Brain (1998), 121, 1459–1467

Let not thy left hand know what thy right hand
knoweth
The case of a patient with an infarct involving the callosal
pathways
P. Marangolo,1 E. De Renzi,2 E. Di Pace,1 P. Ciurli1 and A. Castriota-Scandenbeg 1
1Centro Ricerche di Neuropsicologia, IRCCS S. Lucia,

Rome and 2Clinica Neurologica, University of Modena,
Italy

Correspondence to: Ennio De Renzi, Clinica Neurologica,
Via Del Pozzo 71, 41100 Modena, Italy.
E-mail: derenzi@unimo.it

Summary
Following a cerebral vascular accident, a patient showed
a classical disconnection syndrome: left-hand tactile
anomia, apraxia and dysgraphia and right-hand
constructional apraxia. What made the case unusual was
the presence of hand asymmetry in the performance of
some matching-to-sample tasks carried out in foveal
vision. The left hand committed significantly more errors
than the right hand when it was not possible to identify
on a perceptual basis the stimulus that was to be matched,
because it was removed (memory condition) or was
indicated verbally (verbo-visual matching), or had the
same name but not the same physical appearance as the
match (capital and lower-case letter matching). No hand
difference emerged when the stimulus remained in full
view throughout the matching task (perceptual condition).
The hand effect, however, was limited to colours and
letters. Objects, geometrical shapes and unfamiliar faces
were matched with equal proficiency by both hands under
every condition of presentation. Left-hand errors also
significantly outnumbered right-hand errors in sorting

colours according to hue and colouring drawings. MRI
showed an infarct in the left cingulate white matter that
ran parallel to the trunk of the corpus callosum, and an
infarct of the splenium. However, the latter did not
prevent the transmission of colour and letter information
between the two hemispheres, as shown by the
performance on perceptual equivalence tasks and by
the correct right-hand responses to stimuli projected
to the left visual field. We propose that this pattern of
deficit is contingent upon the specific role that the left
hemisphere plays in categorizing a given colour patch as
belonging to a definite colour region (red, blue, etc.) and
in grapheme recognition. Without the assistance of the
left side the right hemisphere lacks the benefit provided
by meaning identification. In our patient the left brain
did receive information from the right brain and was
able to process it, but was prevented by the paracallosal
lesion from transmitting what it knew to the right motor
area. No hand effect emerged for objects and geometrical
shapes, because their semantic memory is bilaterally
represented.

Keywords: hemisphere disconnection; letter recognition; colour recognition; limb apraxia

Introduction
Complete section of the commissures results in two isolated
hemispheres that are no longer able to carry out tasks
dependent on information processed in the other half of the
brain. Accordingly, neither hand can match-to-sample an
object presented visually or tactually to the ipsilateral
hemisphere. In addition, the isolated right hemisphere fails
tasks involving language, such as reading words and naming
objects projected to the left visual field, naming objects held
in the left hand and performing gestures with the left hand
on verbal command. A question still under debate is whether
or not the right brain has the lexical competence to
© Oxford University Press 1998

comprehend nouns. While some authors (Zaidel, 1990) claim
that this ability is represented in the minor hemisphere, others
(Gazzaniga, 1983) argue that there is wide variability across
patients. All agree, however, that the right brain can interpret
the meaning of the visual representation of an object, as
shown by the left hand’s ability to match a palpated object
or an object that has been presented to the left visual field
with a semantically related object (Cronin-Golomb, 1986;
Trevarthen, 1990).
In patients who have undergone surgical section of the
callosum, the disconnection syndrome is contingent upon the

1460

P. Marangolo et al.

Fig. 1 MRI horizontal sections showing in A the paracallosal infarct and in B the splenial infarct.

complete interruption of splenial fibres. If they are partially
spared, no deficit appears (Bogen, 1987). Indeed, this was
the interpretation of Goldstein and Joynt (1969) of the
absence of interhemispheric transfer deficits in the patients
reported by Akelaitis (1944). Rather more positive were
cases of spontaneous pathology, such as the patient reported
by Geschwind and Kaplan (1962), who showed definite signs
of hemispheric disconnection following an infarct damaging
the anterior half of the trunk. The most widely accepted
account (Gordon, 1990) of this contradiction is that most, if
not all, of the symptoms are due to diaschisis.
We report a patient who showed, under certain experimental
conditions, a striking inferiority of his left hand in performing
matching-to-sample tasks carried out in full vision, as a
sequel to a paracallosal lesion of the left hemisphere. This
patient made many more errors with his left hand than with
his right hand, even when visual stimuli were processed by
the right hemisphere and could reach the right motor centre.
We think that this pattern of impairment throws new light
on the hemispheric specialization of function.

