RIGHT-LEFT CONFUSION IN GERSTMANN'S SYNDROME:
A MODEL OF BODY CENTERED SPATIAL ORIENTATION
Michael Gold, John C. Adair, Daniel H. Jacobs and Kenneth M. Heilman
(Department of Neurology, University of Florida College of Medicine, Gainesville, FL;
Neurology Service, Veterans Affairs Medical Center, Gainesville, FL)

ABSTRACT

Gerstmann's syndrome encompasses the tetrad of finger agnosia, agraphia, acalculia and
right-left confusion and is associated with lesions of the dominant angular gyrus. The
localizing value of this syndrome has been questioned because multiple mechanisms can
account for each of the components of the syndrome. We present the case of a man who
developed Gerstmann's syndrome following a focal infarct of the left angular gyrus. The
patient's right-left confusion could not be accounted for by either an aphasia or a degraded
body schema. A series of experiments that investigated the patient's spatial mapping system
by progressively restricting the degrees of freedom for spatial rotation revealed an isolated
defect in deriving the relative position of an object along the horizontal axis. Defective
horizontal mapping can account for the other components of Gerstmann's syndrome because
they all share a common dependency on relative horizontal positioning.

INTRODUCTION

Joseph Gerstmann reported a patient with finger agnosia, agraphia and
acalculia following a left-sided parieto-occipital lesion in 1924 (Gerstmann,
1924). Although right-left confusion was not part of the initial description of
the left parieto-occipital syndrome, by 1930 Gerstmann (1930) had included it.
Subsequent reports of similar collections of deficits in association with lesions
of the left parieto-occipital junction led to the designation of this tetrad of clinical
finding as Gerstmann's syndrome. An active debate has followed for many years
regarding the localizing value or the clinical constancy of this syndrome. Reports
of isolated deficits or of partial combinations of the components of Gerstmann' s
syndrome have raised question about its value as a unitary clinical syndrome
(Benton, 1961; Critcheley, 1966; Heimburger, 1964).
One of the major criticism regarding Gerstmann' s syndrome is that each of
the defects in the classical tetrad might be multi-factorial in origin (Benton,
1977). Since the underlying defects in Gerstmann's syndrome have not been
fully explored, the mechanisms that may be common to some if not all of the
components of this syndrome remain t0 be defined.
Multiple theories .have been invoked to explain the right-1eft confusion
frequently noted in cases labelled as having Gerstmann's syndrome. According
·to Head (1926), right-left confusion may be due to a defect in symbolic thought.
Therefore., patients may have problems applying the verbal labels "right" and
Cortex, (1995) 31, 267-283

268

Michael Gold and Others

Source of
Representation

Representational
Systems

Spatial Rotation
Sytems

Verbal Label
Fig. I - A model of body-centered spatial orientation.

"left" to part of a body or to a part of an object. A second theory (Benton,
1959) posits a defect in a body schema such that one's knowledge about body
parts, including lateralization, is defective. Patients with a defective body schema
would have problems identifying certain body parts in addition to making right­
left errors when asked to identify lateralized body parts. A final theory (Benton
and Sivan, 1993) posits that there is a defect in spatial mapping. Spatial mapping
is thought to be necessary for the identification of the relative position of an
object confronting the patient or in the performance of cross-body commands.
Defects in both tasks are commonly found in patients with Gerstmann's
syndrome.

PROPOSED MODEL

We propose a model (see Figure 1) that can be used to explain the features
of the right-left confusion see in some patients with Gerstmann's syndrome. The
existence of mental representations of body-centered right/left (horizontal), up/
down (vertical) and near/far (radial) knowledge is central to our model. Spatial
representations, as suggested by Stengel ( 1944), incorporate complex knowledge
relating the position of external objects in space to one's body.
These representations may be present to some extent at birth, but as in the
case of horizontal representations, they may develop with experience. Because
changing a patient's orientation in respect to gravity has an effect on their vertical
spatial neglect (Mennemeier et al., 1995), we infer that vertical representations

Right-Left confusion in Gerstmann 's Syndrome

269

(knowledge about what is above/below one's own body) are referenced to
gravity.
We also posit that one's radial representations (knowledge of what is near/
far to one's own body) use the horizon as a reference. Finally, we agree that
self-body horizontal representations (knowledge of what is right/left on one's
own body) are acquired by integration of asymmetrical somatosensory signals
received from both sides of the body (Benton, 1959) or by the development of
hemispheric asymmetry and dominance superimposed onto a pre-existing body
schema.
In contrast to self-body orientation, the determination of other-body right/
left orientation depends on the development of spatial rotation systems that allow
other-body information to be integrated with previously established horizontal
representations, which now serve as a reference. According to a body-centered
model of spatial orientation, in order to label the sides of an object in space
one must have the ability to rotate the image of an object in three dimensions
to derive a body-centered view, which is then compared to the body-centered
representation. Once the comparison occurs and a laterality decision is made,
the appropriate verbal label or motor response is activated.
Spatial rotation involves the visualization of'· movement of an object or of
the viewer through three-dimensional space. This movement can be interpreted
as linear translation along one, two or three orthogonal axes and can be described
as movement along an axis, movement in a plane or movement in space,
respectively.
We report a case of a patient with an angular gryrus lesion who had all the
features of Gerstmann' s syndrome. We performed a series of experiments on
this patent designed to explore some of the previously mentioned theories of
right-left disorientation and to provide data for an integrated model of right-left
confusion based on a defect in spatial mapping. We investigated the patient's
spatial rotational abilities by presenting tasks that progressively restricted the
number of axes available for spatial mapping tasks in order to determine if
faulty spatial mapping could be limited to a single axis or a combination of
axes.
CASE REPORT

