Agrammatism
and Conduction A Chinese Case
Fred C. C. Peng
Yiu-Tong Chu Far Eastern Memorial
Aphasia:
Hospital
International
Christian
University
Hoi-Keung Yiu Far Eastern Memorial
Hospital
ABSTRACT A 60-year-old right-handed Chinese man was well until one week prior to admission because his family noted that he seemed “confused.” Upon examination, we observed that he slurred his speech and that while he could recognize some familiar things he lost his ability to calculate. There was no motor impairment physically nor was there any sensory loss, other than his speech problem. We thus conducted a series of neuropsychological tests for the purpose of our neurolinguistic investigation. The results were that repetition was fluent but paraphasic and that his natural speech was fluent but he was poor in naming. Also noted were severe agraphia, fmger agnosia, acalculia, and left-right disorientation. Neurological examination was otherwise normal. But the CT scan of the brain revealed a left parieto+ccipital infarction. This case was therefore diagnosed as conduction aphasia. Of his various deficits one stands out distinctly, which is agrammatisam. However, this patient’s agrammatism differs considerably from that reported in the two cases of transcortical aphasia. This presentation will focus on such features of agrammatism as may be observed in our patient, although we cannot say that his agrammatism is typical of conduction aphasia among the Chinese patients because this is the fast case ever reported; there are no other case reports in the literature with which to compare.
INTRODUCTION Agrammatism has been a subject of interest in neurolinguistics (or the neurology of language, for that matter) in recent years. But it actually began to receive attention from aphasiologists as early as the turn of the century (1902-22). Although many articles and books have been published in aphasiology on this subject
210
Journal
of Neurolinguistics,
Volume
2, Number
2 (1986)
since then, they have, unfortunately, more or less skimmed over the surface without any attempt of an in-depth investigation of the substance to look into the nature and extent of this form of disturbances in language behavior. The reason for the luke-warm attitude - with the exception of the German school of aphasiology towards this sort of study may be best illustrated by Critchley’s statements (1970: 16; 36-7) in which he tried to separate the usages of agrammatism and telegrammatism but claimed that “We do not pay as much attention nowadays to these concepts as formerly.” He then went on to define the former as “an aphasic disorder which impairs syntax rather than vocabulary” and the latter as “the type of speech which results after the sacrifice of the shorter terms which are less heavily charged with reference-function, or ‘meaning’.” More astounding, however, was Critchley’s criticism of Pick’s attempt to bring the science of language to bear upon the problems of aphasia. Thus, Crititchley said: “In his monograph upon agrammatism (1917) he repeatedly resorted to the evidence of linguistics. Maybe Pick endowed syntax with an undue importance in the role of breakdown of grammar in the fragmentary utterances of aphasiacs (Critchley 1970: 36-7). Whether or not Pick, as Critchley said, endowed syntax with an undue importance is not an issue here. What is important is that from the point of view of neurolinguistics agrammatism and its related forms of language disturbances are very significant indicators of how the brain as a whole (or even the mind) works uis-b-vis language behavior which represents humans as against nonhuman primtes. Put differently, how the normal brain is supposed to work in its intact state is reflected in how the damaged brain continues to function in language behavior which is found to be defective and agrammatic. It is, therefore, rather appalling to find such indifferent statements in the literature, even though only a few leading specialists have been interested in the breakdown of grammar in aphasias. Perhaps, we should not be too critical of our precedessors’ somewhat naive remarks on the importance of the role of grammar (at the sentence or discourse level) in the study of aphasic speech, if only because there were not many linguistically sophisticated neurologists working in this area. But the fact remains that agrammatism and its related forms of language disturbances are still relatively little known, chief among the main reasons being that not enough data have been collected outside the domain of European languages (notably German and English). The present article is intended to fill such a gap, no matter how small its contribution. One comforting but highly relevant point to make in our contribution is that the Chinese language, unlike Critchley’s grossly mistaken view, is not virtually devoid of syntax, but rather fuB of syntax (1970: 16); as a matter of fact, Chinese grammar is nothing but syntax and semantics and, thus, any breakdown of its syntactico-semantic properties constitutes a serious violation of its grammar. it is this breakdown or violation
Agrammatism
and Conduction
Aphasia:
A Chinese
Case
211
of grammar in our aphasic patient’s language behavior that is our main focus. The following is a brief medical history of this patient whom we shall call L.K. L.K. was a 60-year-old right-handed Chinese man who was well until one week prior to admission because his family noted that he seemed “confused.” His family members also noted that he slurred his speech and that while he could not name some familiar objects around the house he also lost his ability to calculate, calculation being part of his daily business. When he was admitted to our hospital, we observed and examined him and found that there was no motor impairment physically nor was there any sensory loss, other than his speech problem, because he was alert, well-developed, and cooperative; however, his ability to identify objects placed in the right hand was impaired. He denied headache, chest pain, cardiac palpitation, alterations of consciousness, visual changes. Hypertension had been diagnosed four years prior to admission, however. There was no history of venereal disease, diabetes, or head trauma. Review of all systems was unremarkable, except for his BP which was 160/90 mmHg. There was no respiratory difficulty either. The left carotid pulsation was slightly weaker than the right. Nonetheless, he was alert and able to follow commands. For instance, he could write his name with his right hand, when told; upon request, he