Brain (1991), 114, 1977-1996

PHONOLEXICAL AGRAPHIA
SUPERIMPOSITION OF ACQUIRED LEXICAL AGRAPHIA ON
DEVELOPMENTAL PHONOLOGICAL DYSGRAPHIA
by ANDREW KIRK,1 LEE X. BLONDER,1 ELIAHU WERTMAN2 and
KENNETH M. HEILMAN1
(From the 'University of Florida College of Medicine, Department of Neurology, and the Veterans
Affairs Medical Center, Gainesville, FL, USA, and 2Ezrath Nashim Hospital, Hebrew University
Medical School, Jerusalem 91001, Israel)

Study of neuropsychological sequelae of a focal acquired brain lesion may bring out and help delineate
the features of a compensated developmental language disorder and its anatomical substrate. A left-handed
man with a history of phonological developmental dyslexia and dysgraphia learned in early adulthood to
read and write using a lexical system. Following a small posterior right parietal infarct when aged 56 yrs
he developed a severe agraphia displaying features of phonological dysgraphia with impaired segmentation
and features of lexical agraphia. Writing was severely impaired for all classes of word and nonword stimuli
but his errors did not resemble those attributable to a deficit in the system responsible for the short-term
storage of the graphemic representation of a word (graphemic output buffer). These observations imply
that an acquired lexical agraphia has been superimposed on his developmental phonological dysgraphia,
resulting in a combined or 'phonolexical' agraphia.
INTRODUCTION

Acquired agraphias can be broadly classified as either linguistic or motor agraphias
(Roeltgen and Heilman, 1984). Linguistic agraphias have been further subdivided, on
the basis of patterns of impairment, into at least two major types. Patients with
phonological agraphia have difficulty spelling nonwords but have a relatively preserved
ability to write phonetically irregular words. Errors tend not to be phonologically correct
(Shallice, 1981). Part of speech and imageability of the target word are important
determinants of spelling accuracy in this disorder. In particular, functors are written
less accurately than are content words and performance is better for high than for low
imagery nouns (Assal et al., 1981; Bub and Kertesz, 1982; Roeltgen et al., 1983ft).
Writing includes semantic or derivational errors, neologisms, mistakes with visual
similarity to the intended word, and incorrect real words unrelated to the stimulus
(Shallice, 1981; Bub and Kertesz, 1982; Roeltgen et al., 1983ft; Bolla-Wilson et al.,
1985). High frequency words are written more accurately than less frequent words
(Shallice, 1981), and short words are more accurately written than long words (Shallice,
1981; Roeltgen et al., 1983ft; Temple, 1988).
Correspondence to: Dr K. M. Heilman, J. Hillis Miller Health Center, Box J-236, University of Florida, Gainesville,
FL 32610, USA.
Present address: Dr Andrew Kirk, Division of Neurology, Department of Medicine, College of Medicine, University
of Saskatchewan, Saskatchewan, Canada S7N 0X0; Dr Lee Blonder, Center on Aging, 101 Sanders-Brown Building,
University of Kentucky, Lexington, KY 40536-0230, USA.
© Oxford University Press 1991

Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016

SUMMARY

1978

A. KIRK AND OTHERS

Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016

In contrast to phonologically agraphic patients, those with lexical agraphia spell irregular
words poorly but have a relatively spared ability to write regular words and pronounceable
nonwords. Errors tend to be phonologically correct (e.g., NEPHEW — NEFFUE;
Patterson, 1982). Imageability and grammatical class do no necessarily affect writing
accuracy, and functors may thus be written as well as content words (Beauvois and
DeYouesne", 1981; Roeltgen and Heilman, 1984; Gonzalez Rothi etai, 1987; Rapcsak
et al., 1988). These patients may also be impaired at spelling the correct homophone,
producing, for example, PALE instead of PAIL (Hatfield and Patterson, 1983; Roeltgen
and Heilman, 1984; Goodman and Caramazza, 1986).
These patterns of agraphia have contributed to the development of models of writing
that include at least two routes that can be used to spell (e.g., Beauvois and DeYouesne",
1981; Nolan and Caramazza, 1983; Roeltgen and Heilman, 1984). Spellings of familiar
words may be retrieved from a lexicon, whereas spellings for unfamiliar words or
nonwords must be arrived at via a phonological route as no lexical entries exist for
them. Irregular words could thus only be spelled using the lexical route and nonwords
only phonologically or perhaps, in some cases, by analogy with real words (Shallice,
1981).
Ellis (1982) has suggested that phonological transcoding entails two functions:
segmentation of words into phonemes and then conversion of phonemes into graphemes.
Roeltgen et al. (1983ft) found support for this model in their analysis of errors made
by subjects with phonological agraphia.
Models of writing often include a graphemic output buffer responsible for short-term
storage of a graphemic representation of a word. This abstract representation, which
may have been generated via phonological or lexical means, can then be used to produce
letter names or shapes for oral or written spelling (Morton, 1980; Newcombe and
Marshall, 1980; Ellis, 1982). Several cases have been documented in which agraphia
has been ascribed to damage to this graphemic buffer (Miceli et al., 1985; Caramazza
et al., 1987; Posteraro et al., 1988; Hillis and Caramazza, 1989). Graphemic buffer
deficits should affect spelling in specific ways (Caramazza et al., 1987; Hillis and
Caramazza, 1989). Word and nonword spelling should be impaired essentially equally.
Impairment should be independent of modality of input or output. Long words should
be spelt more poorly than short words. Part of speech, word frequency, and imageability
should not affect spelling accuracy. Although some of these criteria can apply individually
to phonological or lexical agraphia, a different type of error is expected of a graphemic
buffer deficit. Errors should not resemble the intended word semantically, phonologically,
or morphologically. Instead, errors should consist of graphemic substitutions,
transpositions, insertions and deletions. These erroneous responses often violate
orthographic rules of the language and are thus unpronounceable.
Lesions causing phonological agraphia have been localized to the left supramarginal
gyrus or underlying insula (Roeltgen et al., 1983ft) whereas lexical agraphia has been
attributed to lesions of the left posterior angular gyrus and parieto-occipital lobule
(Roeltgen and Heilman, 1984). A single case of lexical agraphia with a left precentral
gyrus lesion (Rapcsak et al., 1988) has also been reported. The two reports describing
linguistic agraphia resulting from right hemisphere damage have documented lesions
in the corresponding retrorolandic sites on the right. Thus Bolla-Wilson et al. (1985)
reported a case of phonological agraphia in a left-handed patient with a right supramarginal

PHONOLEXICAL AGRAPHIA

1979

CASE REPORT
O.C., a left-handed man, aged 56 yrs, was examined in March, 1989, 3 mos after a stroke impaired
his reading and writing. As a child he had had great difficulty learning to read and write although his
speech development had been normal. His older brother was also left handed but learned to read and write
without difficulty. His 3 sisters and his parents were right handed and had no literacy problems. O.C.
reported that he never could 'see the sense' in sounding out words and never understood how it could
be possible to 'break up words'. O.C. left school at age 13 yrs with an eighth grade education. He recalled
that at that time he could read and write only the letters of the alphabet and his name. At 18 he joined
the Navy (he recalled that the 'written' entrance examination was read to him by a recruiter) where his
captain discovered his illiteracy and assigned 3 chaplains to teach him to read and write. A trial of teaching
him to sound out words was unsuccessful. The clergymen subsequently gave him lists of words each day
to copy and memorize by rote. By the time he left the service 3 yrs later he was literate enough to work
as a machinist and rose to become a state union president by his early thirties. In this capacity he wrote
reports and read speeches regularly. He later worked with computers in an airplane engine factory monitoring
the flow of parts and finished engines into and out of the factory. As an adult, O.C. described himself
as an avid and rapid reader. He particularly enjoyed books on American labour and politics.
His medical history was unremarkable until December 1988, when he underwent cardiac catheterization
for investigation of recent chest pain. Coronary angiography was normal but shortly after the procedure
he experienced the sudden onset of left hand numbness and slight weakness. When asked to sign a form
he found that he was unable to write intelligibly with either hand. He had begun reading a lengthy political
tome just before his catheterization but was never able thereafter to progress beyond the first chapter despite
repeated attempts.
Although his hand numbness and weakness disappeared completely over the next few weeks, he was
disabled by his inability to read or write and was unable to return to his former employment. CT scan
on the day of the stroke was normal but on the following day showed a small right parietal lucency. MRI
scan performed 4 mos later confirmed a small right posterior parietal lobe lesion but was otherwise normal
(see fig. 1).
When first seen at the University of Florida Teaching Hospital 3 mos after his stroke, cranial nerve,
motor, sensory, and reflex examinations were normal except for minor awkwardness of fine finger

Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016

gyrus lesion, and Gonzalez Rothi etal. (1987) found a right posterior parieto-occipital
lesion in a right hander with lexical agraphia.
Children with developmental dysgraphia have also been classified on the basis of the
phonological accuracy of their misspellings. Boder (1973) used the terms dyseidetic
and dysphonetic to describe children making phonologically correct and incorrect errors,
respectively. Frith (1980) described two groups of poor spellers who varied in their
ability to write nonwords. Temple (1986) has described children with deficits akin to
phonological and lexical agraphia. Roeltgen and Tucker (1988) have divided adults with
developmental dysgraphia into phonological and lexical groups with profiles of spelling
ability virtually indistinguishable from those of patients with acquired agraphias. Case
reports have also documented the persistence of phonological dysgraphia into adulthood
(Temple, 1988; Funnell and Davison, 1989). Campbell and Butterworth (1985) have
described a patient with developmental phonological dyslexia and dysgraphia who
overcame her disability and became highly literate using lexical strategies.
We describe a unique subject with developmental dysgraphia and dyslexia, seemingly
due to phonological impairment, who learned to read and write well in adulthood using
a lexical strategy and then, after a right parietal infarct, developed a superimposed
acquired agraphia. His new deficit does not resemble phonological agraphia and we
argue that the best explanation for his current deficit is an acquired lexical agraphia
combined with his pre-existing phonological impairment.

1980

A. KIRK AND O T H E R S

movements on the left. He was oriented for time, place, and person. He remembered 3 of the past 5 presidents
and 2 of 3 words after 3 min distraction. Digit span was five forward. He stated that an apple and a banana
were both to eat and that a chair and a table were both wooden and had legs. He performed calculations
well (24/24 on the Western Aphasia Battery (WAB) (Kertesz, 1982) calculation subtest). There was no
right-left confusion. He was not apraxic with either hand (59/60 on the WAB praxis subtest). There was
no neglect on cancellation or line bisection tasks. Drawing was unimpaired.
O.C. was not aphasic, scoring a WAB aphasia quotient of 97.7, well within the normal range. In contrast
to his normal verbal communication, he was severely agraphic, unable to write even his own name without
a model from which to copy. WAB reading and writing scores were 88/100 and 52.5/100, respectively.
An investigation of his reading and writing was thus undertaken.

RESULTS

Writing
O.C. made no errors in copying 50 written stimuli comprising letters, nonwords,
and regular and irregular words. He could transcribe upper to lower case flawlessly,
print to script, and vice versa. Letters were always well-formed. He preferred to print
in block capitals but could produce script upon request. Samples of his handwriting
revealed that he had usually written in block letters even before his stroke. O.C. always
began at the left of a word and wrote it in the normal sequence, using his left hand.
When asked to use his right, spelling was similarly affected but letters were not so neatly
formed.
O.C. made no errors when asked to repeat aloud orally presented letters, phonemes,
words, or nonwords. He could correctly write 26/26 single dictated letters and could
write letters for 23/24 dictated single phonemes.
O.C. was asked to write to dictation 30 nonwords, 30 regular, and 30 irregular words
(see Appendix) taken from the Battery of Adult Reading Functions (BARF) (L. J. G.
Rothi, H. B. Coslett and K. M. Heilman, unpublished manuscript; Gonzalez Rothi et al.,
1987). Nonwords were constructed so as to be phonologically plausible and probable
(Venezky, 1979; Kay and Marcel, 1981). The three groups of stimuli were balanced
for length (mean 5.3 letters in each case) and the two groups of real words were balanced

Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016

FIG. 1. MRI scan 4 mos after O.C.'s stroke, showing small posterior right parietal lesion also seen on second, but
not first, CT scan. Centre and right are T 2 weighted axial cuts taken at 3.79 and 2.88 cm above horizontal with centimetre
markings seen in T, weighted sagittal scout film at left.

PHONOLEXICAL AGRAPHIA

1981

TABLE I. O.C.'S PERFORMANCE AT WRITING AND SPELLING
ALOUD TO DICTATION
Stimulus category
Writing to dictation
Letters
Phonemes
Nonwords
Regular words
Irregular words
Functors
Contentives
Spelling aloud to dictation
Nonwords
Regular words
Irregular words

No. of
stimuli

% correct

% omission

% erroi

26
24
30
30
30
30
30

100.0
95.8
10.0
3.3
0
0
23.3

0
0
0
0
0
0
0

0
4.2
90.0
96.7
100.0
100.0
76.7

30
30
30

3.3
10.0
3.3

0
3.3
13.3

96.7
86.7
83.4

Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016

for frequency of usage (43.3/million for regular words, 43.6/million for irregular) (Kucera
and Francis, 1967). Nouns, verbs, and adjectives were equally represented on the two
lists. To examine the effect of grammatical class of stimuli on spelling accuracy, O.C.
was also asked to write to dictation 30 functors (mainly prepositions and adverbs) and
30 contentives (nouns, adjectives, and verbs) matched for length (mean 5.3 letters for
each list) and frequency of usage (403.3/million for functors, 402.8/million for
contentives). Note that, since functors tend to be very frequent words, these two lists
are of higher frequency than the regular and irregular words. To ensure that any
impairment was not due to a simple memory disturbance causing him to forget the stimuli,
O.C. was required to repeat each stimulus both before and after attempting to write
it. This he was able to do without fail. He could define the words presented. When
asked immediately, he was always able to tell whether he had written a word correctly
or not but was never able to make any attempt at correcting his mistakes. As shown
in Table 1, O.C. 's performance on writing to dictation was severely impaired in writing
nonwords and regular and irregular words, although he was better at writing contentives
than functors.
To determine whether modality of output influenced his spelling accuracy, he was
asked to spell aloud 30 nonwords and 30 regular and 30 irregular words taken from
the Battery of Linguistic Analysis for Writing and Reading (BLAWR) (Roeltgen et al.,
1983a; Roeltgen and Heilman, 1984). These stimuli were comparable with those used
for writing to dictation and are also listed in the Appendix. Mean stimulus length was
5.3 letters for all three lists. Mean frequency of usage was 43.8/million for regular
and 43.7/miIlion for irregular words. O.C.'s performance on spelling dictated words
aloud is shown in Table 1. It is apparent that spelling aloud and writing to dictation
are similarly affected. He is severely impaired at writing nonwords and regular and
irregular words. O.C. was also asked to spell aloud 10 high and 10 low imagery nouns
(mean imageability ratings 6.46 and 2.98; Paivio et al., 1968) balanced for regularity,
word length (mean 4.4 letters for each list), and frequency (mean 151.4 and 151.5/million,
respectively). He correctly spelled six high and six low imagery nouns. Thus target
word imageability does not appear to play a major role in spelling accuracy. An attempt
was also made to test O.C. 's ability to assemble a dictated word using its letters printed