A 79-year-old, right-handed man was admitted after the onset of right-sided weakness.
The patient reported four similar, but transient (<30 min) episodes during the month preceding
hospitalization. The patient denied previous focal neurological deficits, head trauma or
seizures. The patients's past medical history was significant for ischemic heart disease and
coronary bypass surgery in 1990.
Neurological examination on admission revealed an alert and cooperative patient whose
spontaneous speech appeared to be normal. The patient was anosodiaphoric in relation to
his right-sided weakness, but there was no evidence. of Gerstmann's syndrome. The day after
admission the patient experienced a worsening of his symptoms, developing right-sided
hemiplegia, finger agnosia, right-left confusion, apractic agraphia and acalculia. The right­
left confusion noted during our clinical evaluation was notable for the increased difficulty
the patient experienced when asked to name or point to body parts of an examiner facing
him as compared to milder problems naming or pointing to parts of his own body. Magnetic
resonance imaging performed on the third day of hospitalization revealed a hyperintense

270

Michael Gold and Others

signal on T2-weighted images that corresponded to the left angular gyrus and its underlying
white matter.
STANDARDIZED NEUROPSYCHOLOGICAL TESTS

The patient was given a battery of standardized tests including the Western Aphasia
Battery .(Kertesz, 1979), Boston Naming Test (Kaplan, Goodglass and Weintraub, 1978),
Letter Fluency (Benton and Hamsher, 1976) and Category Fluency (Chertkow and Bub, 1990),
Wisconsin Card Sorting Test (Berg, 1948), Trail-Marking Test (Army Individual Test Battery,
1944), California Verbal Learning Test (Delis, Kramer, Kaplan et al., 1983), Continuous
Visual Memory Test (Trahan and Larrabee, 1983), digit-span subtest from the Wechsler
Adult Intelligence Scale-Revised (Wechsler, 1981), Money Road Map Test (Money, 1976)
and line bisection and target cancellation tests for hemispatial neglect (Heilman, Watson and
Valenstein, 1993). The patient's performance on these tests was compared to published norms
for each specific test. The patient achieved a normal aphasia quotient of 81 on the Western
Aphasia Battery; however, the test revealed mild problems with reading comprehension and
severe problems with calculations (chance performance level). The performance on the Boston
Naming Test was normal (56/60).
Letter and Category fluency tasks were both defective. Executive funcion as assessed by
the Wisconsin Card Sorting Test and by the Trail-Making Test was defective. Verbal memory
assessed by the California Verbal Learning Tes! was defective, however, non-verbal memory
as assessed by the Continuous Visual Memory Test was normal. The patient was only able
to achieve a digit span of four, which is considered defective. Line bisection and target
cancellation tests failed to reveal any hemispatial neglect. Directional sense as assessed by
the Money Road Map Test was absent. The number of errors committed by the patient (15/
32) matched the performance of normal seven- to eight-year-old children.
Although multiple domains of cognition are affected, the type of cognitive defects and
the natural history of the patient's illness fit in quite well with the description of patients
with dominant angular gyrus lesion (Benson, Cummings and Tsai, 1982). We also concluded
from this standardized testing that neither our patient's comprehension defects nor his language
output were affected enough to interfere with the investigation of his right/left confusion.
STATISTICAL METHOD

Statistical comparisons were performed using a two-sided Pearson Chi-square test for
homogeneity with one degree of freedom unless stated otherwise. Following Daniel's (Daniel,
1990) recommendations, Yate's correction is not used and data from normal controls are
pooled.

EXPERIMENT 1

Is Right-Left Confusion a Problem with the Misapplication of Verbal Labels?

In this ex,periment, we compared the accuracy in a task that required the
patient t0 name ~body parts or to point to a named body part. According to the
theory 'm ·verblli mislabelling, when ·asked to name a 1ateralized (right/left) body
part, he ·should perform abnormally. However, when asked to point :to a body
part, .he should perform normally. We ~compared the patient's accuracy in :naming
'and ~pointing to lateralized body parts to a ,group of four age-, sex- and
handedness-matched normal •controls.

Right-Left confusion in Gerstmann's Syndrome

271

Materials and Methods
The examiner pointed to lateralized body parts (e.g. right hand) and' asked the patient
or control subject to name that lateralized body part. The examiner then named the same
body parts (e;g. left foot) in a different order and asked the patient to point to the, named
part. Seven different parts were chosen from each side of the body for a total ef 14 test
items.