Case report
Antonio (a fictitious name) was an 82-year-old man whose
medical history was unremarkable except for arterial
hypertension over the previous few years. On December 7,
1996, the patient presented with right hemiparesis and

aphasia, from which he made a partial recovery. He came to
our attention in January 1997. Neurological examination
showed a mild paresis, marked grasping and sensory deficits
of the right hand. The results of visual field examination
(Goldman perimetry) were normal. His speech was halting
and laborious, with some phonological distortions,
reminiscent of those found in Broca’s aphasia.

Neuroradiological findings
An MRI, carried out on February 18, 1997, showed a
lesion located in the left cingulum (areas 23 and 24),
involving the roof of the lateral ventricle, from the anterior
margin of the frontal horn to the level of the paracentral
lobule, parallel to the corpus callosum (Fig. 1A). It
probably interrupted the callosal pathways of the trunk
within the left hemisphere, with the possible exception of
a few posterior fibres. In addition, there was evidence of
a small lesion of the genu and a wide lesion of the
splenium (Fig. 1B). It is uncertain whether the latter was
an extension of the callosal infarct or a separate infarct.
Punctiform lesions were seen in the left temporal lobe and
the white matter of both hemispheres.

Neuropsychological examination
Formal language examination (Ciurli et al., 1996) was carried
out in January 1997 and repeated in March 1997, when a

Left-hand-impaired letter and colour recognition
Table 1 Performance on language tests
Number of correct responses/
number of test items
Token test
Right hand
Left hand
Visual naming
Objects
Visual naming
Colours
Tactile naming
Right hand
Left hand
Tactile equivalence
Right hand
Left hand
Reading words
Reading non-words
Writing object names
Right hand
Left hand
Writing to dictation
Right hand
Left hand
Imitating movements
Right hand
Left hand

29/36
14/36
16/20
4/10
6/20
11/20
12/20
19/20
19/20
9/20
13/20
0/20
11/20
0/20
54/72
36/72

moderate improvement was observed in the results of all
tests. Here we will focus on the scores of the second
examination (Table 1), which was carried out at the same
time as the experimental investigations to be reported below.
It should be pointed out, however, that the hand effect was
already present at the time of the first examination, when it
was fortuitously discovered whilst the patient was performing
the token test. Since the performance of the right hand was
sometimes hampered by grasping, the examiner asked the
patient to point to the tokens with his left hand and was
amazed to find that the number of errors increased remarkably:
the number of correct responses was 26 out of 36 with the
right hand and 7 out of 36 with the left hand. Table 1 shows
that the hand asymmetry on the token test was replicated in
March (χ2 5 12.99, P , 0.001). It was also apparent in
writing. When Antonio used his right hand to write the names
of 20 visually presented objects, he gave 13 correct responses,
five responses with one spelling error and two responses
with more than one spelling error. On writing to the dictation
of the same words, he gave 11 correct responses, eight
responses with one spelling error and one response with
more than one spelling error. By contrast, on both tests his
left-hand writing was characterized by the perseveration of
the same ill-formed letters. No improvement was observed
when the patient attempted to write in block letters: with his
right hand he made a few omission errors and with his left
hand he perseverated in selecting the same letters, unrelated
to the target. Object visual naming was within the normal
range, while colour naming was impaired. Tactile naming
(tested with the same 20 objects used in visual naming)

1461

revealed an apparent inversion of the hand asymmetry, the
number of correct responses being 6 out of 20 with the right
hand and 11 out of 20 with the left hand. However, it seems
likely that the patient’s right- and left-hand errors did not
share the same origin. Those produced using the right hand
stemmed from the somaesthetic deficits found on neurological
examination. In keeping with this hypothesis, the patient
performed poorly with his right hand in a test in which he
had to feel two objects in succession and to decide whether
they were same or different. When the task was carried out
with the left hand, the performance was almost flawless.
Thus the impairment he showed with his left hand in tactile
naming cannot be attributed to stereognosic deficits. Colour
discrimination was assessed by means of the Ishihara and
Farnsworth test, and was found to be normal. A striking
difference between the hands was also apparent on a gesture
imitation test (Table 1). According to the norms provided by
De Renzi et al. (1980), the patient’s score was at the
borderline level with the right hand and definitely pathological
with the left hand (χ2 5 9.60, P , 0.001). On the contrary,
constructional apraxia (copying drawings) was more impaired
with the right hand than with the left hand.
The unexpected finding that the patient’s performance on
the token test was worse with the left hand than with the
right hand provided the impetus for the following experiments
that were carried out in the period from March to June 1997,
during which the patient’s deficit remained unchanged. The
patient’s co-operation throughout the tasks was good and he
never showed signs of fatigue or lack of concentration. He
gave informed consent to participate.