1982

A. KIRK AND OTHERS

Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016

on cardboard squares and randomly mixed. This was abandoned after five stimuli as
O.C. found it extremely frustrating and was unable to produce any response.
To assess whether modality of stimulus input affected O.C.'s writing, he was asked
to name, in writing, 15 common objects shown to him. Seven were written correctly.
Later, when the same words were dictated, he performed similarly, writing six correctly.
Before examining O.C.'s error types in detail, it is important to establish, as far as
possible, the nature and extent of the developmental dysgraphia from which he suffered
before his stroke. Given his history, there is little doubt that his writing was severely
impaired by his stroke. Support for this is found in a song he composed in 1986. The
original manuscript contained one misspelling (SOON — SOONE) and one mispunctuation (LETTERS - LETTER'S). By contrast, when he was asked, after his stroke,
to write his song from memory he made 18 errors in 50 words. Direct comparison of
this error rate with that for the stimuli discussed above would be misleading as the song
contained many short high frequency words. When provided with a correctly written
copy of his song, O.C. was able to flawlessly copy it, apart from leaving out one word.
O.C.'s history suggested that he had suffered from a developmental phonological
dysgraphia. He had in his possession several postcards that he had written to his family
during his stint in the Navy while he was learning to read and write. These contained
147 words, of which 82% were correctly written. All letters were well-formed. All
words spelled wrongly were always written in the same wrong way—there were no
examples of words being written correctly in one place and incorrectly in another or
of a word being misspelled in two different ways. Errors were often phonologically
incorrect. He correctly spelled 88% of regular words and 76% of irregular words.
Writing samples from recent years before his stroke showed considerably fewer errors.
Perusing notebooks of several hundred words, originally intended only for his own
reference, we uncovered five words spelled incorrectly (ABSENTEE — ABESENTEE;
THIRTY - THIRITY; THIRTY - THEREE; SEARS - SEAR; ONE - OME:
HUNDRED — HUNDER). ABSENTEE appeared only once and was spelled incorrectly
as shown. THIRTY was never spelled correctly on the four occasions it appeared. The
other 3 words appeared more than once and were spelled both correctly and incorrectly.
Given the likelihood that O.C.'s developmental dysgraphia was phonological, we
decided to examine his phonological abilities. Ellis (1982) proposed that writing a word
via a phonological route includes two separable processes, segmentation of the word
into phonemes and then conversion of those phonemes into graphemes. As O.C.'s
conversion of single letters to phonemes and vice versa was well-preserved even after
his stroke, we tested his ability to segment sounds using the Lindamood Auditory
Conceptualization (LAC) Test (Lindamood and Lindamood, 1971) in which coloured
tokens chosen by the subject are made to stand for single phonemes. In LAC Category
I-A, the subject is asked to use a token of a given colour to represent each different
phoneme uniquely in a string of separately enunciated sounds dictated to him. For
example, the examiner dictates, 'Show me s/s' or 'Show me g/b/v' where the appropriate
phoneme is pronounced rather than the letter name. Correct responses to these 2 examples
would be 2 tokens of the same colour for the former and 3 tokens of different colours
for the latter. O.C. represented 8/10 strings correctly with tokens. He quickly corrected
his 2 incorrect responses after the stimuli were repeated. On the similar but more
demanding LAC Category I-B, O.C. correctly formed 4/6 token strings. Again, when

PHONOLEXICAL AGRAPHIA

1983

Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016

the stimuli were repeated, he immediately corrected his 2 incorrect constructions saying,
'I didn't quite catch that the first time.' In contrast to his relatively spared performance
on Category I, essentially a test of phoneme-symbol conversion, O.C. performed poorly
on Category II where the subject is asked to represent nonwords with strings of coloured
tokens, a task requiring segmentation ability. For example, when shown a string of
4 tokens of different colours and told, 'If that says "vops", show me "vaps" ', he
is expected to replace the second token with one of a new colour. In this category, O.C.
correctly assembled only 1/13 responses. He was clearly mystified by the requests and
aimlessly moved tokens about. He commented, 'I'm not getting the breakdown of the
word. I understand in theory what it is you want me to do but I know that's something
I've never been able to do.'
We next attempted to determine what acquired deficit had been superimposed on O.C. 's
pre-existing phonological impairment. His well-formed letters, his normal repetition
and copying, and the presence of a severe deficit for all types of input or output suggested
that the impairment could be no more peripheral than the graphemic output buffer. That
his new difficulty was not merely a worsening of his phonological dysgraphia was apparent
from his uniformly high error rates in writing nonwords and regular and irregular words.
Dysfunction of the graphemic output buffer would impair all three categories of stimuli,
but impairment of the lexical writing route in the presence of his developmental
phonological impairment might also be expected to impair his writing in this fashion.
Analysis of error types may help in distinguishing between these and other alternatives,
as specific error types reflect specific patterns of impairment.
Unlike other cases of agraphia in the literature, O.C. is not now impaired by damaged
function in a single component of the spelling system, and we must consider what errors
are likely to result from impairment of phonological segmentation plus another central
deficit. In recent years before his stroke, O.C.'s writing was marked by only very
infrequent errors. He compensated well for his phonological deficit, presumably through
use of the lexical route. If his stroke were to result in a defect localized to the graphemic
output buffer, then the previous good performance of his lexical route should continue
to supply the graphemic buffer with accurate spellings in almost every instance and
his errors should predominantly be those reported as the typical result of graphemic
buffer damage. In this instance, his errors should consist of graphemic deletions,
substitutions, insertions, and transpositions and should often be unpronounceable
combinations violating rules of English orthography. Error rates should be influenced
by stimulus length but not by lexical characteristics such as word frequency or part
of speech. However, if O.C.'s new deficit was related to an impairment of his lexical
writing route, the resulting errors would be less predictable because he would now spell
by using whatever function remains in his defective lexical and phonological systems.
Although we might expect some combination of the characteristics of lexical and
phonological agraphia, it is impossible to predict with certainty the precise constellation
of error types that would be expected or the influence of word frequency, part of speech,
or imageability. He does not appear to have complete disruption of both routes. If this
were the case, he would presumably be completely unable to spell any words.
Table 2 classifies O.C.'s 224 errors made in writing and spelling aloud 240 words
dictated to him. It is apparent that the great majority of his erorrs (78.6%) are
phonologically incorrect neologisms and that the frequencies of different types of errors

1984

A. KIRK AND OTHERS

T A B L E 2. C L A S S I F I C A T I O N OF O . C . ' S ERRORS ON WRITING AND SPELLING A L O U D TO
DICTATION
Writing to dictation (%)

Pttooologtcalry correct noawonj
Phonologially correct real word
PhonologksJIy incorrect.
icmawically uflrelucd
real word
PtionoJogtcally irjcorrect norrword
Denvalioulry rctucd real word
Semantjcally reUttd raU word

Spelling aloud (%)

Non words

Regular

Irregular

Functors

Conlerttrva

Nonnvrds

Regular

Irregular

Total

_
-

13.8
0

20.0
10.0

13.3
0

4.3
0

-

37

0
10.3

7.1
3.6

3.7
96.3
-

13 8
724
0

13 3
56.7
0
0

10 0
76.7
0

8.7
870
0
0

3.4
86.6
9.9
-

14 8
74 1
0

13 8
75 9
0
0

9.4
78.6
1.3
0

0

0

74

0

TABLE 3. GRAPHEMIC CLASSIFICATION OF 224 ERRORS IN
WRITING AND SPELLING ALOUD TO DICTATION

Single deletion
Single substitution
Single transposition
Single insertion
Multiple deletions
Deletion+substitution
Deletion+transposition
Deletion + insertion
Unclassifiable

