NOTE
PURE WORD DEAFNESS DUE TO LEFT HEMISPHERE
DAMAGE
Nobuyoshi Takahashi\ Mitsuru Kawamura\ Hitoshi Shinotout, Keizou Hirayama1
Kimitaka Kaga2 , and Mitsuko Shindo 2
'
eDepartment of !'feurology, School of Medicine, Chiba University; 2Department of
Otorhmolaryngology, School of Medicine, Teikyo University)

INTRODUCTION

Pure word deafness is defined as a selective deficit of auditory language comprehension,
and is distinguished from cortical auditory deficits for the absence of deficits in understand­
ing music and environmental sounds. Lichtheim (1885) suggested that this syndrome was due
to the isolation of Wernicke's area from auditory linguistic input following damage to the
subcortex of the left temporal lobe. Liepmann and Storch (1902) supported this view in the
light of autopsy studies and more recently Geschwind (1965) has advocated a similar view.
Thereafter, many cases labeled as pure word deafness have been reported with anatomical
documentation of either unilateral left or bilateral cortical/subcortical damage to the tem­
poral lobe. The number of cases that comply with the above definition of pure word deaf­
ness is, however, much smaller and in only a few has the location of the lesion been precisely
delineated (Henschen, 1917; Hamanaka, Asano, Morimune eta!., 1980; Kamei, Nakane,
Nishimaru eta!., 1981; Metz-Lutz and Dahl, 1984; Yaqub, Gascon, Alnosha et al., 1988).
We report a case of pure word deafness, following a left thalamic hemorrhage, whose
deficits were selective for auditory language comprehension and whose lesion was identified
by means of computed tomography (CT), magnetic resonance imaging (MRI) and positron
emission tomography (PET).
CASE REPORT

On October 17, 1982, a 55 year-old right-handed man with a 10 year history of hyper­
tension was admitted to the Chiba Emergency Medical Center following onset of headache,
disturbance of consciousness, right hemiparesis and aphasia. He had graduated from junior
high school, from which time he had been employed as an automobile mechanic. CT ob­
tained on admission showed left thalamic bleeding. The hematoma had penetrated the third
ventricle and extended rostrally to the left centrum semiovale and caudally to the midbrain.
Subsequently, the deficits of consciousness and paresis gradually subsided and no ab­
normalities of speech remained. However, although he could hear the voices of other peo­
ple, he realized that he was almost totally incapable of understanding what was being said.
Since these symptoms persisted, he was transferred to the Department of Neurology of Chi­
ba University Hospital on July 21, 1984. On examination, consciousness was normal and he
was fully oriented. Visual acuity was normal, but he showed lower right quadrantopsia on
confrontation. His visual reflexes to light were brisk. Vertical eye movements were limited.
There was a slight diplopia when the gaze was to the right. Swallowing and articulation were
normal. On the right there was a hemiparesis with increased tendon reflexes and a positive
Babinski sign and sensory deficits including the right side of the face.
Neuropsychological examination did not reveal ideational apraxia, deficits of construc­
tional behavior, hemispatial neglect or visual agnosia. WAIS verbal IQ was 89, performance
IQ was 90 and total IQ was 89 (Because of severe difficulty in auditory comprehension, pre­
sentation of some problems was done visually).
Cortex, (1992) 28, 295-303

