BRAIN

AND

LANGUAGE

19, 346-356 (1983)

Ideogram Writing in a Disconnection Syndrome
ATSUSHI

YAMADORI,

TATSUYA

Kobe

University

NAGASHIMA,
School

AND NORIHIKO

TAMAKI

of Medicine

A case of disconnection-type agraphia coupled with alexia was reported. The
patient showed several asymmetrical manual capacities between the two hands,
i.e., dissociated difficulty of Kanji (ideogram) writing between the two hands,
left unilateral difficulty of Kana (phonogram) writing, right unilateral dyscopia
of letters as well as geometrical figures, and right unilateral difficulty in drawing
without a model. Anatomically, lesions involved most of the corpus callosum in
its posterior portion including the splenium and the left medial occipital lobe.
From these data, a possible liguistic capacity of the right hemisphere was suggested.

INTRODUCTION
One of the prominent effects of destruction of the corpus callosum in
human beings is that linguistic information of the left hemisphere (in the
majority of cases of right-handers) stops arriving at the right hemisphere.
One sign of this effect is agraphia confined to the left hand (Geschwind
& Kaplan, 1962; Bogen, 1969). The linguistic nature of this disconnectiontype agraphia has been made clear in recent years (Geschwind, 1965;
Zaidel & Sperry, 1977), but its exact nature still remains unsettled.
Reports of cases from countries where nonalphabetic writing systems
are being practiced might contribute to further interests in this difficult
issue. In a report of left-sided disconnection agraphia from Japan these
authors have pointed out that Kanji (ideogram) and Kana (phonogram)
writing were differentially
affected, the latter being more disturbed than
the former (Yamadori et al., 1980). Sugishita and others reported a similar
phenomenon (Sugishita et al., 1980).
This is another case report of dissociated agraphia of the disconnection
type for Kanji and Kana letters. A complex combination of lesions caused
destruction of the posterior half of the corpus callosum and the left medial
Send requests for reprints to Atsushi Yamadori, M.D., Neurology Service, Hyogo Brain
and Heart Center at Himeji, 520 Saisho-Ko, Himeji, 670, Japan.
346
0093-934X/83 $3.00
Copyright
All rights

0 1983 by Academic
Press, Inc.
of reproduction
in any form reserved.

IDEOGRAM

WRITING

347

occipital lobe. As a result a unique selective affection of written symbols
resulted. Analysis of the symptom complex led us to a hypothesis that
the neural substrate of both Kanji and Kana graphemes necessary for
writing is stored bilaterally, while the neural substrate for ordering these
graphemes into a meaningful sequence is confined to the left hemisphere.
REPORT

OF A CASE

An 18-year-old right-handed female high school student suddenly developed a headache and lost consciousness on a road while traveling on
August 5, 1980. In the hospital she was in a coma and anisocoria (left
larger than right) was noted. Immediate arteriography revealed a huge
arteriovenous
malformation
in the area of the posterior portion of the
corpus callosum. Computed tomography revealed a high-density area in
the posterior portion of the body of lateral ventricle on both sides and
in the white matter neighboring the left trigone.
Craniotomy
was immediately
performed to evacuate the hematoma.
Slightly left of the corpus callosum, the roof of the body of the left lateral
ventricle was excised in its posterior portion for about 3 cm. The blood
in the ventricle was then evacuated. Resection of the arteriovenous malformation was not tried. Postoperative course was complicated by development of meningitis. She was transferred to the Department of Neurosurgery, Kobe University School of Medicine, for further treatment
on September 29, 1980.
On neurological examination, mild right-sided weakness and slight rightsided hyperreflexia were present. Plantar reflex was extensor on the right.
Sensation was decreased for all modalities on the right. Two-point discrimination
and point localization,
as well as stereognosis, were also
defective on the right. Perimetry revealed dense congruent right homonymous hemianopsia with macular split.
The neuropsychological
findings remained essentially unchanged up to
the day of her operation on December 1, 1980. She was alert, attentive,
and cooperative. Recent memory was moderately impaired. Recall of
three items after 5 min was faulty. Remote memory for events prior to
the onset of the present illness was well preserved. Impoverishment
of
spontaneous speech was noted but articulation as well as grammatical
forms was normal. Repetition was done easily. Comprehension of spoken
language was intact. Mild anomia was present. Neither ideomotor nor
ideational apraxia was present bilaterally.
Visual recognition was well
preserved for common objects, pictures, and colors. Ability of tactile
recognition of the left hand was intact. Details of major defects are
described below.
Reuding Performance