% writing

% spelling

% total

5.0
9.3
1.4
4.3
2.9
4.3
0
4.3
68.4

10.6
9.4
0
7.1
7.1
1.2
0
2.3
62.4

7.1
9.4
0.9
5.4
4.5
3.1
0
3.6
66.1

to classify O.C.'s errors as graphemic errors as done by Miceli et al. (1985), Caramazza
etal. (1987), Posteraro et al. (1988), and Hillis and Caramazza (1989). Clearly, given
orthographic similarities between stimulus and response, some of the errors made by
subjects with lexical or phonological agraphia will also be classifiable under this system.
Any word or letter combination can be converted into any other word through the
application of multiple deletions, substitutions, insertions, or transpositions (as in Lewis
Carroll's word game, 'Doublets', Carroll, 1879). Therefore, one must beware of carrying
this technique to extremes. Graphemic buffer deficits, such as those reported by Hillis
and Caramazza (1989), frequently produce single errors of this type in a word. To keep
the results meaningful, we adopted the system used by Posteraro et al. (1988), classifying
errors as single graphemic errors, multiple errors of the same type, or double errors of
different types. Complex concatenations of 'errors' of different types were thus considered unclassifiable. Very few of the errors in the reported cases of graphemic buffer
impairment are unclassifiable under this system. It is important to bear in mind, however,

Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016

are similar for the different stimulus characteristics and output modalities. Errors
semantically related to the stimuli are decidedly rare (1.3%). Of O.C.'s 224 errors,
only 7.6% were unpronounceable non words violating orthographic rules, a phenomenon
frequent in subjects with graphemic buffer impairment, and may constitute as many
as 74% of their errors (Posteraro et al., 1988). These errors can also occur in phonological
or lexical agraphia (Hatfield and Patterson, 1983; Roeltgen et al., 19836).
To examine the possibility of a graphemic buffer deficit, an attempt is made in Table 3

PHONOLEXICAL AGRAPHIA

1985

Reading
O.C. read aloud 26/26 upper-case printed letters correctly. He correctly pronounced
the corresponding phoneme for 24/24 printed letters. He was asked to read aloud
phonologically plausible nonwords, regular, and irregular words from the BLAWR.
Each word was individually printed on a card in upper-case letters. Word length and

Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016

that these errors are not at all specific for cases with graphemic buffer deficit. For
example, fully 79% of the errors made by Beauvois and D6rouesne''s (1981) oft-cited
case of lexical agraphia are readily classifiable as 'graphemic' errors (most often single
deletions) using this system. From Table 3, it is apparent that very few of O.C.'s errors
are classifiable as graphemic errors attributable to impairment of the graphemic buffer.
The position of errors within a word may also be related to the type of agraphia.
Hillis and Caramazza (1989) described 2 patients with graphemic buffer impairment,
1 with a left hemisphere lesion who was more accurate at writing the beginning of a
word and 1 with a right hemisphere lesion who was more accurate at the end of a word.
They attributed this phenomenon to hemispatial neglect of a spatially coded representation
in the graphemic output buffer. To assess O.C.'s spelling accuracy as a function of
letter position within a word, we used Wing and Baddeley's (1980) formula for equating
letter positions in words of varying lengths. This formula was used by Caramazza et al.
(1987) and Hillis and Caramazza (1989) to examine the spatial distribution of errors;
we instead examined O.C.'s accuracy, rather than his errors, at each of the five
standardized relative letter positions in response to stimuli of five letters or more (168
of the 240 words and nonwords in Table 1 were at least five letters long). Fig. 2 shows
the frequency with which O.C., in writing and spelling aloud, placed the correct letters
in each of these five positions. Fig. 3 shows the frequency with which the correct letters
for each of these five relative positions appeared anywhere in his response, namely,
in either the correct or incorrect position. The figures show that O.C. is most successful
at writing the initial letter of a word and becomes progressively less accurate at later
letter positions.
The effects of stimulus characteristics may also be important in distinguishing between
different types of agraphia, although in O.C.'s case these are complicated by the
interaction of two deficits. For the 350 words in Table 1 and in the homophone lists
(see next paragraph), the mean length of the words correctly written or spelled aloud
was 4.3 letters, significantly shorter than the mean length of the words spelled incorrectly,
5.1 letters (P < 0.05). For the 290 real words in Table 1 and in the homophone lists
(see below), the mean frequency of usage of words correctly spelled was 265.5/million,
significantly higher than that of words spelled incorrectly, 136.5/million (P < 0.05).
An important characteristic of disturbances of the lexical route in writing is a tendency
to produce incorrect homophones (Hatfield and Patterson, 1983; Roeltgen and Heilman,
1984; Goodman and Caramazza, 1986; Rapcsak et al., 1988). Roeltgen and Heilman's
patients with lexical agraphia each produced incorrect homophones in response to about
a quarter of the homophonic words dictated to them. O.C. was asked to write 36
homophonic words and to spell aloud 74. Each word was dictated to him, used in a
sentence, and then repeated. He correctly spelled 22 (20.0%) of these words, 5 (13.9%)
in writing and 17 (23.0%) aloud. He produced 17 (15.5%) homophone substitutions,
5 (13.9%) in writing and 12 (16.2%) aloud.

1986

A. KIRK AND O T H E R S

cor rect letters in the
Tect posiitions

100

80

60

40

00 0

20

1

2

3

4

5

Relative letter position in stimulus

100
an
c
u

80

as
60

11
FIG. 3. Distribution, as a function of stimulus
letter 'position', of letters present in O.C.'s
responses, whether placed in correct or incorrect
positions. This figure is based on responses to 168
word and nonword stimuli. The letters from each
position in the stimulus are here scored 'correct'
if they appear anywhere in O.C.'s response to that
stimulus.

n

40

20

1 2
3
4
5
Relative letter position in stimulus

frequency of usage were similar for the regular (means 6.4 letters and 68.7/million)
and irregular (means 6.4 letters and 70.2/million) words. Even though the nonwords
were shorter (mean 5.1 letters), it can be seen from Table 4 that O.C. read nonwords
much more poorly than he did either class of real word. O.C. was also asked to read
functors and contentive words from the BLAWR. These stimuli were matched for word
length (mean 4.3 letters for each list). Although the functors were more frequently used
(mean 4343.7 vs 262.6 per million) than the contentives, Table 4 shows that O.C. read
the functors much more poorly than he did the contentives. We also had O.C. read
lists of high and low imagery nouns balanced for regularity, length (mean 4.4 letters

Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016

FIG. 2. Distribution of letters correctly written
or spelled aloud in their correct positions by O.C.
as a function of letter 'position'. This figure is
based on 168 word and nonword stimuli. The
formula used to normalize letter positions in words
of different lengths is taken from Wing and
Baddeley (1980).

PHONOLEXICAL AGRAPHIA

1987

TABLE 4. O.C.'S READING PERFORMANCE
Stimulus category
Letters
Nonwords
Regular words
Irregular words
Functors
Contentives
High imagery words
Low imagery words

No. of
stimuli
26
37
31
33
40
19
10
10

% correct
100.0
13.5
77.4
69.7
62.5
100.0
90.0
80.0

% omissions
0
5.4
12.9
21.2
22.5
0
10.0
10.0

% errors
0
81.1
9.7
9.1
15.0
0
0
10.0

DISCUSSION

This left-handed subject presented after a right parietal infarct with linguistic agraphia
and alexia, but not aphasia. His agraphia was not due to a memory disturbance as he
could repeat a dictated word both before and after incorrectly writing it. His early history
was compatible with developmental dysgraphia and dyslexia. He vividly recalled an
inability to read or write phonetically and was illiterate until, as an adult, he was taught
via a regimen of rote word memorization—as lexically oriented a system as one can
imagine. Writing samples from this acquisition period support the hypothesis that he
had a development phonological dysgraphia and learned to write lexically. Many of
his errors then were phonetically incorrect, implying an impaired phonological system,
and there was a remarkable tendency to repeat his errors, spelling a word in the same

Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016

each) and frequency (mean 151.4 and 151.5/million, respectively). As seen in Table 4,
he performed similarly on the two groups of words.
When reading words and nonwords, errors tended to bear some orthographic similarity
to the targets, that is, the two usually had several phonemes in common. Of 30 errors
in reading nonwords, three were orthographically similar real words (e.g., DOP -—
DOPE; NUD — NUN) and 27 were neologistic nonwords with some similarity to the
targets (e.g., JOFTIB - JOBBITY; ZEMNUL - ZAGLUE). In all instances but one,
the correct initial phoneme was preserved in his erroneous verbal response. In reading
real words, all errors were real words with some orthographic similarity to the target
(e.g., EVENT - EVEN; SOVEREIGN - SOUVENIR). No semantic errors were
seen (unless one accepts FABRIC — FIBRE). The only error that could be construed
as derivational was ADVANTAGE — VANTAGE. Omissions were characterized by
the statements, 'I don't know' or 'That's not a real word'. O.C. was never willing to
hazard a guess at a real word he did not recognize although he admitted that when reading
the nonwords, realizing that they were nonwords, he almost always guessed. When
able to read a word correctly, O.C. could always provide a reasonable definition. When
unable to read a word aloud he could not define it. If a word was incorrectly read as
another word, he gave a definition corresponding to his response rather than to the
stimulus.
O.C.'s difficulty reading nonwords with relative sparing of real words, both regular
and irregular, his errors which orthographically resembled the stimuli, and his difficulty
with functors led us to conclude that he had phonological alexia.

1988

A. KIRK AND OTHERS

Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016

incorrect way every time it appeared. In fact, in extant writing samples from this period,
he never spelled a word in more than one way, suggesting a developing, but as yet
imperfect lexical system with some incorrect but stable representations in the lexicon.
Writing samples from recent years reflect a much greater command of written language.
Rare errors that are phonologically incorrect likely reflect the lack of an effective
phonological system. Campbell and Butterworth (1985) reported a subject, R.E., whose
history was in many ways similar to O.C.'s before his stroke. R.E. also had developmental
phonological dyslexia and dysgraphia and learned to read and write using a lexically
oriented system of whole word memorization. She also achieved a high level of literacy
but, like O.C., still made occasional errors which were not phonetically correct. Following
his stroke, O.C.'s writing was severely impaired. With the exception of single letters,
he performed extremely poorly at writing and spelling aloud all classes of stimuli. He
showed neither the relative sparing of regular words and nonwords seen with lexical
agraphia nor the sparing of real words typical of phonological agraphia. His letters were
always well-formed, ruling out a peripheral motor disturbance as an explanation for
his writing impairment. Severe impairment was seen for all output modalities tested
(writing, spelling aloud, and anagram letters). Written naming was as impaired as writing
to dictation. Copying was normal even when O.C. was asked to copy print as script,
and upper as lower case. These characteristics effectively rule out a deficit any more
peripheral in the writing system than the graphemic output buffer.
O.C.'s ability to accurately transcribe single phonemes illustrates preservation of
phoneme-grapheme conversion but his extreme difficulty on LAC category II reflects
a pronounced segmentation deficit, which presumably was responsible for his inability
to learn to write phonetically. Following his stroke, O.C.'s writing deteriorated but
did not display the typical features of phonological graphia. Thus, an acquired deficit
has been superimposed on his developmental phonological dysgraphia with segmentation
deficit. O.C. infrequently produced unpronounceable nonwords violating orthographic
rules, a feature which is, by contrast, common in the writing of subjects with a deficit
of the graphemic output buffer. Word length, frequency of usage, and part of speech
all influenced O.C. 's spelling accuracy. The majority of his errors were not the graphemic
deletions, substitutions, insertions, or transpositions characteristic of patients with
graphemic buffer impairment. Instead, most of O.C.'s erroneous responses in writing
and spelling aloud all classes of stimuli were phonologically incorrect neologisms which
nevertheless often had letters in common with the stimulus. As shown in figs 2 and
3, O.C. was better at spelling the beginning of a word and became progressively less
accurate as he continued towards its end. Although Hillis and Caramazza (1989) have
ascribed a similar pattern (with a different error type) to graphemic buffer impairment
in a subject with a left hemisphere lesion, this has been attributed to right hemineglect
of a spatially coded representation within the buffer. If O.C.'s agraphia were due to
a similar attentional mechanism, his right hemisphere lesion would be expected to
produce, if anything, the opposite profile, with greater accuracy at the end of a word.
For several reasons, it thus does not seem reasonable to attribute O.C.'s agraphia to
graphemic output buffer impairment.
If O.C. 's infarct were to result in a lexical agraphia superimposed on his pre-existing
phonological dysgraphia with segmentation deficit, one would also expect to see severe
disruption in writing all classes of stimuli except single letters. No longer able to accurately

PHONOLEXICAL AGRAPHIA

1989

Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016

write using a lexical route and with no effective phonological route on which to fall
back, he would be forced to rely on remaining function in these two imperfect writing
routes. In arriving at spellings for words in this way, word frequency, length, and part
of speech might all influence accuracy, as they do in O.C. His difficulty in writing
homophones and frequent substitution of incorrect homophones has been reported as
a characteristic of lexical agraphia (Hatfield and Patterson, 1983; Roeltgen and Heilman,
1984; Goodman and Caramazza, 1986).
Roeltgen et al. (19836) and Bolla-Wilson et al. (1985) have shown that subjects with
phonological agraphia with disruption of the segmentation component of the phonological
writing route make characteristic errors. Like O.C.'s, their written productions
share many letters with the targets but initial letters are much more often correct
than are letters later in the word. Horowitz et al. (1968) suggested that the initial
grapheme/phoneme of a word is its most salient, containing the most relevant
discriminative information. O.C. thus made errors typical of phonological agraphia with
impaired phonemic segmentation but was severely impaired at writing all classes of
words, not just nonwords as in phonological agraphia. This is a pattern one would expect
in a subject with phonological agraphia with segmentation deficit who has also sustained
functional damage to the lexical writing route. We thus postulate that O.C. suffered
from a developmental phonological dysgraphia with segmentation deficit, which he
overcame by using lexical strategies localizable to his right parietal lobe. When his small
right parietal infarct impaired his lexical writing route, he was forced to rely more on
his ineffective phonological route, resulting in a combined phonological and lexical or
'phonolexical agraphia'.
Less can be said about the evolution of O.C.'s reading impairment as we have no
prestroke examples of his reading and must rely on self-report. By history it seems most
likely that he had a developmental phonological dyslexia and learned to read lexically.
Testing after his stroke reveals typical characteristics of phonological alexia (Patterson,
1982). Like Campbell and Butterworth's (1985) subject, R.E., O.C. usually produces
neologistic responses to nonwords and does not tend to 'lexicalize' or read nonwords
as visually similar real words as is common in acquired phonological alexia (Patterson,
1982; Funnell, 1983). Patients with acquired phonological alexia may read nonwords
as visually similar words because these nonwords are able to 'activate' lexical entries
for similar real words, unchecked by phonological confirmation. O.C. and R.E., on
the other hand, who have achieved literacy in the absence of an effective phonological
route, have perhaps had to develop a more discriminating lexical reading system which
does not as readily accept a word as another visually similar word, which recognizes
a nonword as a nonword but, due to poor phonological functioning, can only offer an
approximate nonword rendering. Although O.C. sometimes failed to recognize a real
word and said that it was not a word, he always refused to guess at it while he confessed
to usually guessing at nonwords. We might surmise that his alexia worsened following
his stroke but, without prestroke testing, we can conclude little else.
The right posterior parietal localization of O.C.'s stroke is consistent with the location
of lesions responsible for lexical agraphia in the left (Roeltgen and Heilman, 1984) and
in the right hemispheres (Gonzalez Rothi et al., 1987), lending support to the contention
that one or the other posterior parietal lobe is essential to the lexical route in writing.
No lesion is visible to explain O.C.'s developmental literacy problems. Galaburda et al.

