Journal of Neurology, Neurosurgery, and Psychiatry 1982;45:719-724

Alexia and agraphia in Wernicke's aphasia
HOWARD S KIRSHNER, WANDA G WEBB
From the Departments ofNeurology and Hearing and Speech Science, Vanderbilt University School of
Medicine, Nashville, Tennessee, USA
SUMMARY Three patients. with otherwise typical Wernicke' s aphasia showed consistently greater
impairment of reading than auditory comprehension. While this syndrome resembles alexia with
agraphia, the paraphasia of speech, repetition, and naming underline the aphasic nature of the
disorder. Together with previous reports of isolated word deafness in Wernicke's aphasia, these
cases suggest a relative independence of auditory and visual language processing.

Wernicke's aphasia is traditionally characterised' 2
as a disorder of central language processes, affecting
the different modalities of language input and output more or less equally. Thus reading is impaired
along with auditory comprehension, and writing,
like spontaneous speech, is marred by paraphasic
substitutions. Anomia and paraphasic substitutions
in naming and repetition complete the picture. The
equivalence of deficits in reading and auditory comprehension is often explained by the concept that
reading is learned by transcoding of visual language
stimuli to previously acquired auditory word associations, stored in Wemicke's area.3 Disruption of the
auditory association area would thus interfere not
only with auditory comprehension, but also with
reading.
In a previous communication,4 we discussed cases
of Wemicke's aphasia with spared reading ability
and compared this auditory-predominant sub-type
of Wernicke' s aphasia to the traditional syndrome of
pure word deafness. The present report concerns
three cases of Wernicke's aphasia, all secondary to
ischaemic stroke, in which reading was more
impaired than auditory comprehension, both acutely
and after partial recovery. Whereas the auditorypredominant cases of Wernicke's aphasia resemble
pure word deafness, this second, visual-predominant
sub-type resembles the classical syndrome of pure
alexia with agraphia, an acquired illiteracy in which
reading and writing are impaired in the absence of
other language deficits. The paraphasic productions
of our patients in spontaneous speech, naming,
Address for reprint requests: Howard S Kirshner, MD, Department of Neurology, Vanderbilt University School of Medicine,

Nashville, Tennessee 37232, USA.
Received 21 November 1981 and in revised form 6 March 1982.
Accepted 18 March 1982

repetition, ano oral reading, however, distinguish
their deficit from pure alexia without agraphia and
link them more closely to Wernicke's aphasia. These
cases add to previous evidence that the auditory and
visual modalities of language may be affected to different degrees in Wemicke's aphasia and suggest at
least a partial independence of auditory and visual
language processing.
Methods and report of cases
Three aphasic patients, all victims of acute left hemisphere
strokes, were studied within five days after onset and again
four to six weeks later. All had at least a partial high school
education and good premorbid reading skills. The tests
employed are the same as those described previously.4
Case I This 67-year-old right-handed male, in good
health except for hypertension and prior myocardial infarction, entered the hospital after an episode of blurred vision, lightheadedness, nausea, and left-sided weakness. Initial examination disclosed only a left homonymous
hemianopsia and left hemisensory deficit, both of which
resolved over the next few hours. A computed tomographic (CT) scan was normal. Three days later, during
cerebral arteriography, the patient became restless and
"confused". The arteriogram showed nonstenosing plaque
formation at the right carotid bifurcation; the left carotid
bifurcation and intracranial vessels were unremarkable.
On examination the patient repetitively uttered phrases
such as "let me out, . . . let me get over". He followed no
commands, called his daughter "Mama", and made many
paraphasic errors. He had a dense right homonymous
hemianopsia but no apparent motor or sensory deficits. By
the next morning he followed some simple commands but
not multistep directions. He manifested almost total
anomia for objects and colours.
The patient improved steadily over the next few days.
Five days after onset he named eight of ten pictures, five of
six colours, and eight of ten small letters. He correctly
followed seven of eight commands of increasing complexity, either spoken or presented in printed form. On a