Results
The patient correctly named 10/14 body parts and correctly pointed to 12/
14 body parts. Chi-square analysis did not reveal a significant difference between
naming and pointing (X2 = 0.212, p = 0.648). Normal controls correctly named
a body part in 56/56 total trials and correctly pointed to a body part in 55/56
trials. Chi-square analysis revealed that the patient's accuracy in naming body
parts was worse than the normal controls (X 2 = 16.9, p < 0.001), and that his
ability to point to body parts was worse than normal (x2 =4.27, p=0.04). An
analysis of the errors shows that the patient made four body part errors and
two right-left errors.

Discussion
The results of this experiment need to be interpreted carefully. Normal
controls perform almost perfectly so one or two errors on the patient's part can
lead to statistically significant differences, whereas a single error on the part of
a control subject can lead to non-significant results. With this caveat in mind,
the results indicate that our patient's accuracy when naming lateralized (right­
left) body parts or pointing to lateralized (right-left) body parts was not the
same as normal controls. The patient's accuracy when naming body parts was
equivalent to his accuracy whyn pointing to lateralized body parts. Our patient
appears to have more difficulty naming body parts and pointing to named body
parts than do normal controls. If our subject's defect was primarily one of verbal
labeling, we would have expected naming errors, but no pointing errors. These
results suggest that a defect in verbal labeling cannot fully account for the
patient's right-left confusion. Although the patient's error rate was less than
10%, he made errors in both naming and pointing to lateralized body parts,
suggesting a mildly degraded body schema. The patient's performance, although
not perfect, indicates that he has relatively well preserved knowledge about his
own right/left orientation.
EPERIMENT 2

Is Right-Left Confusion Due to a Problem with a Degraded Body Schema?
In experiment 1, the patient made four body part errors and two right-left
errors. Since these errors could not be attributed to a defect in verbal labelling,
the possibility of a degraded body schema contributing to our patient's right­

272

Michael Gold and Others

left confusion needed to be explored further. According to the body schema
theory, the brain contains a representation of the body upon which right/left
knowledge is superimposed. Damage to the brain degrades this body schema,
resulting in a loss of knowledge about the body. Loss of body knowledge can
lead not only to the mislabelling of body parts, but also to right-left errors.
When asked to name parts of our own bodies, we might rely on unique
features or landmarks such as scars, moles, or freckles or afferent asymmetries
(self cues) to identify a body part without referencing the body schema. Since
one cannot use these alternate methods when identifying parts of another person's
body, there must be increased reliance on the body schema. Thus, a subject
with a degraded body schema should have more problems identifying body parts
on another person's body than on his/her own body.
Materials and Methods
The patient was asked to name or point to lateralized body parts on a person standing
in front and facing away from him (congruous position). In this position, there is no need
for any spatial mapping or rotation in order to identify a body part. By having the patient
identify body parts in a physically separate subject, self cuing is avoided. The body parts
used in this experiment were the same ones as in Experiment 1. The sequences of body
parts were presented in a randomized fashion.

Results
The patient correctly named 12114 body parts on another person (one body
part error and one right/left error) and correctly pointed to 14/14 body parts on
another person. Chi-square analyses comparing the patient's ability to name his
own body parts and those of a control subject as well as his ability to point
out body parts on himself and on a control subject did not reveal any differences
(naming: x2 =0.848, p=0.36; pointing: x2 =2.15, p=0.142). Lastly, there was
no difference between the patient's ability to name or point to body parts on
a control subject (x2 =2.15, p=0.142). Chi-square analysis comparing the
patient's accuracy in identifying parts of his own body or those of a control
subject by either naming or pointing also failed to demonstrate a difference
(x 2 =2.33, p=O.l27).

Discussion
The performance on this and the prior task suggests that our patient correctly
applies knowledge about body parts to someone else's body. The paucity of
errors provides further evidence that our patient's body schema is not severely
degraded.
EXPERIMENT 3

Does Spatial Rotation Affect Right-Left Orientation?
The patient appeared to have a particularly difficult time when asked to make
right-left body part decisions on a subject facing him. According to the body­

Right-Left confusion in Gerstmann 's Syndrome

273

centered orientation hypothesis, when one is confronted with an object that has
lateralized parts, but is not isomorphic, a spatial transformation is required to
bring the object into a body-centered view. This transformation allows the person
to make a right/left decision. If the patient failed to spatially rotate an object,
we predicted that he would make more errors when presented with an object
that required spatial rotation in order to be brought into a body-centered view
than when no spatial rotation was required.
Materials and Methods

Using the same technique as the previous experiments, we positioned a control subject
in front of and facing our patient (incongruous position). The patient was now asked to
name and point to the same 14 body parts as in the previous experiments. His performance
in the incongruous position was compared to his performance in the congruous position.