296

N. Takahashi and Others
CLINICAL SYMPTOMS AND EXAMINATION RESULTS OF THE PURE WORD DEAFNESS

Clinical Symptoms
Spontaneous speech was fluent with normal speed and intonation. The amount of speech
and voice volume were unchanged from the premorbid state. There were no word substitu­
tions or paraphasias. Reading aloud, reading comprehension and spontaneous writing were
normal. In contrast, there were severe deficits in auditory verbal comprehension and he was
almost totally incapable of understanding individual words.
Subjectively, he could discriminate a human voice from other sounds, but he stated that
it sounded like a foreign language, except for his own speech which sounded normal. The
hearing abnormality was particularly strong when he could not see the speaker (e.g. when lis­
tening to the telephone, radio or a tape cassette) and with unfamiliar voices. Clearly articu­
lated words and words frequently used in daily conversation were easier to understand than
poorly articulated and uncommon words.
He was however capable of identifying the number of syllables in a word and the number
of words in a sentence. Moreover, if a word was pronounced slowly or was repeated, his un­
derstanding improved. Improvements also obtained with lip-reading.
In contrast, he could immediately identify environmental sounds, such as animal cries or
a ringing telephone, as well as musical instruments and the names of musical melodies. All
of the above symptoms remained unchanged throughout the period of testing and neurora­
diological examination.
Examination Results (September 1984 to December 1986)
Auditory Testing (September 15, 1984)
In pure tone audiograms, he had moderate deficits at high frequencies, but not in the
range of conversational sounds, except for a slight decrease at 2000 Hz. The right ear was
mildly worse than the left at all frequencies above 250 Hz. Auditory brainstem and slow ver­
tex evoked potentials were normal.
Western Aphasia Battery (WAB) (December 10, 1986)
(Lip-reading was permitted when auditory comprehension, oral repetition and writing to
dictation tests were given).
Auditory comprehension and repetition were markedly impaired, as well as writing to
dictation (Figure 1). In spontaneous writing, he made only two errors (once replacing a Kana
character by another and once omitting a Kana character) out of 7 sentences, but otherwise
his writing showed no abnormalities. Speech, reading and object naming were normal.
Auditory Recognition Tests (September 15, 1984)
Speech sounds, environmental sounds and music recognition were tested following the
methods of Shindo, Kaga and Tanaka (1981). Five separate tests of speech recognition were
carried out.
(1) Mono-syllable Identification. The patient was required to repeat the 53 CV syllables
of the Japanese writing system pronounced in random order by the examiner.
(2) Word Recognition. The examiner read out 20 words such as SAKE and TAKE, and
the patient pointed at the corresponding card presented with three distractors and showing
both a picture and the printed word. The same procedure was followed for tests (3) through
(5).

(3) Long and Short Vowel Recognition. Twenty trials involving pairs of words differing
for the length of a vowel, such as KAHDO and KADO, were given.
(4) Accent Recognition. Twenty trials involving pairs of words differing for accent such
as HASHI and HASHI were given.
(5) Prosody Recognition. The patient was required to identify from prosody the mood
expressed by 14 short sentences.
These tests were given separately to the left and the right ear.
In the environmental sound test, the patient listened to an audio tape on which a total of

Pure Word Deafness

I.

Spontaneous speech

II.

Auditory comprehension

Ill.

Auditory repetition

IV.

Naming (aloud and comprehension )

v.

Reading

VI.

Writing
Spontaneous writing of his own name and
address

297

Spontaneous writing
Writing to dictation
Spontaneous writing of Kana and numbers
Writing to dictation of individual Kana,
Kanji and numbers
Copying of Kana and Kanji
VII.

Praxis

VIII. Constructional behaviour

0

50

100%

Fig. 1 - Western Aphasia Battery. Notable deficits of auditory comprehension, repetition and writ­
ing from dictation. The performance was normal in the other subtests.

24 familiar environmental sounds, such as a ringing telephone, a eat's meow, the sound of
wind, etc., had been recorded. He was required to name the sound and to point to an appro­
priate picture corresponding to it, presented with three distractors (Kaga, Shindo and Su­
gishita, 1987). For human voice recognition, the spontaneous conversation of seven people
(his wife, son, daughter, the patient himself, one female and two male examiners) was re­
corded and he was required to identify the voices by naming.
For the music recognition test, tape recordings of the following six categories were made.
(1) Pitch Discrimination. Two notes of different pitch were presented sequentially and the
patient was required to identify the higher.
(2) Loudness Discrimination. Two notes of different loudness were presented sequen­
tially and he was required to identify the louder.
(3) Rhythm Discrimination. Two short rhythmical sequences were presented one after the
other and he was asked whether they were same or different.
(4) Melody Discrimination and Identification. Two short melodies were played one after
the other and he was asked whether they were the same or different. He was also played mel­
odies that were familiar to him and asked to point to one of four drawings representing the
theme of the melody.
(5) Musical Instrument Identification. After hearing the sound of a musical instrument,
he was required to point to one of four pictures identifying the instrument.
(6) Musical Mood Identification. After listening to one of five short melodies (two var­
ieties of energetic music, one sad, one happy and one melodious tune), he was required to
choose among three words that corresponding to the mood of the music.