Experiment 1. Perceptual and mnestic
matching to sample
This experiment was carried out to verify whether the
ability to match-to-sample various types of visual stimuli
was contingent upon the hand used. Two paradigms were
used, both of which required the patient to match the
presented stimulus with the identical one from among
eight alternatives. One made demands on perceptual skills
only (the stimulus and the alternatives were present
throughout the task), while the other also had a mnestic
component: the stimulus was presented for 3 s, during
which the alternatives were covered with a piece of
cardboard which was removed after a further 3 s unfilled
delay. The following sets of stimuli were used: (i) drawings
of objects; (ii) geometrical figures; (iii) unfamiliar faces;
(iv) coloured cards (red, green, yellow, pink, brown, blue,
black and white); and (v) letters. For each type of stimulus,
blocks of 48 trials were performed first with the right and
then with the left hand.
The results showing correct responses made with either
hand in the two tests are shown in Table 2. Unsurprisingly,
the memory task proved to be more difficult than the
perceptual task for all categories, except objects, in which

1462

P. Marangolo et al.

Table 2 Correct responses in the perceptual and memory
matching-to-sample tasks
Number of correct responses/number of test items

Objects
Geometrical
shapes
Faces
Colours
Letters

Perceptual task

Memory task

Right hand Left hand

Right hand Left hand

48/48
45/48

48/48
45/48

48/48
39/48

45/48
34/48

40/48
46/48
48/48

40/48
45/48
46/48

30/48
40/48
45/48

23/48
17/48
30/48

Antonio’s performance was at ceiling in both conditions.
More interesting was the finding that, on the perceptual
task the patient had a similar performance with either
hand, whereas on the memory task his performance was
worse with the left hand than the right hand. The scores
(successes/failures) of the memory task were converted to
natural logarithms (1n) and were analysed by means of a
factorial analysis of deviance with five tests and two hands
(Ross, 1990). The hand factor was significant (χ2 5 30.56,
P , 0.001), because of the patient’s poorer performance
with the left hand. The hand 3 test interaction showed a
trend towards significance [χ2(4) 5 8.50, P , 0.10]. When
the hand factor was analysed within tests, it was significant
for colours [χ2(1) 5 20.26, P , 0.001] and letters [χ2(1) 5
5.37, P , 0.02] but not for objects, geometrical shapes or
unfamiliar faces. In matching colours and letters to sample,
Antonio’s performance with the left hand was 65% and
37% inferior, respectively, to that with the right hand.

Comment
It seems that the hand effect, found for colours and letters,
was contingent upon the absence of the stimulus during
the matching task. Would the same effect occur if the
information about the target were provided verbally?

Experiment 2. Verbo-visual matching
To answer the above question, a verbo-visual matching
test was given. Antonio viewed an array of eight stimuli
and selected from these the stimulus named by the
examiner. Only objects, colours and letters were tested,
given the impossibility of assigning verbal labels to
unfamiliar faces. Blocks of 40 trials per hand were
administered for objects and colours and blocks of 48
trials for letters.
Again the patient’s performance was 100% correct with
either hand when the stimuli were objects. When they
were colours or letters, his correct responses with the right
hand were 31 out of 40 (78%) and 40 out of 48 (83%),
respectively, while those with the left hand were 22 out

of 40 (55%) and 20 out of 48 (42%) (χ2 5 4.53, P , 0.05
for colours and 17.78, P , 0.001 for letters).
In order to rule out the possibility that the absence of
a hand effect with objects was contingent upon the patient’s
ceiling performance, a more difficult verbo-visual test was
devised. An abstract noun was given orally and the patient
had to select from an array of eight to 10 drawings of
objects the one that was related to it (e.g. he had to point
to the drawing of a church in response to the word
spirituality). No significant hand effect emerged, although
his performance was far from perfect: he scored 35 out
of 48 with the right hand and 30 out of 48 with the left hand.
To compare Antonio’s verbo-visual matching of objects
and colours, we devised a further test, in which the same
stimuli were used. The patient viewed arrays of different
coloured drawings of objects and attempted, in separate
blocks, to match-to-sample either the name of the object
or the name of the colour. The selective impairment of
the left hand for colours was confirmed. On object pointing,
his performance was almost perfect (right hand 48 out of
48, left hand 46 out of 48), while on colour pointing his
performance with the left hand showed a 68% decrement
compared with that of the right hand (right hand 37 out
of 48, left hand 12 out of 48; χ2 5 26.05, P , 0.001).

