although the modest group sizes did limit the power of this
latter analysis. Other researchers suggest that Korsakoff patients show a
generalised impairment in attention or perception due to dysfunction of
intralaminar thalamic nuclei or a reduction of brain norepinephrine (Mair et al.,
1986; Paller, Acharya, Richardson et al., 1997). The possibility that our
Korsakoff group may have had a more diffuse attentional deficit was not
addressed by the analysis of median scores and cannot be ruled out.
The results from the allocentric and egocentric conditions showed that the
Korsakoff amnesics were equally and severely impaired in the two conditions
and, in both cases, they appeared to lose information at an accelerated rate. This
latter finding is consistent with related studies of Korsakoff amnesics (Holdstock
et al., 1995; Downes, Holdstock, Symons et al., 1998) which have interpreted
accelerated forgetting as indicating a deficit in memory consolidation (Downes
et al., 1998; Mayes, 1995). In the present case scaling problems, due to
unmatched control and amnesic performance at the initial delay, make it difficult
to determine whether the Korsakoff forgetting rate is truly accelerated. However,
if this accelerated forgetting is genuine, and evidence from previous studies
(Holdstock et al., 1995; Downes et al., 1998) indicates that this is likely, then
this patient group has a memory impairment for allocentric and egocentric
spatial information in addition to a possible encoding deficit suggested by their
performance on the short delay tasks. Evidence suggests that an attentional
deficit does not necessarily lead to accelerated forgetting of verbal and
nonverbal material. Cahn and Marotte’s (1995) study of children with Attention
Deficit Hyperactivity Disorder showed that, although this disorder impairs
attention and organisation of information, which impaired nonverbal memory
when tested immediately, forgetting of verbal and nonverbal material was not
accelerated. We, therefore, consider that, although the Korsakoff patients may
have shown poor encoding of position information, their pattern of performance
on the allocentric and egocentric spatial memory tasks provides important
information concerning these two types of spatial memory in this patient group.
Unlike two patients with medial temporal lobe damage, NM (referred to in this
paper) and YR (Holdstock et al., submitted), who were only significantly
impaired on the allocentric spatial task, the Korsakoff patients were impaired
and showed apparent accelerated forgetting on both the allocentric and
egocentric spatial memory tasks.
Data from animal studies suggest that the hippocampus, mamillary bodies
and anterior thalamic nuclei form a critical system necessary for allocentric
spatial memory (Aggleton and Saunders, 1997). This suggests that the Korsakoff
amnesics’ impairment on the allocentric spatial task may be associated with
abnormalities in the mamillary bodies which invariably show pathology in this
disease. While it is possible that egocentric memory is also dependent on the
mamillary bodies, animal research has not implicated them in this aspect of
spatial memory. Korsakoff’s syndrome has been recently found to result in
dysfunction to a number of brain structures (Kopelman, 1995; Paller et al., 1997)