Results
The patient correctly named 5/14 lateralized (right/left) body parts and
correctly pointed to 6114 lateralized (right/left) body parts on the control subject
in the incongruous position. An analysis of the -errors indicated that they all
consisted in right/left or left/right errors, and that there were no body part errors.
Chi-square analysis of combined naming and pointing comparing the control­
in-congruous position and the control-in-incongruous position demonstrated a
significant difference in accurancy (x2 = 17.9, p < 0.001).
Discussion
The patient demonstrated significant problems identifying and naming
lateralized body parts in a task that required spatial mapping and rotation. In
the absence of a seriously degraded body schema or a verbal labeling defect,
this experiment suggests that our patient has a problem transforming other-body
spatial information to a body-centered view. This problem may reflect a general
problem with spatial transformations or may indicate a case of defective spatial
transformation limited to a particular axis or plane. If a generalized defect in
the rotation of mental images were the basis of the patient's problem, we would
predict that he would misidentify objects presented in a spatially rotated form.
EXPERIMENT 4

Is Non-canonical Object Recognition Impaired in Right-Left Confusion?
Warrington and Taylor (1973) studied several patients who could identify
objects presented in their canonical form (e.g. a length-wise view of an iron),
but were unable to identify objects presented in a non-canonical view (e.g. a
head-on view of an iron). Since focal brain damage can impair systems involved
in the spatial rotation of an object's image, the possibility of just such an
impairment in our patient had to be excluded.

274

Michael Gold and Others

Materials and Methods

We presented the patient with black and white pictures of nine common household items
in their canonical form. For five of the nine objects, two additional pictures of these items,
after rotation within a plane, were also shown. For the remaining four items, a single rotated
picture was displayed for a total of nine canonical presentations and 14 non-canonical
presentations.
The pictures were presented in a pseudorandomized fashion such that pictures of the
same item were never consecutive.

Results
The patient correctly identified 9/9 (100%) canonically presented objects and
12114 (85.7%) of the non-canonical pictures. Chi-square analysis did not reveal
any differences between the items used and their presentation (rotated/canonical)
in this task (x2 =0.184, p=0.67). Normal controls (Warrington and James, 1988)
can identify 90%-95% of objects presented canonically and 65%-80% of objects
presented non-canonically.
Discussion
The patient had no difficulty identifying objects that were presented in a
spatially rotated (non-canonical) form. This suggests that the spatial rotation
systems that map an object to its canonical form are independent of systems
for object-to-body-centered mapping. Although both object- and body-centered
rotations may share a common core of functions in spatial mapping systems,
the results of this experiment suggest that labeling the lateralized (right-left)
portions of an object goes beyond the ability to derive its canonical view.
EXPERIMENT 5

Is Defective Spatial Mapping or Rotation Limited to a Single Axis?
To this point we have determined that our patient has right-left confusion
that cannot be accounted for by defective verbal labelling, a degraded body
schema or a general mental rotation deficit. The patient's problem seems to be
a deficit in mapping lateralized stimuli onto a body-centered view. To learn if
his body-centered mapping defect is due to defective rotation through a single
axis or occurs on more than one axis, we devised an experimental technique
patterned after the one used by Stengel (1944) involving spatial mapping on a
limited number of axes.
The tecnique (see Figure 2) involves the use of three coins, each of which
could be moved independently, placed in a straight line, and presented along
each axis of space (horizontal, radial and vertical). The initial task required the
patient to determine the relative position of one coin (the target coin) to a second
coin (the reference coin). For example, along the horizontal axis, the patient
would be asked: "Is the nickel to the right or to the left of the quarter?". This

Right-Left confusion in Gerstmann 's Syndrome

275

•A
0

8

c
Fig. 2 - The coin task used for spatial rotation. Is coin A to the left or right of coin B?

task was repeated in the radial axis by asking the patient if a coin was closer
or farther away from his body than another coin and in the vertical axis by
asking the patient if a coin was above or below another coin. If the patient's
problem involves multiple axes, then he should have difficulty with a task that
requires him to map objects in all axes. Conversely, if the problem is limited
to a given axis, then we should see a difference in performance between tasks
that involve the affected axis and those that spare the affected axis.
Materials and Methods
A set of three coins (dime, nickel and quarter), placed 5 em apart on a cardboard were
aligned with the patient' s horizontal, vertical or radial axis. All combinations of coin positions
were presented for a total of 36 trials in each axis. The patient correctly identified the
denomination of each coin visually.

Results

The patient correctly named the relative position of the target coin in 31136
trials along the radial axis, in 27/36 trials along the vertical axis and in 19/36
trials along the horizontal axis. There was a statistically significant difference
in the accuracy between the horizontal, radial, and vertical axes (X 2 = 10.135,
d.f. = 2, p = 0.006). The accuracy of responses along the radial and the vertical
axes were comparable (x2 = 1.42, p = 0.234). The accuracy of responses along
the horizontal axis was significantly worse than along the radial and the vertical
axes combined (x2 = 9.05, d.f. = 2, p < 0.025).
Discussion

The results of this task indicate that our patient had the most difficulty with
tasks that require the discrimination of right-left versus near-far or up-down.