Comment
The foregoing experiments provided evidence that neither
hemisphere was able to guide the left hand correctly in
pointing to colours and letters when the matching task
demanded a mental representation of the target and not
merely its perception. This finding is surprising, if one
considers that Antonio processed the stimuli with both
visual areas and that the pathways linking the right visual
cortex with the right motor cortex were apparently viable,
as shown by his good performance with objects.
Given the damage of the splenium, we decided to
investigate whether the interhemispheric transfer of visual
information was effective and whether the hand asymmetry
would persist if the stimulus were to be projected to one
visual field, such that the visual field and the hand effect
could be examined separately.

Experiment 3. Same-different judgements on
the perceptual equivalence of laterally
projected stimuli
Only stimuli that had yielded a hand effect, i.e. colours
and letters, were used in a memory paradigm. The
procedure was as follows. Stimuli were presented on the
screen of a computer 4° to the right or left of the fixation
point for 100 ms. The patient viewed the screen from a
distance of 57 cm. and was requested to fix his gaze on
a central cross. A second examiner checked that he
maintained fixation throughout the test. After a ‘ready’

Left-hand-impaired letter and colour recognition

1463

Table 3 Correct responses in the memory matching-to-sample task with laterally presented
stimuli
Number of correct responses/number of test items

Objects
Geometrical shapes
Colours
Letters

Right hand

Left hand

Right visual field Left visual field

Right visual field Left visual field

29/36
31/36
35/36
36/36

27/36
26/36
16/36
23/36

23/36
30/36
36/36
30/36

signal, a stimulus was flashed to either visual field and,
after a 3 s delay, a second stimulus was projected to the
same or the opposite field. The patient said whether the
two stimuli were ‘same’ or ‘different’. Forty-eight pairs
were presented for each category: half were the same and
half were different. In half of the trials the first stimulus
was projected to the left visual field and in the other half
to the right visual field. Antonio gave correct responses
to 96% of the pairs of colours (46 out of 48) and 88%
of the pairs of letters (42 out of 48).

22/36
28/36
20/36
22/36

and the letter test (χ2 5 14.77, P , 0.01), even when the
error rates were controlled family-wise.

Comment
The results of the previous experiments were replicated.
The performance was not contingent upon the visual field
to which stimuli were sent, but upon the hand used to
give the response. The hand effect was, however, limited
to colours and letters and did not affect either objects or
geometrical shapes.

Comment
When the left hemisphere gave the response, stimuli
projected to one field could be successfully compared with
those projected to the other field.

Experiment 4. The hand effect in delayed
responses to laterally projected stimuli
A closer assessment of the patient’s ability to transmit
interhemispheric visual information in the memory paradigm
was carried out by studying all the possible combinations
of hand with visual field. A single stimulus was flashed
in the right or left visual field for 100 ms. Following a
3 s delay, six alternatives were presented in full view and
the patient was requested to point to the match, using the
right hand in half the trials and the left hand in the other
half. Thirty-six trials were given for each field by hand
combination in separate blocks. Objects, geometrical shapes,
colours and letters were tested. The patient’s scores are
presented in Table 3.
Data (success/failures) were converted to natural
logarithms (1n) and were analysed by means of a factorial
analysis of deviance with four tests, two hands and two
visual fields (Ross, 1990). The significant effects were test
[χ2(3) 5 11.64, P , 0.01], hand [χ2(1) 5 12.93, P , 0.001]
and the hand 3 test interaction (χ2 5 16.38, P , 0.001).
To study this significant interaction, the hand factor was
compared within each test. It was not significant in the
object and geometrical shape tests, even when error rates
were controlled experiment-wise (Miller, 1981), while it
was significant in the colour test (χ2 5 16.99, P , 0.005)

Experiment 5. Colour-form association and
colour sorting
In the following experiment, we extended the study of the
hand effect to colour tasks not involving matching to sample.
The non-verbal tests used, i.e. the colour sorting test and the
colour-form association test, have been widely employed and
reported in the literature to investigate colour disorders
consequent on brain damage. There is substantial evidence
that both tests are sensitive measures of colour agnosia
(Lewandowsky, 1908; Sittig, 1921) and are impaired
following left brain damage (De Renzi and Spinnler, 1967;
De Renzi et al., 1972a, b; Basso et al., 1976).
In the colour-sorting task, 35 small colour patches were
scattered on the table in front of the patient, whose task was
to sort them according to hue. There were five patches of
different shades for each of the following seven colours:
blue, green, red, yellow, brown, grey and violet. A score of
1 was given for each card added to the first card selected to
designate a colour category. For example, if the patient
grouped together all five shades of a colour, he received a
score of 4, if he grouped four of them, a score of 3, etc. The
maximum score was 28. The task was carried out first
with the right hand, then with the left hand. The patient’s
performance on this test was compared with that on a
matching test in which 24 pairs of different coloured cards
were scattered on the table and the patient had to pair the
cards of the same colour. The maximum score was 24.
On the matching task, Antonio performed flawlessly (24
out of 24) with the right hand and made three errors with
the left hand (black was paired with grey, pink with grey,