N. Takahashi and Others

298

TABLE I

Tests ofAuditory Perception
Correct response
Speech Recognition Tests
1. mono-syllable identification (53 CV phonemes)
2. Word recognition (20 words)
3. Long and short vowel recognition (20 words)
4. Accent recognition (20 words)
5. Prosody recognition (14 sentences)
Environmental Sound Recognition Tests
1. Environmental sound identification (pointing)
2. Environmental sound identification (naming)
3. Human voice identification
Music Recognition Tests
1. Pitch discrimination
2. Loudness discrimination
3. Rhythm discrimination
4. Melody discrimination
Melody identification
5. Musical instrument identification
6. Musical mood identification

R. ear

L. ear
211Jo
50%
70%
55%
0%

2%

40%
50%
65%
0%

24/24 (100%)
21/24 (88%)
517
5/8
4/4
l/4
3/6
10/11
8/10
515

Table I shows that, with speech sounds, he manifested severe deficits in the monosyllable
test and the prosody test, and moderate deficits in the other tests. His errors tended to be
phonemic rather than semantic. In contrast, with environmental sound tests, pointing to pic­
tures was 1000Jo correct and naming 88% correct. With the exception of one confusion be­
tween his wife's voice and one of the female examiners' voice, he was flawless in the recog­
nition of human voices. He showed moderate deficits in the rhythm and melody tests, but
good performance in the other musical tests.
Neuroradiologica/ Evidence

CT scan, obtained on January 29, 1987, showed a low density area (LDA) that extended
from the left thalamus and posterior internal capsule to the white matter of the left temporal
and parietal lobes.
In a horizontal MRI image passing through the thalamus [IR method, repeat time (Tr)
of 2080 msecs and inversion time (Ti) of 500 msecs (March 29, 1985)], the lesion of the white
matter was more clearly delineated and an area of low signal intensity (LSI) was seen in the
left temporal white matter running next to the lateral wall of the left lateral ventricle (Figure
2). The left auditory radiations were damaged. A region of LSI was evident in the left thal­
amus and internal capsule. In a more rostral section, an area of LSI was seen in the white
matter of the temporal and parietal lobes running lateral to the left lateral ventricle and ex­
tending to a portion of the centrum semiovale above. There was no notable abnormality of
the left temporal cortex and in the right hemisphere. In a coronal section passing through the
posterior part of the temporal lobe (Figure 3a), LSI was seen in the left temporo-parietal
white matter extending from the wall of the left lateral ventricle to the centrum semiovale.
In a sagittal section 30 mm to the left of the midline (Figure 4a), the left temporal and par­
ietal white matter above the lateral ventricle showed LSI. Using the SE method, high signal
intensity (HSI) regions were seen in the same areas [Tr = 2080 msecs, echo time (Te) = 80
msecs] (Figures 3b and 4b).
In a PET study using the 150-steady state method (March 4, 1987), cerebral metabolic rate
for oxygen (CMR0 2 , Figure 5) and cerebral blood flow (CBF) were measured. The CMR02
of the left thalamus was markedly reduced, and that of the white matter of the left temporal

Pure Word Deafness

299

Fig. 2 - Horizontal MRI. A low signal intens­
ity area encroaches upon the left thalamus, poster­
ior internal capsule and white matter of the left
temporal lobe. The large arrow indicates the left
primary auditory area. The thin arrow indicates
damage to the left auditory radiations.

Fig. 3 - Coronal MRI using theIR method (a) and theSE method (b). The lesion running along the
cerebral ventricle in the white matter of the left temporal and parietal lobes is seen as a low signal intens­
ity area using the /R method and as a high intensity area using theSE method (arrows).

and parietal lobes was also low. Although the oxygen consumption of the left auditory cor­
tex and the superior temporal gyrus, including Wernicke's area, was somewhat lower than
that on the right (3.48 vs. 3.95 ml/100 ml/min), both were within a normal range. Similar
CBF findings were obtained.

300

N. Takahashi and Others

Fig. 4- Sagittal MRI using theIR method (a) and theSE method (b). Th~ lesion in the white matter
of the left temporal and parietal lobes above the lateral ventricle is seen as a low signal intensity area in
theIR method and as a high signal intensity area in theSE method (arrows).