Spatial memory in amnesia

495

beyond those observed in the classic neuropathology. PET has revealed
pathology of the frontal and parietal cortices of Korsakoff patients (Paller et al.,
1997) and both of these regions have been implicated in egocentric spatial
processing (Fuster, 1980; McDaniel, Via, Smith et al., 1995; Niki, 1975; Pohl,
1973; Save and Moghaddam, 1996; Semmes et al., 1963). This may account for
the egocentric memory deficit. Although results of our correlational analysis did
not support a frontal explanation, the modest size of the amnesic group did limit
the power of such analyses.
The results of the patients with medial temporal lobe damage, RS and NM,
will now be considered. These two patients were both unimpaired on the ‘short
delay’ condition of the spatial memory task suggesting that encoding of spatial
information and working memory for spatial information remain intact following
bilateral medial temporal lobe damage. This finding is consistent with a number
of studies reporting normal encoding of spatial information following medial
temporal lobe damage (Mayes, Downes, Shoqeirat et al., 1993; Warrington and
Baddeley, 1974), bilateral damage restricted to the hippocampus (Cave and
Squire, 1992) and right temporal lobectomy, including the hippocampus (Smith
and Milner, 1989).
Both patients RS and NM were impaired on the allocentric spatial memory
task but only RS was impaired on the egocentric task. The control subjects’
performance did not differ significantly on the two tasks but, if anything, they
found the allocentric task slightly easier than the egocentric task. Therefore,
greater impairments on the allocentric than the egocentric task cannot be
attributed to the relative difficulty of the tasks. The impaired performance of
patients RS and NM on the allocentric task suggests that medial temporal lobe
structures are critical for this aspect of spatial memory. This result is consistent
with animal studies (Feigenbaum and Rolls, 1991; Jarrard, 1993; Morris et al.,
1982; Muller et al., 1987; O’Keefe, 1991; O’Keefe and Conway, 1978, 1980;
O’Keefe and Dostrovsky, 1971; O’Keefe and Nadel, 1978; O’Keefe and
Speakman, 1987) and human studies (Abrahams et al., 1997; Goldstein et al.,
1989; Holdstock et al., submitted; Maguire, Burgess, Donnett et al., 1998;
Maguire, Burke, Phillips et al., 1996; Maguire, Frackowiak and Frith, 1996,
1997; Morris et al., 1996) which have implicated the medial temporal lobe,
including the hippocampus, in allocentric spatial memory. The severity of their
allocentric spatial memory deficit was no greater than that shown on the same
task by patient YR (Holdstock et al., submitted), who had relatively selective
hippocampal damage. This result was particularly interesting in the light of a
recent study (Bohbot et al., 1998) which found that right parahippocampal
lesions, but not right hippocampal lesions, impaired performance on a human
analogue of the Morris Water maze which was probably tapping allocentric
spatial information. That study suggested that under some circumstances the
parahippocampal cortex itself may play a critical role in spatial memory which
is not shared by that part of the hippocampus which was damaged in Bohbot et
al.’s right hippocampally lesioned patients. However, the results from our study
suggested that this was not the case for the allocentric and egocentric spatial
memory tapped by our tasks. Our results showed that additional damage to the
parahippocampal gyrus did not accentuate the allocentric memory deficit for

496

J.S. Holdstock and Others

place information found after more selective hippocampal damage which
affected the entire length of the hippocampus. This finding is consistent with a
serial relationship between the parahippocampal gyrus and the hippocampus
which is suggested by anatomical evidence (Aggleton and Brown, 1999).
Both patients showed accelerated forgetting of the allocentric spatial
information. This pattern of results suggests a deficit in the consolidation of
visuo-spatial information which is consistent with the hypothesis that the
hippocampus mediates the consolidation and storage of associative information
(Mayes, 1995). It might be argued that, given the very short initial delay, the
pattern of scores could also be interpreted in terms of the differential
contribution to performance of intact short-term memory and impaired long-term
memory at different delays. This latter explanation is unlikely, at least for
patient NM, whose performance was still matched to the controls at 20 seconds,
but who showed accelerated forgetting between this delay and 60 seconds.
Although the duration of visual short-term memory has not been established, the
duration of verbal short-term memory was estimated by experimental methods to
be two seconds or less (Muter, 1980) when rehearsal is not possible and amnesic
patients with normal short-term memory (assessed by digit span) were severely
impaired at verbal recall after a filled delay of only 15 seconds (Isaac and
Mayes, 1999). These two sources of evidence indicate that the duration of verbal
short-term memory is less than 15 seconds. Therefore, if similar mechanisms
underlie visuo-spatial short-term memory it is unlikely that it is still contributing
to performance after 20 seconds (see Downes et al., 1998; Isaac and Mayes,
1999, for a discussion of these issues). If this is the case, the accelerated
forgetting shown by NM is most likely to be due to a consolidation deficit.
The discrepancy in the pattern of performance of the two patients with
medial temporal lobe damage on the egocentric task may be attributed to their
different patterns of brain pathology which have been revealed by detailed
investigation of their structural MRI scans. A number of brain regions were
found to be damaged to a greater extent in RS than NM and so may account for
RS’s impairment on the egocentric task. (1) Within the medial temporal lobes,
the anterior part of the hippocampus was found to be smaller in RS than NM.
(2) RS had more cortical atrophy than NM, particularly of the temporal lobes.
(3) He also had less white matter than NM which may suggest poorer
transmission of information between brain regions in RS than NM. (4) RS’s
lateral and third ventricles were enlarged which may suggest damage to
diencephalic structures such as the thalamus. Surprisingly, there was no evidence
that RS had more atrophy than NM to the brain regions which have been
previously associated with egocentric spatial processing in the animal literature,
that is, the frontal cortex and parietal cortex (Fuster, 1980; McDaniel, Via,
Smith et al., 1995; Niki, 1975; Pohl, 1973; Semmes et al., 1963). In fact, RS
had much less atrophy to the parietal lobe than NM who was unimpaired on the
egocentric task. RS’s egocentric spatial memory deficit may therefore be due to
damage to the anterior hippocampus, atrophy of the temporal lobes, reduction of
white matter or possibly damaged diencephalic structures.
Patient NM’s unimpaired performance on the egocentric condition was
consistent with previous animal studies which have indicated a much larger