The patient had difficulty reading. She read presented materials slowly
and hesitantly, pointing at each letter with the hand. Paralexia was present

348

YAMADORI,

NAGASHIMA,

AND TAMAKI

for single letters as well as for words. She was able to read about 80%
of the test material correctly. Performance of oral reading for Kanji and
Kana letters was essentially at the same level. Comprehension of written
material was also defective. But meaning of Kanji was recognized better
than Kana word. This was well manifested by the abundant semantic
paralexia for Kanji word. Number reading was also defective. Kinesthetic
cueing (Schreibendes Lesen) was effective.
Writing Performance

Writing performance was repeatedly tested with each hand. The ability
was different depending upon which hand was tested.
The RIGHT-hand writing was completely normal for Kana usage. Letters,
words, and sentences were flawlessly written to dictation. However, for
Kanji letters, many incomplete graphemes were produced. Most often
only a portion of a Kanji grapheme was produced. Unnecessary iteration
or omission of stroke was also noted. Thus only half of the tested Kanji
letters were correctly written.
LEFT-hand
writing performance was opposite that of the right hand.
This hand was totally unable to produce correct words or sentences in
Kana. However, Kana grapheme itself was correctly produced most of
the time. For instance, a four-letter word, SU-I-E-I
(swimming) was
written correctly by the right hand but the left hand produced a fourletter jargon, SU-SE-I-YA
for the same word. Letters produced were
often perseverative. On the contrary Kanji writing by the left hand was
much superior to that by the right. Often the left hand was able to write
the very letter the right hand had failed. For instance, a Kanji letter UMA (horse) which was not written by the right hand was correctly written
by the left hand with ease. It has to be noted that Kanji writing by the
left hand was not perfect as was the case with Kana by the right hand.
There were many errors. But the nature of errors was different from
that of the right hand. In contrast to graphic errors of the right hand,
they consisted of perseveration of previously produced Kanji or semantic
substitution of a required Kanji letter. For example, a word meaning
foreign country made of two Kanji letters was substituted by a two-letter
Kanji word meaning overseas.
The difference in writing performance between the two hands was also
noted in dictation of digit series. Thus the right hand could produce the
five-digit series 1 3 8 4 0 with ease but the left hand produced a wrong
sequence, 1 3 6 5 4 0.
Copying and Drawing Performance

COPYING
OF LETTERS
by the right hand was defective. She would
give up copying after having tried a few strokes. No difference was
present between Kanji and Kana letters. On the contrary copying of
letters by the left hand was quite good. But success of copying did not

IDEOGRAM

WRITING

349

FIG. I. Example of Kana writing. Right column in descending order, a correct fourletter word HO,U,SO,U (broadcasting) by the right hand. Left column in descending order,
a jargon JI,YU.U,MA
and an unidentifiable letter by the left hand for the same target
word.

necessarily lead to restoration of reading. The same dissociated performance
between the two hands was also seen in copying geometrical figures.
For example, she could not copy a cube with the right hand but could
copy it correctly with the left hand.
DRAWING
WITHOUT
A MODEL was also dissociated. The command
“draw a square and then draw a circle in it” was beyond the ability of
the right hand. Only perseverative squares were produced. But the left
hand drew the pattern easily. If only the right hand was tested, this
result might have been taken as a sign of comprehension
difficulty.
To sum up, the patient showed alexia of the right visual field, dissociated
difficulty of Kanji writing between the two hands, left unilateral difficulty
of Kana writing, right unilateral difficulty of copying letters and geometrical
figures, and right unilateral difficulty of drawing geometrical figures without
a model.
Neurorudiological

Evaluation

Computed tomography (CT) scan of the brain with enhancement showed
a high-density area in the region of the posterior two-thirds of the corpus

FIG. 2. Example of Kanji writing. Left column in descending order. a two-letter word
HO-U, SO-U (broadcasting) by the left hand. For the same word the right hand produced
only a portion of the frst letter HO-U. as shown to the right.

350

YAMADORI,

NAGASHIMA,

AND TAMAKI

FIG. 3. Writing of digit series. Right column was written by the right hand and all are
correct. The left column was written by the left hand. Only the second series (8 5 6) is
correct. The columns should be identical.

callosum mainly on the left side. The medial part of the left occipital
lobe showed irregular enhancement coupled with a low-density area. The
left trigone was enlarged. The coronal section showed an enhancing lesion
in the posterior corpus callosum protruding into the left lateral ventricle.
The deep white matter of the left cerebral hemisphere neighboring the
posterior end of the body of the lateral ventricle showed a low-density
area.
Arteriographic
study confirmed a large arteriovenous
malformation
situated at the posterior part of the corpus callosum on the left side.
Main feeders consisted of the left and right pericallosal arteries, left

FIG. 4. Copy of a cube. The upper one is a model. The lower left cube was copied by
the left hand and is essentially correct. The lower right was copied by the right hand.