719

720
matching-to-sample test battery modified from Sidman et
al,5 he performed with greater than 90% accuracy in
matching single, dictated or printed small letters to printed
capital letter choices and in matching single words presented in printed, dictated, or dictated, spelled form to
printed word choices. He read aloud 10/10 3-letter words.
On the Boston Diagnostic Aphasia Examination
(BDAE),6 performed four days after onset, the patient's
oral expression was fluent but marked by both literal and
verbal paraphasias and word-finding difficulty. Confrontation naming showed prolonged latencies and some errors
on pictures, body parts, and colours. Auditory comprehension subtests showed normal responses to multistep commands and complex ideational material. Repetition was
marked by paraphasic substitutions, especially on low
probability phrases. On reading tests he was unable to
comprehend passages of more than two sentences, and oral
reading was slow, effortful, and contaminated by paralexic
errors. He could write short sentences to dictation but
made spelling errors on longer words. Re-examination
with the BDAE four weeks after onset showed improvement in all areas, but with continued paraphasic errors in
spontaneous speech, repetition, and oral reading. His most
significant disability remained that of alexia, with inability
to comprehend paragraph-length material.
The Token Test7 was performed four weeks after onset,
both in the standard, auditory form and with the instructions presented in printed form. He correctly performed
94% of the auditory and 82% of the visual instructions (by
weighted scores).
A repeat CT scan (fig 1), performed 6 days after onset,
showed an area of low density involving predominantly the
left angular gyrus, with extension into the supramarginal
gyrus, posterior-most superior temporal gyrus, and

Kirshner, Webb
inferior-most superior parietal lobule, along with subjacent
white matter.8 The patient continued to improve in speech
and language, with mild paraphasia of spontaneous speech
and continued reading difficulty when last seen three
months after onset.
Case 2 This 69-year-old hypertensive, diabetic righthanded lady presented with several episodes of left monocular blurred vision and one of transient left hemiparesis
associated with "slurred" speech. Examination revealed
right carotid and left supraclavicular bruits. No fixed
neurologic deficits were found. CT scan of the head was
normal. Cerebral arteriography revealed tight stenosis of
the right internal carotid artery, severe plaque formation
and 50-60% stenosis of the left internal carotid artery, and
stenosis of the proximal left vertebral and subclavian
arteries. Following the arteriogram the patient was noted
to be "confused"; on examination, her speech was rambling, with numerous spontaneous paraphasic errors and
neologisms. She did not respond to commands or yes/no
questions. She showed no apparent appreciation of visual
stimuli and failed to count fingers or blink to visual threat.
Spontaneous movements and withdrawal to pinprick were
strong and symmetric in all extremities. Reflexes were
symmetic, with flexor plantar responses. On the following
day she remained confused but reached out to grasp
objects with either hand, more consistently in the left field
of vision. She continued to speak paraphasically but expressed some appropriate phrases, including a complaint of
headache. She failed to identify objects or colours. By the
fourth day she looked about the room, recognised family
members, and spoke in full sentences, though with many
word substitutions. Over the next few days she regained
the ability to name some objects but not uncommon
objects or parts of objects. Her auditory comprehension

Fig 1 Computed tomographic
(CT scan), Case 1, showing an area
of reduced attenuation, consistent
with infarction in the region of the
left angular gyrus.

721

Alexia and agraphia in Wernicke's aphasia

-;
i4
F

and following of commands recovered substantially, except
for left-right errors. In contrast, she remained almost
entirely unable to read or write. She could not read more
than single words aloud, follow the simplest printed commands, or sign her name. She was unable to perform even
simple calculations, such as "7 + 5", to copy or construct
geometrical figures, or to identify left and right. She was
able to point to specific fingers on her own and the
examiner's hand.
Speech and language evaluations were carried out 2-3
weeks after onset. She named 16/20 capital letters, 17/20
pictures, and read aloud 19/20 picture names. In contrast,
she could not write the names of any of ten pictures. On
the matching-to-sample battery, she correctly matched
19/20 dictated and 8/10 printed small letters to the correct
capital letters. She also matched 20/20 printed, 20/20 dictated, and 19/20 dictated, spelled words to pictures. Given
a series of word lists differing in length and frequency of
occurrence, she read aloud the frequent words accurately,
whether short or long, but she made frequent errors on the
unfamiliar words and was totally unable to pronounce nonsense trigrams. At the time of retesting six weeks after
onset, the patient named 20/20 pictures orally and 10/20 in
writing. On the BDAE, spontaneous speech was fluent,
with frequent literal paraphasic errors. The patient exhibited moderate dysnomia for objects, actions, colours,
numbers, and body parts. Auditory comprehension was
impaired only for multistep commands and complex ideational material. Single word and sentence repetition
showed literal paraphasic errors. Oral reading was severely
impaired, with marked hesitancy and both literal and verbal paraphasic errors. She failed to read sentences or paragraphs for meaning and made occasional errors even in
single word recognition. Writing ability was limited to the
copying of single letters and numbers; she failed to write