Spatial memory in amnesia

497

involvement of the hippocampus in allocentric than egocentric spatial memory
(Feigenbaum and Rolls, 1991; Jarrard, 1993; Morris et al., 1982; Muller et al.,
1987; O’Keefe, 1991; O’Keefe and Conway, 1978, 1980; O’Keefe and
Dostrovsky, 1971; O’Keefe and Nadel, 1978; O’Keefe and Speakman, 1987) and
studies of patients which have also implicated the right temporal lobe (Goldstein
et al., 1989) and the hippocampus bilaterally (Holdstock et al., submitted) in
allocentric but not egocentric spatial memory. Previous work suggests that the
parahippocampal gyrus may be necessary when the positions of particular objects
are encoded within an egocentric framework (Milner et al., 1997; Owen et al.,
1995). However, NM’s results suggest that the hippocampus, the amygdala and
the parahippocampal gyrus including the perirhinal cortex and entorhinal cortex
are not necessary for normal egocentric spatial memory when position alone has
to be remembered. An interesting finding to arise from NM’s data was that
considerable atrophy to the parietal lobes was not sufficient to impair egocentric
spatial memory. This finding appears inconsistent with previous studies which
have implicated the parietal lobe in egocentric spatial memory (e.g. Berthoz,
1997; McDaniel, Via, Smith et al., 1995; Save and Moghaddam, 1996). However,
focal damage is much more likely to cause serious disruption of neural circuits
than atrophy and our results do not exclude the possibility that more focal lesions
to the right areas of the parietal cortex would cause such a memory impairment.
The results from the allocentric and egocentric spatial memory tasks indicate
a possible dissociation between amnesia due to Korsakoff’s syndrome and
medial temporal lobe damage. A number of studies have reported no significant
difference between the disproportionately impaired performance of Korsakoff
and medial temporal lobe amnesic patients on spatial memory tests (Kopelman,
Stanhope and Kingsley, 1997; Mayes et al., 1991; Shoqeirat and Mayes, 1991).
However, Chalfonte, Verfaellie, Johnson et al. (1996) report that when object
recognition performance was matched to that of the controls, memory for object
position was not disproportionately impaired in Korsakoff patients but was
disproportionately impaired after medial temporal lobe damage when tested
incidentally. Kopelman et al. (1997) also report a trend for their Korsakoff
patients to be less impaired at object-position memory than medial temporal lobe
amnesics when matched to controls for object recognition. The difference
between Korsakoff and medial temporal lobe patient performance in the present
study was in the opposite direction to this. The Korsakoff patients showed a
considerable impairment of both allocentric and egocentric spatial memory
whereas medial temporal lobe damage only consistently impaired allocentric
spatial memory. These results suggest that Korsakoff patients are not less
impaired than medial temporal lobe amnesics on all spatial memory tasks and
that this patient group may actually have additional impairments affecting
performance on the egocentric spatial memory task and encoding.
To summarise, allocentric and egocentric spatial memory were assessed
separately in amnesic patients with either medial temporal lobe or diencephalic
damage. The main findings of the study suggested that medial temporal lobe
damage does not affect short-term spatial memory, but it does impair memory
for allocentric spatial information. However, this impairment appeared to be no
greater than that found following more selective hippocampal damage