1DEOGRAM

WRITING

TABLE
SUMMARY

Kanji writing
Kana writing
Copying of letters
Copying and drawing of
geometrical figures

OF RIGHT-LEFT

3.51

I
DISSOCIATION

Right hand

Left hand

Graphic difficulty
Normal
Faulty
Faulty

Paragraphia
Jargon
Good
Good

posterior choroidal artery, and left posterior cerebral artery. They drained
to veins of Galen and Labbe.
Thus lesions were composed of destruction of the posterior part of
the corpus callosum, medial left occipital lobe, and deep left parietal
white matter. In addition, the posterior 3 cm of the medial white matter
on the roof of the left lateral ventricle was sectioned in the first operation,
adding damage to callosal fibers. Also at the later operation destruction
of the splenium by hematoma was confirmed.
Hospitul Course

On December 1. 1980, the patient underwent an operation and most
of the mass was resected. On February 2, 1981, the last operation was

FIG. 5. CT scan with enhancement. Most of the corpus callosum and the left lateral
ventricle is occupied by an enhancing mass. In addition, the medial portion of the left
occipital lobe shows a low-density area.

352

YAMADORI,

NAGASHIMA,

AND TAMAKI

FIG. 6. Coronal CT of the same patient. Enhancing callosal lesion is exclusively situated
in the left side. A low-density area is seen in deep white matter around the left lateral
ventricle.

performed to remove the remainder of the AV malformation. Unfortunately
her memory, language, as well as the right hemiparesis deteriorated after
the operation. She became unable to maintain a long attention span. As
a result examination became very sketchy, but the essential feature of
the asymmetrical writing performance between the two hands remained
unchanged.
DISCUSSION
Lesion distribution
of the present case could be summarized very
schematically as an addition of an anteriorly extended callosal lesion to
the classical combination of splenial and left medial occipital lesion often
seen in cases of left posterior cerebral artery infarction. Neuropsychologically it can be understood as a combination of left unilateral agraphia
and alexia without agraphia. Thus the case is unique in a sense that
symptoms are mostly limited to affection of written language.
As has been described in several previous papers (see, for example,
Yamadori, 1973, the Japanese script system is composed of two different
types of letters, i.e., Kanji (Chinese characters or ideograms) and Kana
(phonograms).
Kanji in Japanese usage represents various substantive
concepts ranging from abstract ideas, such as pleasure, to concrete objects,
such as an egg. These concepts are expressed in single Kanji or plural

IDEOGRAM WRITING

353

Kanji combinations. These Kanji words naturally have their own linguistic
names, i.e., phonetic labels. But in usual silent reading these words are
very often scanned only for meanings. As a matter of fact, an average
Japanese would have many Kanji vocabulary words whose meanings are
quite familiar but whose exact phonetic labels are not. Kana, on the
other hand, represents only sound. A Kana letter represents a syllable.
They are small in number (46). With Kana all Japanese expressions are
possible. Kanji has no such function. In modern Japanese, Kanji is used
for a substantive portion of a sentence and Kana is used for the remainder.
Dissociation of Letter Writings between the TMY~Sides
Kanu letters. The RIGHT hand showed the normal ability of language
expression with Kana letters. All words and sentences produced in Kana
by this hand were linguistically
correct.
Her LEFT hand, on the contrary, produced correct Kana letters but
in nonmeaningful
sequence. The same phenomenon was also noted in
number dictation.
These two aspects of the left Kana agraphia, correct grapheme production
and incorrect grapheme sequencing, lead to the following speculation.
The right hemisphere seems to possess neural substrate for motor output
of letter graphemes just like the left hemisphere. What the right hemisphere
lacks is the higher linguistic function. i.e., how to combine the available
units into a meaningful series, which is a left-hemisphere property.
This line of reasoning is consistent with the recorded performance of
the left hand of the commissurotomized
or the commissure-injured
patients.
Thus a patient of Levy et al. (1971) wrote “PI” and “PENCIL”
in
capital letters for a tobacco pipe. A patient of Geschwind and Kaplan
(1962) wrote “yonti”
for “yesterday.”
Did neural impulses for these
letters also come from the left hemisphere? It seems more natural to
assume that these letters P,I,E,N,C,I,L,y,o,n,t,
and i must be the property
of the right hemisphere, while “pipe” or “yesterday” may not. The right
hemisphere tries to mobilize these letter units, but without the correct
plan for sequencing only jargon is produced.
Kanji fetters. In Kanji writing both hands showed subnormal capacity
but the underlying mechanism seemed to be qualitatively
different.
The RIGHT hand, having shown no difficulty in Kana writing, had
difficulty in realizing Kanji graphemes. The Kanji graphemes produced
were similar in their Gestalt to correct ones but not precise. This was
not a linguistic defect.
A similar kind of defect limited to Kanji letters has long been known
to occur in cases of alexia without agraphia in Japanese patients (Miura,
1933). The features of this Kanji dysgraphia of pure alexics were summarized recently by Iwata (1980). Of these, partial production of a target
grapheme and, duplication or omission of a constituent stroke are important.