_

Fig 2 CT scan, Case 2, showing
infarction in the vicinity of the left
angular gyrus. The small area of
increased density represents
haemorrhagic infarction.

letters, numbers, or words to dictation. Retesting six weeks
after onset revealed almost normal auditory comprehension and mild, persistent dysnomia and paraphasic speech.
Reading and writing, though improving, remained severely
deficient. The Token Test, performed three weeks after
onset, showed 90% correct for auditory and 39% correct
for visual instructions.
A second Cf scan, 18 days after onset, revealed a low
density area involving the left angular gyrus and subjacent
white matter, with slight extension into the supramarginal
gyrus and superior parietal lobule. The entire temporal
lobe appeared spared." The patient's subsequent course
was one of gradual improvement, with mild paraphasic
speech and word-finding difficulty, but persistent and frustrating impairments in reading and writing.
Case 3 This 60-year-old right-handed male was previously in good health except for hypertension, amputation
of the right leg, and a progressive left ear hearing loss.
Audiometry disclosed a moderate conductive hearing loss
on the left, with only high frequency loss on the right.
While hospitalised for evaluation of the hearing loss he
suddenly developed a right hemiparesis and aphasia, with
initial descriptions of his speech as "slurred" and " nonsensical". By day 5 he manifested a clearcut fluent aphasia,
while the hemiparesis had improved to mild weakness of
the right face and arm. Visual fields were intact. Bedside
aphasia testing six days after onset revealed fluent speech
with word-searching pauses, circumlocutions, and frequent
literal and verbal paraphasic errors. He named 8/8 objects,
3/3 parts of objects, and 10/10 pictures. He repeated simple and complex phrases with few errors. Despite the hearing loss he followed 8/8 spoken commands of increasing
complexity and answered appropriately to yes/no and nonsense questions. In contrast, he was able to carry out or
read aloud only 5/8 printed commands. Writing was

722
limited to his signature and single, dictated words; he could
not write even short sentences to dictation. Matching was
nearly errorless for the single letter, word, and picture
items.
On the BDAE, performed on days 5-7, spontaneous
speech was characterised by fluent phraseology with many
cliches, few meaningful nouns, paraphasic errors, and some
neologistic jargon. Naming was mildly impaired, and repetition was accurate but contaminated by paraphasic errors.
Auditory comprehension was impaired only on four and
five part commands and complex ideational material. In
contrast, reading comprehension was severely impaired for
more than single word items. He correctly answered only
20% of questions on sentences and paragraphs. Oral reading of sentences contained many paraphasic errors. Writing
was poor even for dictated single words and copying of
words and sentences. Retesting five weeks after onset
showed improvement in spontaneous speech and auditory
comprehension, with persistent paraphasic errors and
word-finding difficulty. Reading of sentences, both aloud
and for comprehension, remained impaired. The Token
Test was performed 16 days after onset, by which time
comprehension had improved in both modalities. The
patient correctly carried out 93% of the auditory and 88%
of the visual instructions.
A cT scan, performed within five hours of onset, was
normal. A repeat study was not obtained. The clinical
course was one of substantial improvement in all aspects of
language. Only mildly paraphasic speech and reading
difficulty remained as clinically evident deficits.

Discussion

Since its original description in 1874,9 Wernicke's
aphasia has been considered a disorder of central
language processes, independent of specific language modalities. Disruption of the auditory word
association area in the superior temporal gyrus
(Wernicke's area) interferes with auditory processing. Reading, by the traditional formulation,39 is
learned by the transcoding of visual language stimuli
to previously learned auditory word associations;
alexia would thus also be expected with lesions in
this region.
In a previous communication,4 we discussed cases
of Wernicke's aphasia with poor auditory comprehension but relatively spared reading. This disorder resembles the syndrome of pure word deafness, except for the paraphasic speech, repetition,
and naming. The preserved reading ability in these
cases appears to reside in the left parietal and occipital lobes,4 10 or possibly even in the right hemis-

phere."I

The three cases discussed here suggest a second
modality-specific deficit pattern in Wernicke's
aphasia, that of greater impairment of reading than
auditory comprehension. All three patients had
completed at least ten years of school and read