498

J.S. Holdstock and Others

(Holdstock et al., submitted). The effect of medial temporal lobe damage on
egocentric spatial information needs to be clarified by testing further patients in
whom we can be confident that dysfunction is restricted to the medial temporal
lobes, but the present findings suggest that these structures, other than possibly
the anterior hippocampus, may not be critical for egocentric spatial memory. The
role of the diencephalic structures in allocentric and egocentric spatial memory
is still to be fully determined. Our results suggest that Korsakoff amnesics are
equally impaired under allocentric and egocentric spatial memory task
conditions, but interpretation of the results is complicated by the possibility of
dysfunction to brain areas other than the diencephalic region (Paller et al., 1997).
It will, therefore, be critical to test patients with selective damage to
diencephalic structures such as the mamillary bodies and anterior thalamic nuclei
in order to further illuminate the role of these structures in allocentric and
egocentric spatial memory.
Acknowledgements. Testing of control subjects and Korsakoff patients was carried out
while JSH was a member of the Department of Psychology, University of Durham, Durham,
U.K.
The authors wish to thank Mr. P.R. Hunt for designing and constructing the lightboard.
This research was supported by a grant from the Medical Research Council of the United
Kingdom awarded to J.P. Aggleton and Grant No. G9300193 from the Medical Research
Council of the United Kingdom awarded to A.R. Mayes.
REFERENCES
ABRAHAMS, S., PICKERING, A., POLKEY, C.E., and MORRIS, R.G. Spatial memory deficits in patients with
unilateral damage to the right hippocampal formation. Neuropsychologia, 35: 11-24, 1997.
AGGLETON, J.P., and BROWN, M.W. Episodic memory, amnesia and the hippocampal-anterior thalamic
axis. Behavioral and Brain Sciences, 1999 (in press).
AGGLETON, J.P., and SAUNDERS, R.C. The relationships between temporal lobe and diencephalic
structures implicated in anterograde amnesia. Memory, 5: 49-71, 1997.
AGUIRRE, G.K., DETRE, J.A., ALSOP, D.C., and D’ESPOSITO, M. The parahippocampus subserves
topographical learning in man. Cerebral Cortex, 6: 823-829, 1996.
AGUIRRE, G.K., ZARAHN, E., and D’ESPOSITO, M. Neural components of topographical representation.
Proceedings of the National Academy of Sciences of the United States of America, 95: 839-846,
1998.
ANDERSON, R.A. Multimodal integration for the representation of space in the posterior parietal cortex.
Philosophical Transactions of the Royal Society of London, series B, 352: 1421-1428, 1997.
BADDELEY, A., EMSLIE, H., and NIMMO-SMITH, I. Doors and People. Thames Valley Test Company,
1994.
BARTZOKIS, G., MINTZ, J., MARX, P., OSBORN, D., GUTKIND, D., CHIANG, F., PHELAN, C.K., and MARDER,
S.R. Reliability of in-vivo volume measures of hippocampus and other brain structures using MRI.
Magnetic Resonance Imaging, 11: 993-1006, 1993.
BENTON, A.L. Differential behavioural effects in frontal lobe disease. Neuropsychologia, 6: 53-60, 1968.
BERTHOZ, A. Parietal and hippocampal contribution to topokinetic and topographic memory.
Philosophical Transactions of the Royal Society of London, series B, 352: 1437-1448, 1997.
BOHBOT, V., KALINA, M., STEPANKOVA, K., SPACKOVA, N., PETRIDES, M., and NADEL, L. Lesions to the
right parahippocampal cortex cause navigation memory deficits in human. NeuroImage, 5: 626,
1997.
BOHBOT, V., KALINA, M., STEPANKOVA, K., SPACKOVA, N., PETRIDES, M., and NADEL, L. Spatial memory
deficits in patients with lesions to the right hippocampus and to the right parahippocampal cortex.
Neuropsychologia, 36: 1217-1238, 1998.
CAHN, D.A., and MARCOTTE, A.C. Rates of forgetting in attention deficit hyperactivity disorder. Child
Neuropsychology, 1: 158-163, 1995.
CAVE, C.B., and SQUIRE, L.R. Equivalent impairment of spatial and nonspatial memory following
damage to the human hippocampus. Hippocampus, 1: 329-340, 1991.
CAVE, C.B., and SQUIRE, L.R. Intact verbal and nonverbal short-term memory following damage to the