276

Michael Gold and Others

The results indicate a relatively selective difficulty with this task on the horizontal
axis compared to either the vertical or the radial axis.
According to the model proposed previously, a defect in this task can arise
from three possible sources: horizontal (right-left) spatial representations are
defective, leading to a loss of knowledge relating the relative horizontal position
of an object to the body. The loss of these representations precludes a comparison
between spatially rotated information and a body-centered reference frame.
Defective spatial translation or mapping along the horizontal axis would lead
to an erroneous comparison to a body-centered reference and cause the
mislabelling of parts.
Alternatively, the module that computes this spatially translated information
may not be able to access spatial representations and a reliable comparison
between spatially translated information and a body-centered reference cannot
be made, causing the mislabeling of parts.
EXPERIMENT 6

Does Defective Translation along the Horizontal Axis Contribute to Right-Left
Confusion?
In the previous experiment the patient demonstrated a problem deriving the
relative position of two coins along the horizontal axis. The task required
translating the position of the reference coin to a body-centered view and then
determining the direction to the target coin. In an attemp to isolate our patient's
spatial mapping defect, our patient was given a simplified version of the coin
task.
Materials and Methods

The patient was now given the name of the reference coin and the direction to the target
coin. The patient was asked to name the coin in a specific direction from the reference coin
("What is the coin to the right of the quarter?"). This task was carried out for a total of 36
trials. In one third of the trials the reference coin lay along the midsagittal plane, in another
third it was to the left and in the last third it was to the right of the midsagittal plane. If
the patient has a problem with translation along the horizontal axis, he should do better
when the reference coin is in the midsagittal position (no spatial translation required) than
when it is to either side.

Results
The patient correctly identified the target coin in 10112 trials when it was in
the midsagittal position, in 4/12 trials when the reference coin was left of center,
and in 7/12 trials when it was right of center. Chi-square analysis demonstrates a
significant difference among the three conditions (X 2 = 6.17, d.f. = 2, p = 0.046),
but cannot identify a difference between the right-of-center and the left-of-center
conditions (X 2 = 1.51, p = 0.219). There is a significant difference between the
midline and left-of-center condition (X 2 = 6.171, p = 0.013), but not between the
midline and right-of-center conditions (x2 = 1.82, p. = 018).

Right-Left confusion in Gerstmann 's Syndrome

277

Discussion
The patient identified the target coins more accurately when no translation
along the horizontal axis was needed than when the reference coin was out of
the midsagittal plane and a translation to a body-centered view was required.
The results of this experiment suggest that our patient has a deficit in establishing
the relative position of an object along the horizontal axis and that our patient's
ability to translate the position of an object initially placed left of center is more
impaired that when an object is initially placed right of center.
Since the neural computation of the spatial relationship between an object
and a person's body is not necessarily based on an Euclidean model of space,
the two previous task may not have required an explicit mental mapping of an
object in a plane. We used a variation of the coin tasks in which the subject
was asked to make. a radial distance (nearer/further) decision in relation to a
non-self reference point. This task forced the patient to perform a mental spatial
rotation in order to make a radial distance judgement.
ExPERIMENT 7

Is Impaired Translation Along the Horizontal Axis Associated with Impaired
Rotation on the Axial Plane?
If mental spatial rotation in the axial plane involves translation along the

horizontal axis, then we would predict that our patient would have difficulty
with tasks that involve rotation in the axial plane. However, if our patient used
an alternative non-metric approach (Poucet, 1993) for spatial mapping or an
alternate plane for mental spatial rotation, then tasks requiring a rotation in the
axial plane would be spared:
Materials and Methods
The patient was presented with the same coins as in experiment 5 along the radial axis.
In one condition a subject stood opposite the patient and the patient was told to make the
near-far decision relative to the control subject's position (eg "Is the nickel closer or further
away from the subject than the dime?"). The patient was presented with a total of 18 trials
in this condition. In the next condition, the control subject was removed and the patient was
told to imagine himself standing opposite himself and to make a near-far decision relative
to that imagined position. The patient was presented with 18 trials in this condition.

Results
The patient correctly identified the coin's position in 18/18 trials with the
other person's body as the reference position and correctly identified 16/18 target
coins when using his imaginary body position as the reference point. The
performances on this task were statistically equivalent (x2 = 0.53, p = 0.47).
Furthermore, when the patient's performance using his imaged position as the
reference point (16118) was compared to his performance using his actual

278

Michael Gold and Others

position (31136), no difference was detected (x2 = 0.08, p = 0.78).
Discussion
The ability of our patient to accurately make a radial distance decision in a
task that was designed to emphasize a spatial rotation in the axial plane was
unexpected. However, had our patient rotated the position of the coins through
the sagittal plane, he could have avoided the axial plane and its component
horizontal axis. Because our patient has problems with spatial mapping along
the horizontal axis, we posited that if given the choice between, rotating an object
in a axial plane or the sagittal plane, our patient would avoid the axial plane.
EXPERIMENT 8

Does the Patient Prefer Rotation in the Sagittal Plane over Rotation in the
Axial Plane?
In this task we presented the patient with models of letters that were placed
in a position such that their correct orientation could be achieved through either
axial or sagittal plane rotations (anagram letter task). We posited that our patient
would depend on rotations involving the sagittal plane and in those cases where
he used the axial plane, he would make mistakes.
Materials and Methods
The patient was presented with models of letters made from pipe cleaners. The models
were presented in a rotated fashion (e.g. V, A, 1.) such that the correct configuration of the
letter could be derived by rotation in either the axial or sagittal planes.