354

YAMADORI,

NAGASHIMA,

AND TAMAKI

This is exactly what happened in the present case. As in alexia without
agraphia which results in damage to the left occipital lobe and the splenium,
our patient also showed the same lesional distribution
and the same
clinical picture in addition to the left agraphia. The occipital lesion may
be responsible for this Kanji dysgraphia of the contralateral hand. Because
Kanji has more complicated
graphic structures than Kana, an intact
contralateral occipital visual system may be necessary for normal writing
of Kanji.
The LEFT hand showed remarkable capacity in that it could write the
very Kanji word that the right hand failed to produce, which may reflect
the intactness of the right occipital lobe. The fact is doubly interesting
because this left hand could not produce the same word as a combination
of Kana letters. Thus it is quite likely that the right hemisphere has the
neural substrate for Kanji graphemes just as it does for Kana and digit
graphemes. Since Kanji is an ideogram some meaning could easily be
expressed even with a single letter realization. Surprisingly, many correct
Kanji combinations
were also produced. It is true that errors were also
seen. However, these were paragraphic substitutions in the same semantic
sphere and were not graphic errors. Thus the right hemisphere seems
to possess a certain degree of semantic processing capacity at word level
which is probably unrelated to phonetic sequencing capacity. This vague
semantic capacity without phonetic exactness may have contributed to
the surprising success of correct Kanji letter combinations by the left
hand.
From the above arguments these authors come to the conclusion that
Kanji and Kana letters as an element of linguistic expression are bilaterally
represented. It is a different mode of processing strategy which produced
the apparent dissociation of Kanji and Kana writing in the present case.
This conclusion is consistent with what was proposed in another paper
(Yamadori,
1980a).
Dissociation of Copying and Drawing between the Two Sides
Copying of letters. The present case showed selective difficulty in
copying letters with the right hand. The left hand copied them correctly.
This unilateral dyscopia of letters was first observed in patients of alexia
without agraphia and was explained as a result of isolation of normal
visual perception of letters in the right visual cortex from the left hemisphere
because of the combined lesions of the splenium and the left medial
occipital lobe (Yamadori, 1980b). The same mechanism must be in operation
in the present case.
Copying and drawing of geometrical jigures. Difficulties
in copying
geometrical figures limited to the right hand have been observed in commissurotomized
patients and have been attributed to the disconnection
of the left hemisphere from the dominant visuospatial influence of the

IDEOGRAM

WRITING

355

right hemisphere (Gazzaniga et al.. 1965; Zaidel & Sperry, 1977). In the
present case, defects in spontaneous drawing of figures can be explained
by this hypothesis. But as far as copying is concerned, dyscopia of figures
can be as easily explained by the same mechanism as letter dyscopia
without recourse to the influence of the right parietal lobe. For instance,
dyscopia of cubes limited to the right hand was reported in a case of
alexia without agraphia (Yamadori, 1980b). Thus there is a possibility
that unilateral dyscopia of figures can be caused by different mechanisms.
Reading

Perj2mnance

Moderate alexia in the present case affected both Kanji and Kana to
the same degree. Some semantic processing capacity for Kanji words
was noted but not for Kana words. Based on CT findings, the alexia
seems to represent a typical example of alexia without agraphia. Since
she had a complete right hemianopsia and complete callosal destruction
at least in the posterior portion, her reading capacity can be taken as a
sign of right hemisphere function.
Our evidence obtained from a single Japanese case favors the notion
that the right hemisphere has linguistic capacity to a certain degree.
First, the neural substrate for letter graphemes necessary for writing is
stored. Second, some capacity of semantic processing at word level is
present. This capacity may remain latent and can only be tapped if an
appropriate strategy such as ideogram reading or writing is available.
The conclusion is different from that obtained with commissurotomized
patients by Gazzaniga and Sperry (1967). They concluded that linguistic
expressive capacity seemed to be organized almost exclusively in the
left hemisphere. But later studies (Levy et al., 1971; Gazzaniga et al.,
1977) showed that there could be patients whose minor hemispheres have
some capacity to express at least simple language through control of the
left hand.
Levy (1978) concluded that language comprehension is apparently quite
good in the right hemisphere, for both spoken and written words and
phrases in spite of its essentially total lack of phonetic analyzers. This
comment seems to be quite true also for written expression at least for
the present case. Since our experience lacks encounters with patients
with complete callosal destruction, this conclusion remains tentative.
SUMMARY