1990

A. KIRK AND OTHERS

(1985) have documented microscopic abnormalities predominantly affecting the left
perisylvian regions of patients with developmental dyslexia and dysgraphia.
O.C.'s lack of aphasia after his right parietal stroke may reflect left hemisphere
dominance for speech, or may relate instead to the small lesion size as aphasia is usual
but not invariable in dextrals with agraphia or alexia after left hemisphere lesions (Dubois
et al., 1969; Damasio and Damasio, 1983).
This case, like that of Campbell and Butterworth (1985), illustrates that written language
deficits can be overcome in at least a subtest of cases by using an alternate educational
strategy, in this case lexical rather than phonetic. In O.C.'s instance, a right parietal
anatomical substrate can be inferred for this acquisition. To what extent this localizing
information and the characteristics of the agraphia can be generalized remains to be
seen. Further cases with acquired deficits superimposed on developmental language
disorders are likely to be rare but instructive.

We thank Dr Melvin Greer for referring O.C. and Dr Kytja Voeller for introducing us to the LAC test.
Most of all we thank O.C. and his wife for their willing cooperation. This study was supported in part
by the Memory Disorders Clinic of the Division of Aging, Department of Health and Rehabilitative Services
of the State of Florida and by the Medical Research Service of the Department of Veterans Affairs.

REFERENCES
ASSAL G, BUTTET J, JOUVET R (1981) Dissociations in aphasia: a case report. Brain and Language, 13,
223-240.
BEAUVOIS M-F, DEROUESNE' J (1981) Lexical or orthographic agraphia. Brain, 104, 2 1 - 4 9 .
BODER E (1973) Developmental dyslexia: a diagnostic approach based on three atypical reading-spelling
patterns. Developmental Medicine and Child Neurology, 15, 663—687.
BOLLA-WILSON K, SPEEDIE LJ, ROBINSON RG (1985) Phonologic agraphia in a left-handed patient after
a right-hemisphere lesion. Neurology, Cleveland, 35, 1778—1781.
BUB D, KERTESZ A (1982) Deep agraphia. Brain and Language, 17, 146-165.
CAMPBELL R, BUTTERWORTH B (1985) Phonological dyslexia and dysgraphia in a highly literate subject:
a developmental case with associated deficits of phonemic processing and awareness. Quarterly Journal
of Experimental Psychology, 37A, 435—475.
CARAMAZZA A, MICEU G, VILLA G, ROMANI C (1987) The role of the Graphemic Buffer in spelling:

evidence from a case of acquired dysgraphia. Cognition, 26, 59 — 85.
CARROLL L (1879) Doublets—a verbal puzzle. Vanity Fair, 21, 185-186.
DAMASIO AR, DAMASIO H (1983) The anatomic basis of pure alexia. Neurology, Cleveland, 33, 1573-1583.
DUBOIS J, HECAEN J, MARCIE P (1969) L'agraphie 'pure'. Neuropsychologia, 7, 271-286.
ELUS AW (1982) Spelling and writing (and reading and speaking). In: Normality and Pathology in Cognitive
Functions. Edited by A. W. Ellis. London: Academic Press, pp. 113—146.
FRITH U (1980) Unexpected spelling problems. In: Cognitive Processes in Spelling. Edited by U. Frith.
London: Academic Press, pp. 495—515.
FUNNELL E (1983) Phonological processes in reading: new evidence from acquired dyslexia. British Journal

of Psychology, 74, 159-180.
FUNNELL E, DAVISON M (1989) Lexical capture: a developmental disorder of reading and spelling. Quarterly
Journal of Experimental Psychology, 41 A, 471 —487.
GALABURDA AM, SHERMAN GF, ROSEN GD, ABOITIZ F, GESCHWIND N (1985) Developmental dyslexia:

four consecutive patients with cortical anomalies. Annals of Neurology, 18, 222-233.

Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016

ACKNOWLEDGEMENTS

PHONOLEXICAL AGRAPHIA

1991

GONZALEZ ROTHI LJ, ROELTGEN DP, KOOISTRA CA (1987) Isolated lexical agraphia in a right-handed

(Received December 14, 1989. Revised December 14, 1990. Accepted January 3, 1991)

Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016

patient with a posterior lesion of the right cerebral hemisphere. Brain and Language, 30, 181 —190.
GOODMAN RA, CARAMAZZA A (1986) Dissociation of spelling errors in written and oral spelling: the role
of allographic conversion in writing. Cognitive Neuropsychology, 3, 179-206.
HATFIELD FM, PATTERSON KE (1983) Phonological spelling. Quarterly Journal of Experimental Psychology,
35A, 451-468.
HILUS AE, CARAMAZZA A (1989) The Graphemic Buffer and attentional mechanisms. Brain and Language,
36, 208-235.
HOROWITZ LM, WHITE MA, ATWOOD DW (1968) Word fragments as aids to recall: the organization
of a word. Journal of Experimental Psychology, 76, 219—226.
KAY J, MARCEL A (1981) One process, not two, in reading aloud: lexical analogies do the work of non-lexical
rules. Quarterly Journal of Experimental Psychology, 33A, 397—413.
KERTESZ A (1982) Western Aphasia Battery Test Manual. New York: Grune and Stratton.
KUCERA H, FRANCIS WN (1967) Computational Analysis of Present-Day American English. Providence,
RJ: Brown University Press.
LINDAMOOD CH, LINDAMOOD PC (1971) Lindamood Auditory Conceptualization Test. Boston: Teaching
Resources.
MICELI G, SILVERI MC, CARAMAZZA A (1985) Cognitive analysis of a case of pure dysgraphia. Brain
and Language, 25, 187—212.
MORTON J (1980) The logogen model and orthographic structure. In: Cognitive Processes in Spelling.
Edited by U. Frith. London: Academic Press, pp. 117—133.
NEWCOMBE F, MARSHALL JC (1980) Transcoding and lexical stabilization in deep dyslexia. In: Deep
Dyslexia. Edited by M. Coltheart, K. Patterson and J. C. Marshall. London: Routledge and Kegan
Paul, pp. 176-188.
NOLAN KA, CARAMAZZA A (1983) An analysis of writing in a case of deep dyslexia. Brain and Language,
20, 305-328.
PAIVIO A, YUILLE JC, MADIGAN SA (1968) Concreteness, imagery, and meaningfulness values for 925
nouns. Journal of Experimental Psychology Monograph Supplement, 76, No. 1, Pt. 2, 1 -25.
PATTERSON KE (1982) The relation between reading and phonological coding: further neuropsychological
observations. In: Normality and Pathology in Cognitive Functions. Edited by A. W. Ellis. London:
Academic Press, pp. 77—111.
POSTERARO L, ZINELU P, MAZZUCCHI A (1988) Selective impairment of the Graphemic Buffer in acquired
dysgraphia: a case study. Brain and Language, 35, 274-286.
RAPCSAK SZ, ARTHUR SA, RUBENS AB (1988) Lexical agraphia from focal lesion of the left precentral
gyrus. Neurology, Cleveland, 38, 1119-1123.
ROELTGEN DP, CORDELL C, SEVUSH S (1983a) A battery of linguistic analysis for writing and reading.
The INS Bulletin, October.
ROELTGEN DP, SEVUSH S, HEILMAN KM (19836) Phonological agraphia: writing by the lexical-semantic
route. Neurology, Cleveland, 33, 755—765.
ROELTGEN DP, HEILMAN KM (1984) Lexical agraphia: further support for the two-system hypothesis of
linguistic agraphia. Brain, 107, 811-827.
ROELTGEN DP, TUCKER DM (1988) Developmental phonological and lexical agraphia in adults. Brain
and Language, 35, 287-300.
SHALUCE T (1981) Phonological agraphia and the lexical route in writing. Brain, 104, 413-429.
TEMPLE CM (1986) Developmental dysgraphias. Quarterly Journal of Experimental Psychology, 38A,
77-110.
TEMPLE CM (1988) Developmental dyslexia and dysgraphia persistence in middle age. Journal of
Communication Disorders, 21, 189-207.
VENEZKY RL (1979) Orthographic regularities in English words. In: Processing of Visible Language, Volume
1. Edited by P. A. Kolers, M. E. Wrolstad and H. Bouma. New York and London: Plenum Press,
pp. 283-293.
WING AM, BADDELEY AD (1980) Spelling errors in handwriting: a corpus and a distributional analysis.
In: Cognitive Processes in Spelling. Edited by U. Frith. London: Academic Press, pp. 251-285.