Results
The patient was given a total of 12 trials and derived the correct orientation
in 11112 trials. The patient used a sagittal rotation in 7/11 trials, an axial rotation
in 3/11 trials and a combination of rotations in 2/11. The single error was made
during an attempt to use an axial rotation. Three age-, sex-, handedness- and
education-matched controls were given the same task and derived the correct
orientation of the letter using an axial rotation in 32/44 trials and using a sagittal
rotation in 12/44 trials. Chi-square analysis comparing the number of trials using
sagittal rotations revealed a significant difference between the patient and normal
controls (x2 =5.146, p=0.023).
Discussion
The patient rotated objects in the sagittal plane more frequently than normal
controls. The data in this and the previous experiment suggest that the patient
avoids using the axial plane for spatial rotations by using alternate planes of
rotation. Since the data also suggest that our patient's right-left confusion

279

Right-Left confusion in Gerstmann's Syndrome

Road Mappi~g Task

180) ~ c
0

""

8

D

270°

Fig. 3 - The road mapping task used for spatial rotation. If you are travelling along road B--+ D,
which way should you tum to get to point C?

involves a defect establishing the relative position of an object along the
horizontal axis, we devised. a task that required a rotation in the axial plane
while also constraining the patient from using alternate planes of rotation.
EXPERIMENT 9

Is Mental Rotation in the Axial Plane Defective?
The technique in this experiment uses an idealized four-way 90° intersection
created by printing two perpendicular lines on an 8 112" X 11" piece of paper
(see Figure 3). The end of each line segment was labelled with a letter (A-B,
C-D). The paper was placed on the axial plane such that, from a body-centered
view, one line lay along the mid-sagittal plane and the second line lay along
the coronal plane. For each trial the patient was told to imagine· himself
"travelling" along a road (eg point A to point B). He was then asked which
way he should tum (right or left) to reach point :(B or D) when he reached the
crossroads. By selecting different starting points, the patient was presented with
trials that had 0, 90, 1'80 and 270 degrees df counter-clockwise rotation from
:a body~centered point :of viiew. The intersection was presented in the same

280

Michael Gold and Others

perspective in each trial. This task allowed us to measure the effect of different
degrees of rotation on our patient's ability to make a right-left decision. The
intersecting figure precluded the patient from using a rotation in the sagittal
plane to solve this task. If spatial mapping in the axial plane is defective then
conditions with any degree o(rotation (90, 180, 270) should be defective when
compared to a condition with no rotation (0). If the effort required for mental
rotation is related to the degree of rotation (Shepard and Metzler, 1971), then
our patient's accuracy should decrease as conditions increase in their degrees
of rotation.
Materials and Methods
The patient was presented with all possible combinations of road and end-points for a
total of 36 trials. This task presented him with conditions with 0, 90, 180 and 270 degrees
of counter-clockwise rotation from a body-centered point of view. The patient was told along
which line segment he was to imagine himself travelling. He was then asked to name the
direction (right/left) he should tum to reach a given target point when he reached the
intersection of the two line segments.

Results

The patient chose the correct direction of movement 11112 times with 0° of
rotation (body centered view), 6/12 times with a rotation of 90°, 4/12 times
with a 180° rotation and 2/12 times with a rotation of 270°. Chi-square analysis
comparing the number of correct decisions between the three conditions with
spatial rotation (non-body centered) failed to reveal a significant difference
between conditions with rotation of 90, 180 and 270 degrees (x2 = 3.00, d.f. = 2,
p = 0.223). A second Chi-square analysis comparing the number of correct
decisions in the body-centered condition with non-body-centered conditions
revealed a significant difference (X 2 = 12.75, p < 0.001).
Discussion

When our patient was given a task that emphasized spatial mapping or
rotation in the axial plane, he performed significantly better when no rotation
was required when compared to the three conditions requiring rotation. A
comparison between various degrees of rotation (90, 180, 270) failed to reveal
a significant difference; however, our patient's performance consistently
deteriorated as more clockwise rotation was required to bring the intersection
into a body-centered view. The combination of a clear effect of any rotation
along the axial plane and the implied effect of the degree of rotation suggest
that either spatial rotations systems responsible for mapping positions along the
horizontal axis or its. output are defective.
GENERAL DISCUSSION

We systematically approached the analysis of right-left confusion in a patient
with Gerstmann's syndrome. A comparison of his ability to name or point to