1464

P. Marangolo et al.

and red with orange). He ran into greater difficulties on the
sorting task and the hand difference was much more marked:
right hand 15 out of 24, left hand 2 out of 24 (χ2 5 15.39,
P , 0.001). When requested to sort with the right hand, he
was slow and uncertain and took 8 min 34 s to group together
all the green and yellow cards, four blue, four violet and two
grey cards. Although defective, this right-hand performance
was definitely superior to the left-hand performance: after
4 min he managed to sort out only two green and two blue
cards and had grouped the rest haphazardly. The correct rate
was significantly higher with the right hand (χ2 5 15.39,
P , 0.001)
In the colour form association task, Antonio was requested
to colour drawings of objects, animals and fruits having a
characteristic colour. Ten coloured pencils and 40 outline
drawings were presented and the patient was requested to
colour them in first with his right hand and then, after a
delay, with his left hand.
The right hand performance was poor (11 out of 40), but
significantly less impaired than the left hand performance (3
out of 40) (χ2 5 5.54, P , 0.02).

Comment
Antonio’s right hand performance was poor on both the
colour-sorting and the colour–form association task,
indicating that his ability to consider a hue as representative
of a colour category and to imagine the characteristic colour
of an object was impaired, probably as a consequence of left
brain damage. However, not even these impoverished skills
were available to the hand controlled by the right hemisphere,
with which the patient performed at chance level. When it
was possible for him to sort colours by perceptual identity
basis, his performance was much more successful.

Experiment 6. Letter name identity
The final experiment probed the patient’s ability to match
lower-case letters with upper-case letters. Its aim was to
assess whether the presence of a stimulus that could not be
matched according to perceptual identity would result in left
hand errors.
In half of the trials a lower-case (target) letter was presented
together with eight capital letters, and in the other half an
upper-case (target) letter was presented together with eight
lower-case letters, in a total of 48 trials. The patient’s task
was to point to the letter that matched the target. Letters with
similar shapes in both forms (such as C and c) were not used.
Antonio scored 44 out of 48 with the right hand and 31
out of 48 with the left hand (χ2 5 10.30, P , 0.001).

Discussion
Antonio showed the classical signs of hemispheric
disconnection: tactile agnosia, ideomotor dyspraxia and
agraphia of the left hand and constructional apraxia of the

right hand. This pattern of impairment is typical of patients
with total section of the commissures, and it was probably
contingent upon the left hemisphere infarct that severed
callosal fibres. What makes this case surprising and prima
facie radically different from other callosotomy patients is
that the patient’s left-hand impairment was apparent even
under conditions in which the information was received
by both sides of the brain and hence did not require
interhemispheric transmission. The left-hand inferiority
emerged when the stimulus was no longer present (memory
condition), when it was identified by its name (verbo-visual
matching) and when the test involved stimuli (the same capital
and lower-case letter) that were not physically identical.
Moreover, it was confined to colours and letters and was
absent for objects, geometrical shapes and unfamiliar faces.
It must be stressed, however, that Antonio’s left-hand
performance was always well above chance level. Thus we
are faced with a partial rather than a total inability to perform
the tasks.
The issue to be addressed is why the information that
reached the right visual cortex could not assist the ipsilateral
motor centre in programming left hand’s movements towards
the target. The right brain cannot be blamed for this failure,
because the patient’s performance under the perceptual
condition and in tasks involving objects, geometrical shapes
and unfamiliar faces provided evidence that the right motor
centre did receive sufficient visual and verbal information to
steer left-hand movements. Therefore, we must turn to
the left hemisphere to solve the enigma of the left-hand
impairment.
Despite its lesions, the left hemisphere still had the
necessary competence to carry out the tasks in which the left
hand failed, as shown by the patient’s nearly perfect righthand performance in most of the tasks. Therefore, the
contribution of the left brain to left-hand errors must be
perceived as reflecting a deficit of assistance rather than a
deficit of function, and must be sought in the after-effects of
either or both of the lesions, as revealed by MRI. The
relevance of splenial damage can be readily dismissed, given
that the left-hand impairment was present when the stimuli
were displayed in free vision and could, therefore, be seen
by both hemispheres. Moreover, Experiments 3 and 4 showed
that the output of the right visual cortex could be transferred
to the left visual cortex with no apparent loss of information.
Hence, it appears that the crucial lesion was that located
in the left white matter, close to the trunk of the corpus
callosum, which interrupted the pathways linking the left
brain with the right motor cortex. In essence, the question is
why letters and colours had to be initially processed by the
left hemisphere in order to be recognized. With regard to
letters, we suggest that the left hemisphere plays a crucial
role in the identification of their graphemic value. As long
as letters are processed by the right hemisphere, they are
shapes without a linguistic status and are not recognized as
the same entity when presented in different forms (D and d).
To become graphemes and subsequently undergo grapheme-