Kirshner, Webb
fluently prior to their strokes. The greater involvement of reading than of auditory comprehension
persisted over time and was apparent in bedside testing, on the BDAE, and on the auditory and visual
forms of the Token Test. Cases 2 and 3 showed
significant agraphia as well as alexia, suggesting a
primary diagnosis of pure alexia with agraphia. AlR
three cases, however, had persistent paraphasic
speech, repetition, naming, and oral reading, linking
them more closely to Wernicke's aphasia.
The anatomic substrate of this syndrome, as
judged by the CT scans of Cases 1 and 2, involved
the left inferior parietal region, especially the angular gyrus and subjacent white matter. In Case 1 the
posterior-most part of the superior temporal gyrus
also appeared involved. In Case 2 the aphasia
remained as a persistent focal deficit after the resolution of a more diffuse postarteriography syndrome
of confusion and cortical blindness. Cortical blindness is a recognised complication of cerebral
arteriography and may represent a direct effect of
the contrast material.'2 The focal lesion seen on CT
scan, however, suggests embolisation to posterior
branches of the left middle cerebral artery, as does
the sudden onset of the language disorder in all
three cases.
While Wernicke himself assigned the auditory
word association area to the posterior part of the
superior temporal gyrus,9 subsequent aphasiologists
have differed in its localisation. The inferior parietal
lobule, or supramarginal and angular gyri, has been
included by some authors.'3 Paraphasic speech and
impaired auditory comprehension are thus not
unexpected with lesions in this vicinity. The greater
involvement of reading than of auditory comprehension, however, cannot be explained by involvement
of the auditory association area alone, but rather
must reflect the importance of the angular gyrus region to reading.
Dejerine,'4 in 1892, described the syndrome of
pure alexia with agraphia in a patient with an infarct
of the left angular gyrus. He believed the angular
gyrus to be a centre for visual word images, analogous to the superior temporal gyrus as a centre for
auditory word images. Wernicke,9 and more
recently Geschwind,'5 emphasised the role of the
angular gyrus in cross-modal sensory associations.
Instead of removing a cortical word-image centre, a
lesion of the angular gyrus would disrupt reading by
disconnecting visual to auditory transcoding. By
either explanation, an angular gyrus lesion would
produce alexia without affecting either auditory
comprehension or speech production. Dejerinel4
stated that alexia and agraphia could either accompany a Wernicke's aphasia or exist in pure form.
Some recent authors2 16 17 have attempted to distin-

Alexia and agraphia in Wernicke's aphasia
guish alexia with agraphia from "aphasic alexia", or
the alexia associated with Wernicke's aphasia. On
the other hand, both Kertesz ' and Brain'8 noted
that alexia with agraphia can remain as a residual
deficit after recovery from Wernicke's aphasia. Benson'9 stated that only in rare instances does alexia
with agraphia occur without any signs of aphasia.
Modem classifications of alexia have emphasised
three principal groups: alexia with agraphia, alexia
without agraphia, and alexia associated with
aphasia.3 Alexia with agraphia has already been discussed. Alexia without agraphia is an easily separable syndrome seen with lesions of the left occipital
lobe, often including the splenium of the corpus callosum. Dejerinel4 explained this syndrome by a disconnection of visual information perceived by the
intact right occipital lobe from the left angular gyrus.
Aphasic alexia has included different groups of
aphasic cases. Benson'9 has called attention to alexia
in Broca's aphasia, under the term "third alexia".
The nonfluent speech of these patients, of course,
should distinguish them readily from the other alexic
groups. Finally, the alexia of Wemicke's aphasia
may be considered as a separate, fourth group, or it
may be combined with the first syndrome, alexia
with agraphia. Benson'9 has suggested the terms
"central alexia" for the alexia with agraphia cases,
"posterior alexia" for alexia without agraphia, and
"anterior alexia" for the reading disorder seen with
Broca's aphasia.
An alternative explanation for the findings of
fluent aphasia and alexia in our patients would be a
combination of conduction aphasia and alexia with
agraphia. Conduction aphasia, a syndrome in which
paraphasic speech and impaired repetition occur in
the presence of intact auditory comprehension, is
associated with lesions of either the left supramarginal gyrus or the left superior temporal gyrus.20 Factors against the diagnosis of conduction aphasia in
our patients include the presence at least initially of
auditory comprehension deficit, the absence of a
severe repetition disturbance out of proportion to
other language deficits, and the absence of disproportionate literal as compared to verbal paraphasic
errors. For these reasons we prefer the diagnosis of
Wemicke's aphasia with alexia. The most important
point, however, is the linkage of alexia with agraphia
to the fluent aphasias.
Numerous studies have attempted to distinguish
alexic syndromes by their relative impairments of
word versus letter reading. In alexia without
agraphia, preservation of individual letter reading
has been amply documented.2'-24 The alexia of
Broca's aphasia, on the other hand, produces a
major deficit in the naming and reading of single
letters, even when some words can be read in a