Right-Left confusion in Gerstmann's Syndrome

281

a lateralized body part indicates that the use of verbal labels such as "left" and
"right" or "east" and "west" is not defective. A comparison of his ability to
name body parts on his own body versus that of another person without requiring
any spatial rotation indicates that the patient does not have a severely degraded
body schema. However, when a task requires the patient to name body parts
on a person facing him, his poor performance indicates that he has a problem
with the mental operations required to deal with the relative nature of right/left,
such as spatial mappings or rotations.
The nature of this clinical defect was further investigated by carrying out a
task that tested the patient's ability to establish a body-centered view and derive
the relative position of two objects along different axes of space. The results
of this task demonstrate that our patient has a problem limited to translation
along the horizontal axis. A simpler form of this task in which the patient needed
to establish the relative position of a single object confirmed the problem with
translation along the horizontal axis.
The isolated deficit in horizontal translation suggests that the systems
mediating translation along the horizontal axis may be especially vulnerable to
brain damage. Some of the features of the neural systems that embody horizontal
representations or horizontal spatial transformation'S that might make them more
susceptible to brain damage include late development and the lack of a constant
physical reference such as gravity. Lateral orientation for one's own body (self­
body right/left knowledge) appears to be reasonably well established in children
by age eight, identification of crossed-over body parts by age nine and complete
right-left orientation by age eleven (Clark and Klonoff, 1990; Benton and Sivan,
1993). Since the acquisition of right-left orientation appears after children are
able to use propositional language, right-left confusion in children cannot be
accounted for by language deficits. The hypothesis that the neural systems
involved in determining other-body knowledge, which is dependent on spatial
rotation, may be more susceptible to brain damage is supported by the fact that
patients with Alzheimer's disease have more difficulty identifying the lateralized
parts when they have to make a spatial rotation (Marterer and Danielcyk, 1990).
Additional evidence for this hypothesis comes from Stengel's (1994) report of
a patient with Gerstmann's syndrome whose right-left orientation was profoundly
impaired but had relative sparing of vertical orientation.
We gave our patient a task that seemed to require a rotation along the axial
plane. However, when our patient performed normally, we realized that an
alternate plane of rotation (sagittal plane) could have led to a normal
performance. To test our patient's use of alternate planes of rotation, we gave
our patient an anagram letter task in which a letter was presented in a rotated
form and could be corrected by rotation through either the axial or the sagittal
plane. The patient opted for sagittal plane rotations more frequently than normal
control subjects and avoided rotations that involved the horizontal axis.
To prevent the subject from using alternative planes of rotation, we gave
our patient a road mapping task that required him to make a right/left decision
after having to perform an axial rotation of 0, 90, 180 or 270 degrees of
clockwise or counter-clockwise rotation. The results of this task indicate that
under any condition requiring an axial rotation, our patients' performance was

Michael Gold and Others

282

defective. The application of our model to the other components of Gerstmann' s
syndrome may elucidate a common mechanism. Stengel (1944) suggested that
finger agnosia was due to an inability to identify the position of a finger among
its fellows. Critchely (1966) reported that the index and middle fingers are the
most commoly misnamed fingers in patients with Gerstmann's syndrome.
Dinsboume and Warrington's (1962) detailed study of finger agnosia in a group
of patients with Gerstmann's syndrome demonstrated that all the patients had
problems with tasks that required the determination of the relative position of
one finger to another one. Identifying a finger may involve determining its
relative horizontal position within the hand. A defect in horizontal translation
could affect all the fingers, but would be predicted to affect the middle digits
more.
We conclude that a defect in horizontal translation accounted for the right­
left disorientation in our patient. Although we did not test our hypothesis on
the other features of the syndrome, the spatial features of mathematical operations
and the fact that these operations are defective in Gerstmann's syndrome (Levin,
Goldstein and Spiers, 1993; Critchley, 1953) invites the application of our model
to a patient with Gerstmann's syndrome.
REFERENCES
ARMY INDIVIDUAL TEST BATTERY. Manual of Directions and Scoring. Washington, D.C.: War
Department, Adjutant General's Office, 1944.
BENSON, D.F., CUMMINGS, JL, and TsAI, S.Y. Angular Gyrus syndrome simulating Alzheimer's Disease.

Archives of Neurology, 20: 616-620, 1982.

BENTON, A.L. Right-left Discrimination and Finger Localization: Development and Pathology. New
York: Hoeber-Harper, 1959.
BENTON, A.L. The Fiction of the Gerstmann Syndrome. Journal of Neurology, Neurosurgery and

Psychiatry, 24: 176-181, 1961.

BENTON, A.L. Reflection!! on the Gertsmann's Syndrome. Brain and Language, 4: 45-62, 1977.
BEN'I"ON, A.L., and HAMSHER, K. DE S. Multilingual Aphasia Examination. Iowa City: University of
Iowa, 1976 (revised 1878).
BENTON, A.L., and SrvAN, A.B. Disturbances of the body schema. In K.M. Heilman and E. Valenstein
(Eds.), Clinical Neuropsychdlogy, 3rd ed. New York: Oxford University Press, 1933, Ch. 6, pp.