Left-hand-impaired letter and colour recognition
to-phoneme transcoding, they must be transferred to the left
hemisphere. This assumption is in keeping with the role
assigned to the left hemisphere by previous studies carried
out in patients with focal brain lesion (Dejerine, 1892;
Geschwind, 1962; Hécaen and Kremin, 1977; Sevush and
Heilman, 1984) and in patients undergoing complete
callosotomy, most of whom have been found to lack
grapheme-phoneme correspondence rules, when words are
projected to the isolated right hemisphere (Baynes, 1990;
Baynes et al., 1992). If the right brain is unable to specify
the graphemic value of a letter, it will fail to recognize not
only that two different forms of the same letter have identical
meaning (name identity task), but also which letter
corresponds to a given letter name (verbo-visual matching
task). In a similar vein, in the delayed matching-to-sample
task letters will be encoded as meaningless shapes, giving
rise to traces that, without the reinforcement of their graphic
identity, will be weak and difficult to retrieve.
Less is known about the contribution of the left brain to
colour identification. Prima facie, the presence of a hand
effect in the delayed but not in the perceptual condition
would suggest that the right brain is equipped with a reduced
short-term memory for colours, which hampers the retention
of this kind of information during a delay. Davidoff and
Ostergaard (1984) reported a patient with a large infarct of
the left temporo-occipital region, who committed more than
50% of errors in matching-to-sample colours (but not
meaningless shapes) that had been presented 4 s before.
However, the idea that discrete short-term mechanisms are
dedicated to different categories of stimuli is not entirely
convincing and does not account for the whole spectrum
of conditions in which Antonio manifested a hand effect
for colours.
One may wonder whether the left hand inferiority in the
verbo-visual matching test had a verbal origin, namely, that
it was dependent upon the fact that colour names, unlike
object names, were poorly comprehended by the right
hemisphere. The only data we have on this subject come
from the administration to two callosotomy patients of the
token test, which involves both shape names and colour
names (Zaidel, 1976, 1977). When the tokens were shown
to one hemisphere, both patients achieved lower scores with
the right than with the left brain, but there was no evidence
that colour names were more difficult to understand than
shape names, as was the case for Antonio.
Therefore, we propose that the left-hand errors, committed
in both the verbo-visual and the memory matching test, stem
from the same difficulty, i.e. the limited competence of the
right hemisphere in colour categorization. In nature, colours
do not exist in isolation, but are surface properties of objects.
Identifying a colour patch requires an internal colour space
(Davidoff, 1991), which stores the mental representations of
regions of colours. These representations provide the standard
for categorizing the colour of a sample as red, blue, etc. and
are activated when a colour must be retrieved from memory.
We suggest that this ability is a predominant function of the

1465

left hemisphere and that, without it, the encoding of colours
by the right hemisphere is weak and at risk of dwindling
altogether. Likewise, the ability to categorize a colour is a
prerequisite for retrieving the colour that must be matched
with a name, and accounts for the left-hand errors made
in the verbo-visual matching test. The advantage of this
interpretation is that it also holds for the left-hand impairment
shown by the patient in colouring drawings and sorting
colours according to hue, tasks that involve neither shortterm memory nor verbal comprehension, but require the
ability to categorize and evoke colours. Both were found to
be impaired following left brain damage (De Renzi and
Spinnler, 1967; De Renzi et al., 1972a, b; Basso et al., 1976),
and were in fact performed poorly by Antonio when he used
his right hand, suggesting that the failure was in part due
to colour agnosia, consequent on left-hemisphere damage.
However, the score fell to almost zero when he used his left
hand, showing that of its own the right brain was all but
virtually incapable of carrying out the task.
In conclusion, we propose that colours, as well as letters,
need to be processed by the left hemisphere, whenever their
recognition requires the identification of their categorical
value. The absence of a left-hand effect in the recognition
of objects and geometrical shapes can be accounted for by
assuming that their semantic store is represented bilaterally,
as shown by the ability of callosotomy patients to use their
left hand to point to objects that are semantically related to
those presented to the left visual field. As to the lack of hand
asymmetry with unfamiliar faces, by definition these are not
represented in the long-term store and thus undergo merely
perceptual encoding, for which the right hemisphere is
equally, if not better equipped than the left hemisphere.
The left-hand effect also emerged in writing and imitating
gestures. The right-hand writing errors were linguistic, while
those of the left hand were characterized by the dogged
perseveration of the same word and poorly formed letters. It
is apparent that, without the assistance of the left hemisphere,
the right motor centre was incapable of controlling the lefthand writing performance. Interestingly, left-limb apraxia
was present when the patient attempted to imitate movements
seen by both hemispheres. This is at variance with what is
observed in complete commissurotomy patients, in whom
left-limb apraxia occurs only when the hemisphere receiving
the message cannot communicate with that guiding the
gesture, i.e. on verbal command or on the imitation of hand
or finger positions projected to the left hemisphere (Zaidel
and Sperry, 1977; Volpe et al., 1982). These findings led
Nass and Gazzaniga (1987) to challenge Liepmann’s (1905)
classical notion that the motor engrams of the left premotor
cortex exert control over those stored in the right premotor
cortex. Antonio’s left limb apraxia belies this claim. Since
his right brain could see the gestures made by the examiner,
its failure must be contingent upon the faulty transmission
of the motor programmes elaborated by the left hemisphere.
This finding underlines the difference between the
disconnection syndromes observed in patients with