1992

A. KIRK AND OTHERS
APPENDIX
STIMULI GIVEN TO O . C . , FOLLOWED BY HIS RESPONSES
Writing to dictation
(I) Noawords
TRAD
PABLE
TERRIAGE
MANVER
JISP
BIROUGH
ZILLER
QUIM
VATTER
BUME
SOUD
C1MY
SCOMB

nXEND

+
+
SIMAD
CONSE
ELAN
THRUG
CRAG
FLEGT
THOOU
SELAM
CEGFJ
TRUCH
STICH
BARCH

+
ILET
VRLER
ANGON
INTHER
THAWA
NAIC

(d) Functors
ALTHOUGH
HOWEVER
WHOSE
SHALL
THUS
EVER
SHOULD
UPON
THOUGH
ALREADY
ALMOST
MAYBE
WHOM
NONE
NEARLY
ELSE
ALWAYS
AMONG
WHETHER
WHERE
NOBODY
QUITE
STTU.
EACH
BOTH
ONCE
EVERY
THINGS
WTTHIN
ITSELF

ALTHUR
HOWEAER
HOSER
CLERL
THART
EVERN
SHOUD
UPROUN
THPOW
ALREAD
AMOST
MAYBEE
HOME
NUMER
NALER
EALSE
AWAY
ARMOUN
WHER
WHARE
NOBOUY
QU
STRAL
EATECH
BOATH
OUNCE
EVARE
THERTIN
WHHN
tTSALE

(b) Rr(iiu wards
TRANSFER
NEGLECT
MASK
PITIFUL
MOTOR
SINK
FORGET
HOLDER
ADMIT
HOLY
INJURY
MAKER
FACTOR
HUNGER
DAMP
PINCH
SUBMIT
CORN
DIVINE
HOTEL
GRAND
HALTER
OPEN
INFORM
VETERAN
VENT
STAR
DARN
STAND
MATCH

TRAFAD

+
MAX
P1FTFAL
MONTAR
CENK
FORCUT
HLOUD
ADMAT
HOLLAT
INGJRE
MAKRE
FACTER
HUNTER
DIAE
FINER
SBMIAT
CAN
DV1AN
HOLTEEL
GRATIN
HOTLTER
OPTAN
INFROM
VACTION
VEAT
STARE
DA MR
SAN
MATICH

(c) Coraenuvc*
GENERAL
MONEY
HELP
COLLEGE
BEST
INTEREST
PART
PEOPLE
TAKEN
NUMBER
UNITED
METHOD
DEEP
LEADER
CLUB
TRIAL
LIVE
POINT
PUBLIC
YEARS
MAJOR
HISTORY
STATE
WORK
LONG
HIGH
SMALL
LOOKED
CHURCH
MATTER

OENA
MONDAY
HAEP
CALGET
BESTER
INTREST
PERT
POPUTY
TAKING
NUMTER
UNIGHT
MADFOED
+
+
+
TRYRALL
+
PONT
PUBLEAR
YEAR
MARGER
HISTORE

+
+
LOND
+
SMALE
LOOK
CHERE
MATLAR

(c) Irregular words
MISCHIEF
TOMB
HONOR
DENY
CIRCUIT
HEIR
ENGINE
BOUQUET
SHOES
WEAPON
CHORD
YACHT
PALACE
FRIEND
BUSY
ISLE
CORPS
ANSWER
BUOYS
HYMN
BREAST
EARTH
MENACE
BOMB
SURE
GLACIER
MORGUE
SUBTLE
AISLE
YOLK

MIHCER
TOO AM
ONHAR
DIERE
CE
HIRE
ENGRON
BOUKEY
SHOUE
WIPIN
CORD
EYOUT
PALTIST
FOREN
B1SEZ
IL
CORE
ANCENT
BOUT
H1MM
BRESET
EARS
MEASET
BOND
SCHURE
GIASHR
MOBGER
SETRAL
IRALE
YOKE

(0 Homorjhones
SEA
NIGHT
HAIR
SLAY
YOKE
HOES
CORD
CLOTHES
FLEE
BAWLED
CAUGHT
SELL
FIR
RING
ROSE
PANE
PAIR
SON
SEE
CLOSE
KNIGHT
HARE
SLEIGH
YOKE
HOSE
CHORD
FLEA
BALD
COT
CELL
FUR
WRING
ROWS
PAIN
PEAR
SUN

SEIE
MIGAT
HERIE
SL
(Omen km)
HOLDS
SORED
CLAST
FLED
BALD
(Omission)
SEALD
FUR
RIND
+
PAINE
PARY
SUN
SEAE
+
NIGHT
HERE
SLAID
YOUE
HOUSE
CORE
FLEED
BOULD
COUD
SEALD
FURD
RING
ROW
+
+
+

Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016

TROOCE
CRANO
PHEKE
THUSE
SLEM
SUO
TRALF
SHUCK
BARCLE
MOFER
1SLECK
VYTE
ANDON
INTRET
THEAWAY
NACE

TAIRD
PALJE
THI
NAMIFD
GIGF1
BARORD
ZEAUD
QUMEN
FTTED

PHONOLEXICAL

AGRAPHIA

1993

(g) Written naming
Stimulus

Written naming

Writing to dictation

BAND
RING
CARD
SPOON
BRACELET
PENCIL
ASHTRAY
PURSE
PAPER
WALLET
DOLLAR
KEY
BOOK
PEN
SHIRT

+
+
+
SPO
BRACLET
PENNCEL

+
+
CORD
SPOOLEN
BRACLET
PINCEIL
+
PURESH
PAPPER
WAILLOTE
+
QUUE
BOOOK
+
+

+
POT
PAPPER

+
DOLLER
QUTLER

+
+
TSH

Spelling aloud to dictation
(a) Nonwords
NEID
BAT
RENG
THUOGH
RITCOME
BUOGH
RIDRON
GOODING
SISBRY
BUMPKEN
GIBSAW
FUSHBR
CROWER
TRIAG
LUCO
CATMUOT
HIMFEN
TTAL
BIFLOC
RABSULD
IOBIFLE
KLBON

+
NIBBIC
OLBJOR
JUISE
FROUGH
BLUD
COLLLAP
FURLJNE

(c) Irregular words
KNOT
MENACE
PLUMB
MISCHIEF
WRAP
WEAPON
SCENE
SIGHED
PRESTIGE
BOROUGH
TROUBLE
EYE
SHOES
KNOWS
HEIR
PRAIRIE
SCENT
PALACE
ISLE
EARTH
CHORD
SKY

(Omission)
MINNES
PLUM
MISSCH
WIP
WHEPIN
SLEN
SIDE
PRESTAGER
BROW
TRUBELEF
(Omission)
SHOUSE
KNO
ARIE
PRE
CENTER
PAUCE
ILEREY
EARGH
CORD
SKI

(b) Regular words
SELL
PATIENTS
ARTIST
INSECT
ABDOMEN
NEWSPAPER
PROSPERITY
UBRARY
HOLE
ARMY
GENTLEMAN
FRUIT
GRASS
ARC
ANGER
EAR
ANIMAL
BARE
BUY
CHAIR
MEET
BALE
THRONE
LENDS
BAIL
BUTTER
FLESH
PAIL
BEAST
CHIEF