123-140.

BERG, E.A. A simple objective test for measuring flexibility in thinking. Journal of General Psychology,
39: 15-22, 1948.
CHERTKOW, H., and BuB, D. Semantic memory loss in dementia of Alzheimer's type. What do the
various measures measure? Brain, 113: 397.:417, 1990.
CLARK, C.M., and KLONOFF, H. Right and left orientation in children aged 5 to 13 years. Journal of

Clincal and Experimental Neuropsychology, 12: 459-466, 1990.

CRITCHLEY, M. The Parietal Lobes. London: Arnold, 1953.
CRITCHLEY, M. The enigma of Gertsmann's Syndrome. Brain, 89: 186-198, 1966.
DANIEL, W.W. Applield Nonparametric Statistics: Second Edition. Boston: PWS-Kent Publishing
Company, 1990.
DELIS, D.C., KRAMER, J.H., KAPLAN, E., and 0BER, B.A. California Verbal Learning Test: Research
Edition . .New YorK: The.Psychological Corporation, 19&3.
FisHER, P., MARTHRER, A.,.and DANIELCZYK, W. Right-left,disorientationin dementia of.the Alzheimer's
type. Neurology, 40: 1619-1620, 1990.
BERSTMJIINN, J. Fingeragnosie eine umschriebene Smrung uer ·Orienterung am eigener Koerper. Wiener

klinische Wochenschrift, 37: '1010-~10.12, 'f924 .

.GERSTMANN, J. Zur Symptomatologie der Nirnlasionen im -Uebergangs.gebeit der ·unteren .•P..ari:etal- .und
mittlerem Occipitalwindung. Nervenartz, 3: 691;695, 1930.
nRAFMAN, J., PASSA'FIUME, D., FAGLIONI, :p., ani1 1BOLLER, .F. Calculation disturbances in adults with
focal hemispheric damage. Cortex, 18: 37-50, 1982.

Right-Left confusion in Gerstmann's Syndrome

283

HEAD, H. Aphasia and Kindred Disorders and Speech. Cambridge, England: The University Press,
1926.
HECAEN, H. Introduction ii la Neuropsychologie. Paris: Larousse, 1972_
HEILMAN, K.M., WATSON, R.T., and V ALENSTEIN, E. Neglect and related disorders. In K.M. Heilman
and E. Valenstein (Eds.), Clinical Neuropsychology, 3rd edition. New York: Oxford University
Press, 1993, Ch. 10, pp. 279-336.
HEIMBURGER, R.F., DEMYER, W.C., and REITAN, R.M. Implications of Gertsmann's syndrome. Journal
of Neurology, Neurosurgery and Psychiatry, 27: 52-57, 1964.
KAPLAN, E., GOODGLASS, H., and WEINTRAUB, S. The Boston Naming Test. Lea and Febiger,
Philadelphia, 1983.
KERTESZ A. Aphasia and Associated Disorders. New York: Grune and Stratton, 1979.
KINSBOURNE, M., and WARRINGTON, E.K. A study of finger agnosia. Brain, 85: 47-66, 1962.
LEVIN, H.S., GoLDSTEIN, F.C., and SPIERS, P.A. Acalculia. In K.M. Heilman and E. Valenstein (Eds.),
Clinical Neuropsychology, 3rd ed. New York: Oxford University Press, 1993, Ch. 5, pp. 95-122.
MENNEMEIER, M., CHATTERJEE, A., and HEILMAN, K.M. A comparison of the influences of body and
environment centered reference frames on neglect. Brain, 117: 1013-1021, 1994.
MoNEY, J. A Standardized Road Map of Direction Sense. San Rafael, CA: Academic Therapy
Publications, 1976.
PoucET, B. Spatial cognitive maps in animals: New hypotheses on their structure and neural mechanisms.
Psychological Review, 100: 163-182, 1993.
SHEPARD, R.N., and METZLER, J. Mental rotation of three-dimensional objects. Science, 171: 701-703,
1971.
STENGEL, E. Loss of spatial orientation, constructional apraxia and Gerstrnann's Syndrome. Journal of
Mental Sciences, 90: 753-760, 1944.
TRAHAN, D.E., and LARRABEE, G.J. Continuous Visual Memory Test. Odessa, FL: Psychological
Assessment Resources Inc., 1983.
WARRINGTON, E.K, and JAMES, M. Visual apperceptive agnosia: A clinico-anatomical study of three
cases. Cortex, 24: 13-32, 1988.
WARRINGTON, E.K., and TAYLOR, A.M. The contribution of the right parietal lobe to object recognition.
Cortex, 9: 152-165, 1973.
WECHSLER, D. Wechsler Adult Intelligence Scale - Revised. Manual. New York: The Psychological
Corporation, 1981.
WEISKRANTZ, L., and SAUNDERS, R.C. Impairments of visual object transforms in monkeys. Brain,
107: 1033-1072, 1984.
Kenneth M. Heilman, M.D., Box 100236, University of Florida, Gainesville, FL 32610-0236, U.S.A.