723
global manner.'92526 In Wernicke's aphasia, single
letters may be processed well, as seen in our cases,
and difficulty increases with increasing complexity of
words and phrases.'6 In alexia with agraphia, some
authors2 27 have found letter greater than word
alexia, much like that seen in Broca's aphasia. Other
authors,'6 however, have not found letter alexia in
alexia with agraphia, and indeed occasional cases of
alexia with agraphia have been reported with intact
oral spelling and comprehension of dictated, spelled
words.2830 Performance in letter versus word reading would thus not appear to be a reliable factor in
distinguishing alexia with agraphia from alexia
related to Wemicke's aphasia, though certainly it
does distinguish alexia without agraphia, or "posterior alexia", from the alexia associated with frontal
lesions. Inability to perceive a word as a whole is the
most consistent deficit in "posterior" alexia, while
single letter reading and phonetic processing of
graphemes may be intact. In both the "anterior" and
"central" alexias, on the other hand, whole word
perception may be the only preserved function,
while letter reading and phonetic processing are
impaired.'9 25 26
Cases of Wemicke's aphasia with alexia out of
proportion to auditory comprehension deficit have
not been generally recognised, though as discussed
above some cases of alexia with agraphia may
belong to this category. H6caen'6 described a subset
of Wemicke's aphasics with auditory predominant
deficits, but he related this to a deficit in the perception of phonemes. His second subgroup, with
deficient semantic comprehension, was said to have
equal impairment of auditory comprehension and
reading. Mohr, Hier, and Kirshner,'0 in a retrospective review of CT scan lesions, found two cases of
Wemicke's aphasia with greater impairment of reading than of auditory comprehension. Both had
parieto-occipital lesions. This compared to six cases
of auditory-predominant deficit and 23 in which no
disproportion could be judged. The present three
cases are part of an ongoing investigation of selective involvement of specific language modalities in
aphasia secondary to ischaemic stroke. Of 18 cases
of Wernicke's aphasia studied, two showed the
auditory and three the visual-predominant syndrome. Selective impairment of either auditory
comprehension or reading thus appears to be not
uncommon in Wernicke's aphasia.
The independence of deficits in reading and
auditory comprehension in Wernicke's aphasia suggests at least partial independence of the brain
mechanisms for auditory and visual language processing. This independence belies the simple concept that reading can proceed only via transcoding
of visual to phonetic information. The evidence

Kirshner, Webb

724

from Wernicke' s aphasia thus confirms evidence
from other sources31 32 that reading can occur from
purely visual to semantic processing, without access
to auditory or phonetic information.
This research was supported by NINCDS Teacher
No.
Award
Development
Investigator
5K07NS00429-03 to Dr Kirshner and also by the
Veterans Administration. The authors thank Mrs
Anne Morgan for technical assistance and Mrs Jane
S Smith for assistance with preparation of the
manuscript.