SILER
PANC
ARTTESE
INSCENCE
ADDMENT
NEWPAPPER
PROSTTER
UBBRE
HOLD
ARIME
GENTMAN
FRU
GREET
ARCH
ANGLER
EIR
AN
(Omission)
BY
CHIR
MEAT

+
THROUGH
LENDSEN
+
BUTER
FL
+
BES
CHEF

(d) High imagery words
CITY
BOOK
CHAIR
BODY
DRESS
ANIMAL
CORN
ARM
CHILD
ARMY

+
+
CHIR
+

+
AN1

+
+
CHI
ARIME

(e) Low imagery words
+
FATE
+
LAW
+
EVENT
+
DUTY
(Omission)
HONOR
ADVICE
+
IDEA
ID
FAT
FACT
HOPE
+

Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016

NID
BAP
RIN
FALXO
RITKON
BUTCHO
RIBBLE
GUDIN
SIBWAR
BLUMPK1N
GIBSOG
FUSHIB
KROGUD
TRAG
LUTCHO
KAFMAL
HANFIM
TEDDLE
BAFLOD
RABZIL
JOFT1B
KELBON
VAGGLE
NIBER
OBJOL
JISH
FOSH
BUB
KOLAT
FUG

A. KIRK AND OTHERS

1994

(c) Irregular wofds

(e) Low imagery words

GUILT
VILLAGE
HOUR
OCEAN
HONOR
DOUGH
WEAR
REIGN

GULED
VILGET

TRUTH

CELL
RODE
MAIL
GALE

SEAL
ROUD

TRUE

+
OCEEN
(Onnssion)
DO
(Omission)
RANNER
( 0 Homophoaes

+
GALER
YALE
MEAT

AIL

MEET
BAIL
WHERE
GAIL
HEIR

+
WHIRE
(Omission)
HIRE
IL

WHOLE
SELL
MALE
ROAD

HOLD
SILER
MILE

ALE

+
+
+

ROB

MEAT
BALE
WEAR

(Omission)

AIR

+

SIGHED
HOLE

SIDE
HOLD
(Omission)
BEEAT

EYE

BEAT
DEAR
FEET
THRONE
ONE
FLEA
PAIL
SALE
DEW
KNOT
ARC

SEAUNG
THYME
DOUGH

+
+
THOROUGH

+
FLEED

+
+
DOW

(Omission)
ARCH
CELN
THTHME
DO

BOW

+
BENN
(Omission)
CORER
NEED
LED

LENDSN
ACTER
ALD

AWL

BASS
BEAU
BOW (ship's)
BARE
CHORD
GUILT
KNEW
HEARD
HIGHER
HYMN
WHOLLY
HOUR
KNIGHT
CACHE
CASTE
SIGHT
CLIMB
MIGHT
KNIT
KNOW
NONE
TAIL
EIGHT
PATIENTS
PLUMB
PREY
HARE

+
BOW

(Omission)
(Omission)
CORD
GULED
NEW

HERD

+
HIMMER
HOLER
-t-

NIGHT
CASH
CASH
(Omission)
CLAME
M1NET
KJ

+
NUNN
TALE

+
PANC
PLUM
PRA

HIRE

Reading aloud
(a) Nlonword.
NID
BAP

BUTCHO
BLUMPKIN
FLIM
NUD
JAF

HOMFIS
JOFTIB
LODAR
RABZIL
KAKMIN
TIBBLE
VAGTOB
TRAG
BEP

KAFMAL
GOF

KACN1M
RIBBLE
SIBWAR
JISH

+

(Omiksion)
but{tu

+
FLIN dim)
NUN

+
dyAsli
lab

+
kinmtn
libu
vijarin
-t-

BEEP
(Onrasjoa)
gvdz
Id
nn
SEf

dj^3

DOD
JAD

dab

FIMKO

Hpe

djxl

(h) Regular words
ANIMAL
GREEN
INTEREST
FREEDOM
FROG
ADVICE
HISTORY
REFLEX
CONTRACT
MASTER
ATTITUDE
CORNER
SADNESS
SPEECH
CHIN
MAGNITUDE
LETTER
BOTTLE
INDUSTRY
GARDEN
BLANDNESS
HARNESS
SUNSET
LOYALTY
VICTIM

+

+
+
+
(Omuvwon)

+
REFLUX

+
+
+
+
+
+
(Omission)

+
+
+
+
+
BUNDNESS

+
+
4.

Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016

ILL

BOWLED
RIGHT
BEEN
COLONEL
CORPS
KNEAD
LEAD(Pb)
LENDS
ACTS

1995

PHONOLEXICAL AGRAPHIA
(l) Nonwords
HUD
VAGCLE
LUMRIN
KOLAT
PISHNEP
KELBON
DOP
Z1MNUL
OBJOL
FUBBIK
JOSF1D
FUSHIB

(b) Rcgulir words
purjjo
vcdpdjidsi
lAmbaruu
LolEp
ptjmir
itibon
DOPE
zjeglu
ob=d3^
f<|.-.bi
djasofidi
fiibor

(c) Irregular words

+
(Om 155 ion)
VANTAGE
+
+
+
+
+
(Omission)
(Omission)
SOUVENIR
+

+
+
+
+
(Omission)

+
+
+
4-

+

+
(Omission;
+
FIBRE
+

(d) Higb imagery words
CITY
BOOK
CHAIR
DRESS
ANIMAL
BODY
CORN
ARM
CHILD
ARMY

f

(Omission)
f
f
f

(c) Low imagery words
FATE

LAW
EVENT
DUTY
HONOR
ADVICE
IDEA
FACT
HOPE
TRUTH

(OmiSHOO)
f

EVEN
4f
f
f
f
f
f

(Omission)

+
MORTGAGE
+
+
+
+
+
+
(Omisston)
(Omission)

(0 Functors
WHY
SHALL
THE
HAS
FROM
ABOUT
WHO
THESE
WHILE
SHOULD
UNTIL
WAS
ALSO
QUITE
MUCH
WITH
THOUGH
AND
VERY
EVER
SINCE
GET
WHEN
THAT
USUALLY
HAD
THEN
WHATEVER

(Omission;

(Omission)
(Omission)
+
(Omission)

+
+
+
(Omission)
(Omission)
SHOULDER
+
+
+
QUIET
(Omission)
WIT
THOROUGH
+
+

+
SIN
+

+
+
(Omission)
+

+
+

(g) Conlciu VCS
ARM
RIVER
BODY
WALL
CHAIR
BLOOD
BOY
GOLD
FRIEND
WORLD
MAN
BED
BOOK
WATER
FOOT
MOTHER
CAR
WOMAN
TOWN

+
+
+
+
+
f-

+
+
i
^
H
H
H
H
Hh
^
Hh
Hh

Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016

CHTEF
DEATH
ADVANTAGE
ENGINE
TOMB
CIRCUIT
VILLAGE
HEAVEN
JEALOUSY
HEALTH
SOVEREIGN
SWAMP
ANSWER
THOUGHT
BREAST
PLEASURE
UMB
EARTH
RITUAL
MARRIAGE
JUSTICE
WALL
EDIFICE
BEVERAGE
MORGUE
KERCHIEF
COMRADE
MEADOW
FRIEND
HONOR
HOSTAGE
GLACIER
EPISTLE

BRAVERY
EJTORT
NUTMEG
FLASK
FABRIC
BANDIT

1996

A. KIRK AND OTHERS
ONCE
THOSE
WERE
HOW
NOR
OUGHT
THUS
WHERE
AS
HOWEVER
ELSE
THIS

+

+
+
+
(Omission)
OUCH
(Omus*on)
+
+
+
+
+
— correct response.

Downloaded from http://brain.oxfordjournals.org/ by guest on June 7, 2016