A case of disconnection-type agraphia coupled with alexia was reported.
Neuropsychologically
the patient showed several asymmetrical
manual
capacities between the two hands, i.e., dissociated difficulty of Kanji
(ideogram) writing between the two hands, left unilateral difficulty of
Kana (phonogram) writing, right unilateral difficulty of copying letters
and geometrical figures, and right unilateral difficulty of drawing geometrical
figures without model.

356

YAMADORI,

NAGASHIMA.

AND TAMAKI

Anatomically lesions involved most of the corpus callosum in its posterior
portion including the splenium and the left medial occipital lobe.
Analysis of the data led us to following conclusions.
1. The right hemisphere seems to possess neural substrate for letters
regardless of its type.
2. The right hemisphere seems to have no capacity to organize Kana
or digits into a correct sequence.
3. The right hemisphere seems to have some capacity to process words
at the semantic level.
4. For correct Kanji writing, particularly visual support seems to be
a prerequisite.
This visual support seems to come from the normally
functioning occipital lobe contralateral to the writing hand.
REFERENCES
Bogen, J. E. 1969. The other side of the brain. I. Dysgraphia and dyscopia following
cerebral commissurotomy. Bulletin of the Los Angeles Neurological
Society,
34, 73105.

Gazzaniga, M. S., Bogen, J. E., & Sperry, R. W. 1965. Observations on visual perception
after disconnection of the cerebral hemispheres in man. Bruin, 88, 221-236.
Gazzaniga, M. S., & Sperry, R. W. 1967. Language after section of the cerebral commissures.
Bruin, 90, 131-148.
Gazzaniga, M. S., Ledoux, J. E., and Wilson, D. H. 1977. Language, praxis and the right
hemisphere: Clues to some mechanisms of consciousness. Neurology,
27, 1141-l 147.
Geschwind, N., & Kaplan, E. 1962. A human cerebral deconnection syndrome, Neurology,
12, 675-685.

Geschwind, N. 1965. Disconnexion
and 585-644.
Iwata, M. 1977. Neuropsychological
rological

Sciences,

Tokyo,

syndromes in animals and man. Bruin,

88, 237-294

aspects of pure alexia syndrome. Advances

in Neu-

21, 930-940.

Levy, J., Nebes, R. D., & Sperry, R. W. 1971. Expressive language in the surgically
separated minor hemisphere. Cortex, 7, 49-58.
Levy, J. 1978. Lateral differences in the human brain in cognition and behavioral control.
In P. A. Buser and A. Rougeful-Buser (Eds.), Cerebral correlates of conscious experience.
INSERM
Symposium
No. 6. Amsterdam: North-Holland. Pp. 285-298.
Miura. T. 1933. Beitrage zur Symptomatologie der reinen Alexie. Psychiutria
et Neurologica
Japonica,
36, 326-371.
Sugishita, M., Toyokura, Y., Yoshioka, M., & Yamada, R. 1980. Unilateral agraphia after
section of the posterior half of the truncus of the corpus callosum. Bruin and Language.
9, 215-225.
Yamadori, A. 1975. Ideogram reading in alexia. Bruin, 98, 231-238.
Yamadori, A., Osumi, Y., Ikeda, H., & Kanazawa, Y. 1980. Left unilateral agraphia and
tactile anomia. Disturbances seen after occlusion of the anterior cerebral artery. Archives
of Neurology,
37, 88-91.
Yamadori, A. 1980. Cerebral laterality in Kanji-Kana processing. Advances in Neurological
Sciences,
Tokyo, 24, 556-564.
(a)
Yamadori, A. 1980. Right unilateral dyscopia of letters in alexia without agraphia. Neurolagy,
30, 991-994. (b)
Zaidel, D., & Sperry, R. W. 1977. Some long-term motor effects of cerebral commissurotomy
in man. Neuropsychologia,
15, 193-204.