References
lBenson DF, Geschwind N. The aphasias and related disturbances. In: Baker AB, Baker LH eds; Clinical
Neurology. Hagerstown, Harper and Row Publishers,
1977, Vol 1, Chapter 8, 3-9.
2 Brown J. Aphasia, Apraxia and Agnosia. Clinical and
Theoretical Aspects. Springfield: Charles C. Thomas,
1972, 56-71.
3Benson DF. Aphasia, Alexia, and Agraphia. New York:
Churchill Livingstone, 1979, 117.
4 Kirshner HS, Webb WG, Duncan GW. Word Deafness
in Wernicke's Aphasia. J Neurol Neurosurg
Psychiatry, 1981 ;44:197-201.
5 Sidman M, Stoddard LY, Mohr JP et al. Behavioral
studies of aphasia. Methods of investigation and
analysis. Neuropsychologia 1971;9: 119-40.
6 Goodglass H, Kaplan E. The Assessment ofAphasia and
Related Disturbances. Philadelphia: Lea and Febiger,
1972.
Spreen 0, Benton A. Token Test. From the Neurosensory Center Comprehensive Examination for
Aphasia, Neuropsychiatry Laboratory, University of
Victoria, 1969.
Gado M, Hanaway J, Frank R. Functional anatomy of
the cerebral cortex by computed tomography. J Comput Assist Tomogr 1979;3:1-19.
9 Wernicke C. The symptom complex of aphasia. A
psychological study on an anatomical basis. Proc Boston Colloquium for Philosophy of Science
1968;4:34-97 (originally 1874).
10 Mohr JP, Hier D, Kirshner HS. Modality bias in Wernicke's aphasia. Neurology (Minneap) 1978;28:395

(Abst).

Heilman KM, Rothi L, Campanella D et al. Wernicke's
and global aphasia without alexia. Arch Neurol
1979;36: 129-33.
12 Stoddard WE, Davis DO, Young SW. Cortical blindness
after cerebral angiography. Case report, J Neurosurg

1981;54:240-4.

13 Bogen JE, Bogen GM. Wernicke's region-Where is it?

Ann NY Acad Sci 1976;280:834-843.
MJ. Contribution a l6tude anatomopathologique et clinique des differenctes varietes de
c6cit6 verbale. CR Soc Biol (Paris) 1892;4:61-90.
.s Geschwind N. Disconnexion syndromes in animals and
man. Brain 1965;88:237-94, 585-644.
16 H6caen H, Kremin H. Neurolinguistic research on reading disorders resulting from left hemisphere lesions:
aphasic and "pure" alexias. In: Whitaker H, Whitaker
H eds; Studies in Neurolinguistics, Vol 2, New York*,
Academic Press, 1977.
17Kertesz A. Aphasia and Associated Disorders. Taxonomy, Localization, and Recovery, New York,
Grune & Stratton, 1979.
18 Brain WR. Speech Disorders. London; Butterworths,
1962.
19 Benson DF. The third alexia. Arch Neurol 1977;

14 Dejerine

34:327-31.
20 Benson DF, Sheramata WA, Bouchard R, et al. Conduc-

tion aphasia. A clinicopathological study. Arch Neurol
1973;28:339-46.
21 Warrington EK, Rabin P. Visual span of apprehension in
patients with unilateral cerebral lesions. Q J Exp
Psychol 1971;23:423-31.
22 Staller J, Buchanan D, Singer M, et al. Alexia without
agraphia: An experimental case study. Brain Lang

1978;5:378-87.
23 Levine DN, Calvanio R. A Study of the visual defect in

verbal alexia-simultanagnosia. Brain 1978;101:6581.
24 Kirshner HS, Staller J, Webb WG, Sachs P. Transtentorial herniation with posterior cerebral artery territory
infarction: a new mechanism of the syndrome of alexia
without agraphia. Stroke, 1982;13:243-6.
2S Benson DF, Brown J, Tomlinson EB. Varieties of alexia:
Word and letter blindness. Neurology (Minneap)
1971 ;21:951-7.
26 Kirshner HS, Webb WG. Word and letter reading and
the mechanism of the third alexia. Arch Neurol

1982;39:84-7.
27Alajouanine T, Lhermitte F, Ribaucourt-Ducarne B.
Les alexies agnostiques et aphasiques. In: Alajouanine T ed, Les grandes activit6s du lobe occipital.
Paris, Masson, 1960.
28 Albert ML, Yamadori A, Gardner H et al. Comprehension in alexia. Brain 1973;96:317-28.
29 Mohr JP. An unusual case of dyslexia without dysgraphia. Brain Lang 1976;3:324-34.
30 Rothi LJ, Heilman KM. Alexia and agraphia with spared
spelling and letter recognition abilities. Brain Lang
1981;12:1-13.
31 Kapur N, Perl NT. Recognition reading in paralexia.
Cortex 1978;14:439-43.
32 Saffran EM, Marin OSM. Reading without phonology:
evidence from aphasia. Q J Exp Psychol
1977;29:515-25.