Spatial memory in amnesia

499

human hippocampus. Hippocampus, 2: 151-164, 1992.
CHALFONTE, B.L., VERFAELLIE, M., JOHNSON, M.K., and REISS, L. Spatial location memory in amnesia:
Binding item and location information under incidental and intentional encoding conditions.
Memory, 4: 591-614, 1996.
CORKIN, S. Lasting consequences of bilateral medial temporal lobectomy: Clinical course and
experimental findings in H.M. Seminars in Neurology, 4: 249-259, 1984.
DELAY, J., and BRION, S. Le Syndrome de Korsakoff. Paris: Masson, 1969.
DE RENZI, E. Disorders of spatial orientation. In J.A.M. Frederiks (Ed.), Handbook of Clinical
Neurology, vol. 1. Amsterdam: Elsevier, 1985, pp. 405-421.
DOWNES, J.J., HOLDSTOCK, J.S., SYMONS, V., and MAYES, A.R. Do amnesics forget colour pathologically
fast? Cortex, 34: 337-355, 1998.
EPSTEIN, R., and KANWISHER, N. A cortical representation of the local environment. Nature, 392: 598601, 1998.
FEIGENBAUM, J.D., POLKEY, C.E., and MORRIS, R.G. Deficits in spatial working memory after unilateral
temporal lobectomy in man. Neuropsychologia, 34: 163-176, 1996.
FEIGENBAUM, J.D., and ROLLS, E.T. Allocentric and egocentric spatial information processing in the
hippocampal formation of the behaving primate. Psychobiology, 19: 21-40, 1991.
FUSTER, J.M. The Prefrontal Cortex: Anatomy, Physiology, and Neuropsychology of the Frontal Lobe.
New York: Raven, 1980.
GOODGLASS, H., and KAPLAN, E. Assessment of cognitive deficit in brain injured patients. In M.S.
Gazzaniga (Ed.), Handbook of Behavioural Neurobiology, Vol. 2: Neuropsychology. New York:
Plenum Press, 1979, pp. 3-22.
GOLDSTEIN, L.H., CANAVAN, A.G.M., and POLKEY, C.E. Cognitive mapping after unilateral temporal
lobectomy. Neuropsychologia, 27: 167-177, 1989.
HABIB, M., and SIRIGU, A. Pure topographical disorientation: A definition and anatomical basis. Cortex,
23: 73-85, 1987.
HEATON, R.K. Wisconsin Card Sorting Test Manual. Odessa, Florida: Psychological Assessment
Resources Inc., 1981.
HOLDSTOCK, J.S., SHAW, C., and AGGLETON, J.P. The performance of amnesic patients on tests of delayed
matching-to-sample and delayed matching-to-position. Neuropsychologia, 33: 1583-1596, 1995.
HOLDSTOCK, J.S., MAYES, A.R., CEZAYIRLI, E., ISAAC, C.L., AGGLETON, J.P., and ROBERTS, J.N. A
comparison of egocentric and allocentric spatial memory in a patient with partial selective
hippocampal damage. Submitted to Neuropsychologia.
HUNKIN, N.M., PARKIN, A.J., and LONGMORE, B.E. Aetiological variation in the amnesic syndrome:
Comparisons using the list discrimination task. Neuropsychologia, 32: 819-825, 1994.
ISAAC, C.L., and MAYES, A.R. Rate of forgetting in amnesia I: Recall and recognition of prose. Journal
of Experimental Psychology: Learning, Memory and Cognition, 1999 (in press).
JARRARD, L.E. On the role of the hippocampus in learning and memory in the rat. Behavioural and
Neural Biology, 60: 9-26, 1993.
JONES-GOTTMAN, M., and MILNER, B. Design fluency: The invention of nonsense drawings after focal
cortical lesions. Neuropsychologia, 15: 653-674, 1977.
JOYCE, E.M., and ROBBINS, T.W. Frontal lobe function in Korsakoff and non-Korsakoff alcoholics:
Planning and spatial working memory. Neuropsychologia, 29: 709-723, 1991.
KARNATH, H.-O. Spatial orientation and the representation of space in the posterior parietal cortex.
Philosophical Transactions of the Royal Society of London, series B, 352: 1411-1420, 1997.
KEPPEL, G. Design and Analysis. A Researcher’s Handbook. Englewood Cliffs, N.J.: Prentice-Hall, Inc.,
1973.
KOPELMAN, M.D. The Korsakoff syndrome. British Journal of Psychiatry, 166: 154-173, 1995.
KOPELMAN, M.D., STANHOPE, N., and KINGSLEY, D. Temporal and spatial context memory in
patients with focal frontal, temporal lobe, and diencephalic lesions. Neuropsychologia, 35: 15331545, 1997.
LANDIS, T., CUMMINGS, J.L., BENSON, D.F., and PALMER, E.P. Loss of topographical familiarity: An
environmental agnosia. Archives of Neurology, 43: 132-136, 1986.
MACANDREW, S.B.G., and JONES, G.V. Spatial memory in amnesia: Evidence from Korsakoff patients.
Cortex, 29: 235-249, 1993.
MAGUIRE, E.A., BURGESS, N., DONNETT, J.G., FRACKOWIAK , R.S.J., FRITH, C.D., and O’KEEFE, J.
Knowing where and getting there: A human navigation network. Science, 280: 921-924, 1998.
MAGUIRE, E.A., BURKE, T., PHILLLIPS, J., and STAUNTON, H. Topographical disorientation following
unilateral temporal lobe lesions in humans. Neuropsychologia, 34: 993-1001, 1996.
MAGUIRE, E.A., FRACKOWIAK, R.S.J., and FRITH, C.D. Learning to find your way – A role for the human
hippocampal region. Proceedings of the Royal Society of London. Series B, Biological Sciences,
263: 1745-1750, 1996.
MAGUIRE, E.A., FRACKOWIAK, R.S.J., and FRITH, C.D. Recalling routes around London: Activation of the
right hippocampus in taxi drivers. Journal of Neuroscience, 17: 7103-7110, 1997.