1466

P. Marangolo et al.

spontaneous pathology and in epileptics submitted to
callosotomy. For instance, essentially negative findings were
reported in the latter when the callosal section was incomplete
(Gordon et al., 1971). They were able to transfer visual and
tactile information from one hemisphere to the other and to
carry out matching-to-sample tasks, in the absence of the
stimulus (Bogen, 1987; Gordon, 1990). On the contrary,
left hand tactile anomia, apraxia for verbal commands and
agraphia were manifest when the incomplete callosal
interruption resulted from vascular or neoplastic lesions
(Sweet, 1941; Geschwind and Kaplan, 1962; Schott et al.,
1969; Poncet et al., 1978; Beukelman et al., 1980; Watson
and Heilman, 1983; Graff-Radford et al., 1987). Our patient
did not have a complete section of the commissures: rostral
and genu fibres were not involved in the lesion, a few fibres
of the posterior trunk may have been spared and the splenial
fibres were functional, despite their damage. Yet while
epileptic patients could exploit the few splenial fibres left
intact by commissurotomy to carry out lateralized tasks,
Antonio was unable to use the residual splenial connections
to transfer the output of the left-hemisphere cognitive
processes to the contralateral side. In order to activate the
right motor centre, this information had to travel along anterior
callosal pathways, which were interrupted by the lesion.

Acknowledgements
We wish to thank Dr Piero Faglioni for his valuable aid in
the statistical analysis. This research was supported by a
grant from the University of Modena.

References
Akelaitis AJ. A study of gnosis, praxis, and language following
section of the corpus callosum and anterior commissure. J Neurosurg
1944; 1: 94–102.
Basso A, Faglioni P, Spinnler H. Non-verbal colour impairment of
aphasics. Neuropsychologia 1976; 14: 183–93.
Baynes K. Language and reading in the right hemisphere: highways
or byways of the brain? J Cognit Neurosci 1990; 2: 159–79.
Baynes K, Tramo MJ, Gazzaniga MS. Reading with a limited
lexicon in the right hemisphere of a callosotomy patient.
Neuropsychologia 1992; 30: 187–200.
Beukelman DR, Flowers CR, Swanson PD. Cerebral disconnection
associated with anterior communicating artery aneurysm:
implications for evaluation of symptoms. Arch Phys Med Rehabil
1980; 61: 18–23.
Bogen JE. Physiological consequences of complete or partial
commissural section. In: Apuzzo MLJ, editor. Surgery of the third
ventricle. Baltimore: Williams and Wilkins; 1987. p. 175–94.
Ciurli P, Marangolo P, Basso A. Esame del linguaggio II. Firenze:
Organizzazioni Speciali; 1996.
Cronin-Golomb A. Comprehension of abstract concepts in right and

left hemispheres of complete
Neuropsychologia 1986; 24: 881–7.

commissurotomy

subjects.