Fig. 5 - PET (CMROJ study shows a marked decrease in the left thalamus, and a decrease centering
on the white matter of the left temporal and parietal lobes. Noteworthy, however, is the fact that oxygen
consumption at the posterior half of the superior temporal gyrus (arrow) is within a normal range.

Pure Word Deqfness

301

DISCUSSION

There are inconsistencies in the terminology used to discuss deficits of higher auditory
functions. In the present work, we define auditory deficits of language as word deafness,
those of environmental sounds (meaningful non-language sounds) as auditory sound agno­
sia (=auditory agnosia) and those of music as sensory amusia. When all three symptoms are
present, the syndrome is called cortical auditory deficits, while deficits of auditory percep­
tion (deafness) due to cerebral lesions are referred to as cortical deafness.
The concept of pure word deafness was introduced by Kussmaul (1877) and Lichtheim
(1885) to define a condition in which spontaneous speech, reading and writing were normal,
but there were deficits in the auditory comprehension of language. It was regarded as a pure
form of aphasia and contrasted with other forms of aphasia in which the "internal language"
also was disturbed. In recent years, the selective impairment of auditory language compre­
hension has been contrasted with the intact recognition of environmental sounds and music
(Bauer and Rubens, 1985).
Most previous case reports of word deafness present difficulties of interpretation either
because they were in fact cases of cortical auditory deficits or because the examination was
not sufficiently complete to discriminate pure word deafness from cortical auditory deficits.
There are only seven cases which meet the criteria for the diagnosis of pure word deafness
(Henschen, 1917; Hemphill and Stengel, 1940; Okada, Hanada, Hattori et al., 1963; Ha­
manaka et al., 1980; Kamei et al., 1981; Metz-Lutz and Dahl, 1984; Yaqub et al., 1988). In
addition, there have been reports of selective deficits in the recognition of environmental
sounds (Spreen, Benton and Fincham, 1965; Fujii, Fukatsu, Watabe et al., 1990) and of sen­
sory amusia (Edgren, 1985; Kohl and Tschabitscher, 1953).
In our case, (1) hearing was normal within the frequency range of normal conversation,
except for a slight reduction at 2000 Hz, (2) writing was almost normal, except for rare par­
agraphias, and spontaneous speech and reading were intact, suggesting that the internal lan­
guage processes were not impaired, (3) environmental sound and music recognition was nor­
mal. For this reason, our case can be considered a particularly "pure" instance of word deaf­
ness. The location of the brain lesion can be summarized as follows: (1) Both MRI and PET
study showed that damage was confined to the left hemisphere, with no functional or mor­
phological abnormality of the right hemisphere. (2) In MRI study, the main lesion was found
in the left thalamus and the white matter of the left temporal and parietal lobes. The lesion
of the white matter run rostrally along the lateral wall of the posterior part of the lateral ven­
tricle and reached the centrum semiovale. No abnormality of the temporal cortex, including
primary auditory areas, Wernicke's area and the underlying white matter, was found. (3)
PET study confirmed a decrease in CMR02 and CBF of the left temporo-parietal area, but
no abnormality of the posterior part of the superior temporal gyrus (including the primary
auditory area and Wernicke's area).
From autopsy studies of pure word deafness, the responsible lesion has been said to oc­
cur at either of these sites: (1) left temporal cortex and subcortex, or (2) bilateral temporal
cortex and subcortex, the latter being more frequently reported. Among the seven cases of
unambiguous pure word deafness, only five have had the location of the lesion clearly iden­
tified (Henschen, 1917; Hamanaka et al., 1980; Kamei et al.., 1981; Metz-Lutz and Dahl,
1984; Yaqub et al., 1988). The lesion was bilateral in the case of Henschen (1917) and in that
of Yaqub et al. (1988), and unilateral on the left in the three remaining cases.
More in detail, the case of Hamanaka et al. (1980) had a left-sided hematoma extending
from the basal ganglia to the above-lying white matter along the posterior portion of the la­
teral ventricle. In the case of Kamei et al. (1981), a left-sided infarct extended from the pos­
terior superior temporal gyrus and adjacent portions of the supramarginal and angular gyri
to the posterior lateral ventricle, making a deep wedge-shaped cut into the subcortical white
matter. The locus of lesion was similar in the case of Metz-Lutz and Dahl (1984): a left-sided
infarct extended from the left temporo-parietal cortex to deep subcortical areas. In other
words, all unilateral cases, including our own, had a subcortical lesion of the left temporal
lobe (particularly of the superior temporal gyrus) and of the white matter lying deeply in the
temporo-parietal lobe near the posterior part of the lateral ventricle. These findings strongly
suggest that this region is crucially associated with pure word deafness.