500

J.S. Holdstock and Others

MAGUIRE, E.A., FRITH, C.D., BURGESS, N., DONNETT, J.G., and O’KEEFE, J. Knowing where things are:
Parahippocampal involvement in encoding object locations in virtual large-scale space. Journal of
Cognitive Neuroscience, 10: 61-76, 1998.
MAIR, R.G., DOTY, R.L., KELLY, K.M., WILSON, C.S., LANGLAIS, P.J., MCENTEE, W.J., and VOLLMECKE,
T.A. Multimodal sensory discrimination deficits in Korsakoff’s Psychosis. Neuropsychologia, 24:
831-839, 1986.
MAYES, A.R. Memory and amnesia. Behavioural Brain Research, 66: 29-36,, 1995.
MAYES, A.R., DOWNES, J.J., SHOQEIRAT, M., HALL, C., and SAGAR, H. Encoding ability is preserved in
amnesia: Evidence from a direct test of encoding. Neuropsychologia, 31: 745-759, 1993.
MAYES, A.R., MEUDELL, P.R., and MACDONALD, C. Disproportionate intentional spatial-memory
impairments in amnesia. Neuropsychologia, 29: 771-784, 1991.
MCDANIEL, W.F., VIA, J.D., SMITH, J.S., WELLS, D.L., FU, J.J., BISHOP, J.F., BOYD, P.A., and LEDESMA,
H.M. Unilateral injury of posterior parietal cortex and spatial learning in hooded rats. Behavioural
Brain Research, 70: 165-179, 1995.
MCFIE, J., and THOMPSON, J.A. Picture arrangement: A measure of frontal lobe function. British Journal
of Psychiatry, 121: 547-552, 1972.
MILNER, B., JOHNSRUDE, I., and CRANE, J. Right medial temporal-lobe contribution to object-location
memory. Philosophical Transaction of the Royal Society of London. Series B, Biological Sciences,
352: 1469-1474, 1997.
MORRIS, R.G.M., GARRUD, P., RAWLINS, J.N.P., and O’KEEFE, J. Place navigation impaired in rats with
hippocampal lesions. Nature, 297: 681-683, 1982.
MORRIS, R.G., PICKERING, A., ABRAHAMS, S., and FEIGENBAUM, J.D. Space and the hippocampal
formation in humans. Brain Research Bulletin, 40: 487-490, 1996.
MULLER, R.U., KUBIE, J.L., and RANCK, J.B. JR spatial firing patterns of hippocampal complex-spike
cells in a fixed environment. Journal of Neuroscience, 7: 1935-1950, 1987.
MUTER, P. Very rapid forgetting. Memory and Cognition, 8: 174-179, 1980.
NEAVE, N., NAGLE, S., and AGGLETON, J.P. Evidence for the involvement of the mamillary bodies and
cingulum bundle in allocentric spatial processing by rats. European Journal of Neuroscience, 9:
101-115, 1997.
NELSON, N.E., and WILLISON, J.R. The National Adult Reading Test (second edition). Windsor: NFERNelson, 1991.
NIKI, H. Differential activity of prefrontal units during right and left delayed response trials. Symposia of
the Fifth Congress of the International Primatological Society, 1975, pp. 475-486.
O’KEEFE, J. An allocentric spatial model for the hippocampal cognitive map. Hippocampus, 1: 230-235,
1991.
O’KEEFE, J., and CONWAY, D.H. Hippocampal place units in the freely moving rat: Why they fire when
they fire. Experimental Brain Research, 31: 573-590, 1978.
O’KEEFE, J., and CONWAY, D.H. On the trail of the hippocampal engram. Physiological Psychology, 8:
229-238, 1980.
O’KEEFE, J., and DOSTROVSKY, J. The hippocampus as a spatial map: Preliminary evidence from unit
activity in the freely-moving rat. Brain Research, 34: 171-175, 1971.
O’KEEFE, J., and NADEL, L. The Hippocampus as a Cognitive Map. Oxford, UK: Clarendon Press,
1978.
O’KEEFE, J., and SPEAKMAN, A. Single unit activity in the rat hippocampus during a spatial memory task.
Experimental Brain Research, 68: 1-27, 1987.
OWEN, A.M., MILNER, B., PETRIDES, M., and EVANS, A.C. A specific role for the right parahippocampal
gyrus in the retrieval of object-location: A positron emission tomography study. Journal of
Cognitive Neuroscience, 8: 588-602, 1996.
PALLER, K.A., ACHARYA, A., RICHARDSON, B.C., PLAISANT, O., SHIMAMURA, A.P., REED, B.R., and
JAGUST, W.J. Functional neuroimaging of cortical dysfunction in alcoholic Korsakoff’s syndrome.
Journal of Cognitive Neuroscience, 9: 277-293, 1997.
PARKER, A., and GAFFAN, D. Mamillary body lesions in monkeys impair object-in-place memory:
Functional unity of the fornix-mamillary system. Journal of Cognitive Neuroscience, 9: 512-521,
1997.
PARKIN, A.J. Functional significance of etiological factors in human amnesia. In L.R. Squire and N.
Butters (Eds.), Neuropsychology of Memory. New York: Guilford Press, 1992.
PIGOTT, S., and MILNER, B. Memory for different aspects of complex visual scenes after unilateral
temporal- and frontal-lobe resection. Neuropsychologia, 31: 1-15, 1993.
POHL, W. Dissociation of spatial discrimination deficits following frontal and parietal lesions in
monkeys. Journal of Comparative and Physiological Psychology, 82: 227-239, 1973.
ROLLS, E.T., and O’MARA, S.M. View-responsive neurons in the primate hippocampal complex.
Hippocampus, 5: 409-424, 1995.
SAVE, E., and MOGHADDAM, M. Effects of lesions of the associative parietal cortex on the acquisition and
use of spatial memory in egocentric and allocentric navigation tasks in the rat. Behavioural