Davidoff JB. Cognition through color. Cambridge (MA): MIT
Press; 1991.
Davidoff JB, Ostergaard AL. Colour anomia resulting from
weakened short-term colour memory: a case study. Brain 1984;
107: 415–31.
Dejerine J. Contribution a l’étude anatomo-pathologique et clinique
des différentes variétés de cécité verbale. Cécité verbale pure avec
integrité de l’écriture spontanée et sous dictée. C. r. Seances Biol
1892; 4: 61–90.
De Renzi E, Spinnler H. Impaired performance on color tasks in
patients with hemispheric damage. Cortex 1967; 3: 194–217.
De Renzi E, Faglioni P, Scotti G, Spinnler H. Impairment of color
sorting behavior after hemispheric damage: an experimental study
with the Holmgren skein test. Cortex 1972a; 8: 147–63.
De Renzi E, Faglioni P, Scotti G, Spinnler H. Impairment in
associating colour to form, concomitant with aphasia. Brain 1972b;
95: 293–304.
De Renzi E, Motti F, Nichelli P. Imitating gestures: a quantitative
approach to ideomotor apraxia. Arch Neurol 1980; 37: 6–10.
Gazzaniga MS. Right hemisphere language following bisection. A
20-year perspective. Am Psychol 1983; 38: 525–37.
Geschwind N. Disconnexion syndromes in animals and man. 1.
Brain 1965; 88: 237–94.
Geschwind N, Kaplan E. A human cerebral disconnection syndrome.
A preliminary report. Neurology 1962; 12: 675–85.
Goldstein MN, Joynt RJ. Long-term follow-up of a callosalsectioned patient. Arch Neurol 1969; 20: 96–102.
Gordon
HW.
Neuropsychological
sequelae
of
partial
commissurotomy. In: Boller F, Grafman J, editors. Handbook of
neuropsychology, Vol. 4. Amsterdam: Elsevier; 1990. p. 5–97.
Gordon HW, Bogen JE, Sperry RW. Absence of deconnexion
syndrome in two patients with partial section of the neocommissures.
Brain 1971; 94: 327–36.
Hécaen H, Kremin H. Reading disorders resulting from left
hemisphere lesions: aphasic and pure alexias. In: Whitaker H,
Whitaker HA, editors. Studies in neurolinguistics, Vol 2. New York:
Academic Press; 1977. p. 269–329.
Lewandowsky M. Ueber Abspaltung des Farbensinnes. Mschr
Psychiat Neurol 1908; 23: 488–510.
Liepmann H. Die linke Hemisphaere und das Handeln. Münch Med
Wschr 1905; 49: 2322–6, 2375–8.
Miller RG Jr. Simultaneous statistical inference. 2nd ed. New York:
Springer-Verlag; 1981.
Nass RD, Gazzaniga MS. Cerebral lateralization and specialization
in human central nervous system. In: Mountcastle VB, Plum F,
Geiger SR, editors. Handbook of physiology, Sect. 1, Vol. 5, Pt. 2.
Bethesda (MD): American Physiological Society; 1987. p. 701–61.
Poncet M, Ali Chérif A, Choux M, Boudouresques J, Lhermitte F.
Etude neuropsychologique d’un syndrome de déconnexion calleuse

Left-hand-impaired letter and colour recognition
totale avec hémianopsie latérale homonyme droite. Rev Neurol
(Paris) 1978; 134: 633–53.
Ross GJS. Nonlinear estimation. New York: Springer-Verlag, 1990.
Schott B, Michel F, Michel D, Dumas R. Apraxie idéomotrice
unilatérale gauche avec main gauche anomique: syndrome de
déconnexion calleuse? Rev Neurol (Paris) 1969; 120: 359–65.
Sevush S, Heilman KM. A case of literal dyslexia: evidence for a
disconnection syndrome. Brain Lang 1984; 22: 92–108.
Sittig O. Stoerungen im Verhalten gegenueber Farben bei
Aphasischen. Mschr Psychiat Neurol 1921; 49: 63–8, 169–87.
Sweet WH. Seeping intracranial aneurysm simulating neoplasm.
Arch Neurol Psychiatry 1941; 45: 86–104.
Trevarthen C. Integrative functions of the cerebral commissures.
In: Boller F, Grafman J, editors. Handbook of neuropsychology,
Vol. 4. Amsterdam: Elsevier; 1990. p. 49–83.
Volpe BT, Sidtis JJ, Holtzman JD, Wilson DH, Gazzaniga MS.

1467

Cortical mechanisms involved in praxis: observations following
partial and complete section of the corpus callosum in man.
Neurology 1982; 32: 645–50.
Zaidel E. Auditory vocabulary of the right hemisphere following
brain bisection or hemidecortication. Cortex 1976; 12: 191–211.
Zaidel E. Unilateral auditory language comprehension on the Token
Test following cerebral commissurotomy and hemispherectomy.
Neuropsychologia 1977; 15: 1–18.
Zaidel E. Language functions in the two hemispheres following
complete cerebral commissurotomy and hemispherectomy. In: Boller
F, Grafman J, editors. Handbook of neuropsychology, Vol. 4.
Amsterdam: Elsevier; 1990. p. 115–50.
Zaidel E, Sperry RW. Some long-term motor effects of cerebral
commissurotomy in man. Neuropsychologia 1977; 15: 193–204.

Received January 5, 1998. Revised February 24, 1998.
Accepted March 9, 1998.