302

N. Takahashi and Others

Since the early reports of Lichtheim (1885) and Liepmann and Storch (1902), the idea has
been accepted that pure word deafness represents a disconnection syndrome in which Wer­
nicke's area is isolated from auditory linguistic information. Theoretically, three lesional
patterns could produce such a syndrome: (i) interruption of the fibers connecting the left au­
ditory cortex with Wernicke's area, and of the right auditory radiations; (ii) interruption of
the association fibers linking the right auditory cortex with Wernicke's area, and of the left
auditory radiations; and (iii) interruption of the association fibers linking the right auditory
cortex with Wernicke's area, and of the connections between the left auditory cortex and
Wernicke's area (see Figure 6). Due to the rarity of pure word deafness cases, especially of
those in which the lesion has been adequately studied, it has been impossible to distinguish
among these possibilities. The anatomical course of the central auditory pathways has also
remained problematical. Although the course of the radiations linking the medial geniculate
body with the unilateral auditory cortex has been identified (Kretschmann and Weinrich,
1984), little is known about the location of the interhemispheric fibers crossing in the corpus
callosum. Damasio and Damasio (1979) and Alexander and Warren (1988) have suggested
that auditory commissural fibers run backward and upward from the primary auditory cor­
tex, proceed along the wall of the cerebral ventricles lateral to their trigonal portion, and cross
in the posterior part of the corpus callosum. As discussed above, lesions of the white matter
lateral to the posterior part of the lateral ventricle are common to all of the four cases with
unilateral left hemisphere lesions, strongly suggesting that the commissural fibers have been
damaged in this region.
In our case, MRI showed damage to the left auditory radiations, while no notable abnor­
malities were seen in the left auditory area or Wernicke's area in either MRI or PET studies,
which makes it unlikely that the lesion laid between the left auditory cortex and Wernicke's
area. In the light of the above arguments, we submit that in our patient pure word deafness
was due to type (ii) lesional pattern (see Figure 6), namely, to the combined interruption of
the left auditory radiations and the commissural fibers from the right auditory area.
The reason why our patient was able to name auditory sounds [that is, why there was not
an auditory modality-specific anomia, as described by Denes and Semenza, (1975)], is not
entirely clear. Two possibilities come to mind. The first is that there was partial damage to
the commissural fibers connecting the right auditory cortex and Wernicke's area, so that on­
ly the transfer of linguistic information was compromised. Alternatively and taking into ac­
count cases in which only deficits in the perception of environmental sounds have been re­
ported (Spreen eta!., 1965; Fujii et a!., 1990), the existence of an "environmental sound"
center on the right or bilaterally can be hypothesized. On this account, damage may have af­
fected the commissural fibers from the right auditory cortex to Wernicke's area, but not the
commissural fibers from the "environmental sound" center to Wernicke's area. Unfortun­
ately, clear anatomical or physiological support for either of these hypotheses is lacking, and
their distinction remains a task for the future.

(i)

(ii)

(iii)

Fig. 6 - Three types of lesions which could result in pure word deafness. A type (ii) lesion is thought
to be responsible for the pure word deafness in the present case. (A =auditory area; M =medial genicu­
late body; W = Wernicke's area; x = lesion).

Pure Word Deafness

303

ABSTRACT

We report the case of a 55 year-old right-handed man who presented with a long lasting
pure word deafness following left thalamic bleeding. There was no sign of aphasia. The au­
ditory deficit was specific for language, while recognition of music and environmental sounds
was normal. CT, MRI and PET examinations showed that the lesion was anatomically and
functionally confined to the left cerebral hemisphere, mainly the white matter of the tem­
poral and parietal lobes. Wernicke's area was largely preserved. It is proposed that pure word
deafness was consequent to the isolation of Wernicke's area from incoming auditory infor­
mation due to the interruption both of the association fibers from the right auditory area
traveling across the corpus callosum and of the left auditory radiations.