Spatial memory in amnesia

501

Neuroscience, 110: 74-85, 1996.
SEMMES, J., WEINSTEIN, S., GHENT, L., and TEUBER, H.L. Correlates of impaired orientation in personal
and extrapersonal space. Brain, 86: 747-772, 1963.
SHALLICE, T., and EVANS, M.E. The involvement of the frontal lobes in cognitive estimations. Cortex,
14: 294-303, 1978.
SHOQEIRAT, M.A., and MAYES, A.R. Disproportionate incidental spatial-memory and recall deficits in
amnesia. Neuropsychologia, 29: 749-769, 1991.
SMITH, M.L., and MILNER, B. The role of the right hippocampus in the recall of spatial location.
Neuropsychologia, 19: 781-793, 1981.
SMITH, M.L., and MILNER, B. Right hippocampal impairment in the recall of spatial location: encoding
deficit or rapid forgetting? Neuropsychologia, 27: 71-81, 1989.
SUTHERLAND, R.J., and RODRIGUEZ, A.J. The role of the fornix/fimbria and some related subcortical
structures in place learning and memory. Behavioural Brain Research, 32: 265-277, 1989.
VICTOR, M., ADAMS, R.D., and COLLINS, G.H. The Wernicke-Korsakoff Syndrome. Oxford: Blackwell,
1971.
WARRINGTON, E.K. Recognition Memory Test. NFER-NELSON Publishing Company, 1984.
WARRINGTON, E.K., and BADDELEY, A.D. Amnesia and memory for virtual location. Neuropsychologia,
12: 257-263, 1974.
Dr. J.S. Holdstock, Department of Clinical Neurology, University of Sheffield, N Floor, Royal Hallamshire Hospital, Sheffield, S10 2JF,
U.K. E-mail: j.holdstock@sheffield.ac.uk

(Received 25 May 1998; accepted 2 March 1999)