REFERENCES

ALEXANDER, M.P., and WARREN, R.L. Localization of callosal auditory pathways: ACT case study.
Neurology, 38: 802-804, 1988.
BAUER, R.M., and RUBENS, A.B. Auditory agnosia. In K.M. Heilman and E. Valenstein (Eds.), Clinical
Neuropsychology, 2nd ed. New York: Oxford University Press, 1985, pp. 209-213.
DAMASIO, H., and DAMASIO, A. "Paradoxic" ear extinction in dichotic listening: Possible anatomic sig­
nificance. Neurology, 29: 644-653, 1979.
DENES, G., and SEMENZA, C. Auditory modality-specific anomia: evidence from a case of pure word
deafness. Cortex, 11: 401-411, 1975.
EDGREN, J .G. Amusie (musikalische Aphasie). Deutsche Zeitschrift fiir Nervenheilkunde, 6: 1-64, 1895.
FUJII, T., FUKATSU, R., WATABE, S., 0HNUMA, A., TERAMURA, K., KIMURA, 1., SASO, S., and KOGURE,
K. Auditory sound agnosia without aphasia following a right temporal lobe lesion. Cortex, 26: 263­
268, 1990.
GESCHWIND, N. Disconnection syndromes in animals and man. Brain, 88: 237-294, 585-652, 1965.
HAMANAKA, H., ASANO, K., MORIMUNE, S., and SEKO, K. Ein Fall von reiner Worttaubheit ohne akus­
tische Agnosie. Studio Phonologica, 14: 16-24, 1980.
HEMPHILL, R.E., and STENGEL, E. A study on pure word-deafness. Journal of Neurology and Psychia­
try, 3: 251-262, 1940.
HENSCHEN, S.E. Ober die Hiirsphare. Journal fur Psychologie und Neurologie, 22: 319-474, 1917.
KAGA, K., SHINDO, M., and SuGISHITA, M. Speech and sound recognition in auditory pathway. Vocali­
zation vs. non vocalization. The Institute of Electronics, Information and Communication Engi­
neers, Technical Report (Tokyo), 9-16, 1987.
KAMEl, H., NAKANE, K., NISHIMARU, K., SHIRAISHI, K., and MATUO, M. Pure word deafness after cer­
ebrovascular disease: a case report. Clinical Neurology (Tokyo), 21: 402-408, 1981.
KoHL, G.F., and TscHABITSCHER, H. Ober einen Fall von Amusie. Wiener Zeitschrift fiir Nervenheilk­
unde und deren Grenzgebiete, 6: 219-230, 1953.
KRETSCHMANN, H.J., and WEINRICH, W. Neuroanatomie der kraniellen Computertomographie-Grun­
dlagen und klinishe Anwendung. Stuttgart: George Thieme Verlag, 1984.
KussMAUL, A. Disturbance of speech. In H. von Ziemssien (Ed.), Cyclopedia of the Practice of Medi­
cine. New York: William Wood and Co., 1877, pp. 581-875.
LICHTHEIM, L. On aphasia. Brain, 7: 433-484, 1885.
LIEPMANN, H., and STORCH, E. Der mikroskopische Gehirnbefund bei dem Fall Gorstelle. Monatsschrift
fiir Psychiatrie und Neurologie, 11: 115-120, 1902.
METZ-LuTz, M.N., and DAHL, E. Analysis of word comprehension in a case of pure word-deafness. Brain
and Language, 23: 13-25, 1984.
OKADA, S., HANADA, M., HATTORI, H., and SHOYAMA, T. A case of pure word-deafness (About there­
lation between auditory perception and recognition of speech-sound). Studio Phonologica, 14: 58-65,
1963.
SHINDO, M. KAGA, K., and TANAKA, Y. Auditory agnosia following after bilateral temporal lobe lesions.
Report of a case. Brain and Nerve (Tokyo), 33: 139-147, 1981.
SPREEN, 0., BENTON, A.L., and FINCHAM, R.W. Auditory agnosia without aphasia. Archives of Neu­
rology, 13: 84-92, 1965.
YAQUB, B.A., GASCON, G.G., ALNOSHA, M., and WHITAKER, H. Pure word deafness (acquired verbal
auditory agnosia) in an Arabic speaking patient. Brain, 111: 457-466, 1988.
N. Takahashi, M.D., Department of Neurology, School of Medicine, Chiba University, 1-8-1 lnohana, Chiba 280, Japan.