could also imitate such actions as lighting a match. We thus conducted a series of neuropsychological tests for the purpose of our neurolinguistic investigation. The results were such that repetition was nonhesitant but paraphasic and that although his natural speech was fluent he was poor in naming. Also noted were severe agraphia (other than writing his own name), fmger agnosia, acalculia, and left-right disorientation. Neurological examination was otherwise normal. But the CT scan of the brain revealed a left parieto-occipital infarction. This patient was, therefore, diagnosed as a case of conduction aphasia. Note that L.K. was admitted to our hospital one week post onset of his illness, the admission date being March 13,198s. He underwent the first neuropsychological test on March 14 and subsequent neurolinguistic tests, intermittently, from March 14 through 19, and then March 28 to May 20. During this period, L.K. showed signs of some improvement regarding sentence constructions but the data to be presented below were based on the neurolinguistic testings conducted one week to one month post onset. Of his various deficits one stands out distinctly, which is agrammatism. However, this patient’s agrammatism differs considerably from that reported in the two cases of transcortical aphasia (this volume). Thus, our purpose of presenting this case is to focus on such features of agrammatism as may be observed in L.K. so as to describe them for a comparison with the agrammatism of the two transcortical aphasics just noted. Although we cannot say for sure that our patient’s agrammatism is typical of conduction aphasia among the Chinese aphasic
Figure 1. L. K.‘s CT Scan
population, because this is the first case ever reported, there being no other case report in the literature to our knowledge, we are of the opinion that our description will be of value in the years to come; it offers useful insights regarding the nature of agrammatism in general and Chinese agrammatism in particular. CASE DESCRIPTION Before we describe our case, we should reiterate what we mean by “agrammatism,” because the term in current usage means the same thing as “telegraphic speech” which is reserved for motor aphasics. However, in some literature, for instance, Goodglass (1973: 187) and Maruszewski (1975: log), agrammatism is divided into motor agrarmnatism and sensory agrammatism, whereby the former is said to be characterized by labored speech which is often reduced to isolated substantives and short phrases and the responsible lesion is usually in the frontal lobe of the dominant hemisphere. Sensory agrammatism, on the other hand, is also equated with paragrammatism (Goodglass 1973: 187) the characteristics of which are fluent speech albeit containing garbled syntax and omitted or confused particles and/for inflections (Goodglass and Blumstein 1973: 5) its responsible lesion site being thought to be in the temporal lobe of the dominant hemisphere. The shared feature and difference are then the abundance of grammatical errors, although paragrammatism does not seem to show the impoverishment and stereotypy of available syntactic structures which characterize (motor) agrammatism (Goodglass
Apmmatism
and Conduction
Aphasia:
A Chinese
&se
213
and Blumstein 1973: 5). In the present context, however, we wish to use the term “agrammatism” in a different sense for reasons that will become clear later: (1) We are dealing with a language that has neither inflections nor derivations that are typical of European languages; (2) grammar to us means the mental processing of linguistic information at the conceptual level (not at all at the production level); and (3) such being the case, we must delve into the mental activities of the subject while minimizing motor skills at the production stage, because all motor skills tend to have laterality effects. All this is based on our beliefs that grammar is centrally and bilaterally represented in the brain and, therefore, any test through speech is bound to be heavily biased by laterality effects. Thus, we chose the method that best suits the bilateral representation while confining motor skills to the bare minimum; this method is the use of word cards and follows the procedure listed below: 1. Visual Stimulation with an Auditory Reinforcement 2. Visual Stimulation without any Auditory Reinforcement 3. Auditory Stimulation only 4. No Stimulation provided except an Instruction to construct a Sentence when the Word Cards are provided. The first step of the procedure is to show whether or not the patient can copy a sentence or a phrase, using an identical set of word cards, without involving too much of his motor skills. Unlike copying in the usual sense, which is to write down what one sees word for word, the present method requires only that L.K. arrange his word cards (i.e., a set of Chinese characters) in exactly the same way in which the model sentence is constructed. The objective is twofold: (i) to ascertain whether the patient has any visual problem short of dyslexia and (ii) to see if the patient can coordinate the three modalities (visual, auditory, and manual), a task that is the easiest in the entire procedure, to perform the task. Such a task, of course, demands a high level of cortical functions which is only tacitly at work. A separate task to distinguish (i) from (ii) would be to use a set of symbols the subject does not know (such as Japanese Kana for a Chinese subject who knows nothing about the Japanese writing system) and ask the subject to do the same. (In fact, we did this at the end to determine L.K.‘s memory impairment.) The second step is similar to the previous step except that no auditory stimulus is added; that is, the model sentence is not read to the patient. In so doing, the subject has now only two modalities to work with, viz., visual and manual. If this task makes a difference in the outcome, it means that the auditory reinforcement was a significant help to the patient, which is to say that something is wrong with the visual mode; if not, the auditory reinforcement was a redundant measure. On the other hand, the visual mode could be redundant, which if removed may
214
Journal
of Neurolinguistics,
Volume
2, Number
2 (1986)
or may not make a difference in performance. The third step is to test this redundancy, because the subject now has only the auditory and manual mode to work with; any difference in the outcome, however, would mean that the subject needed the visual stimulus as well for processing (i.e., perceiving the symbols and then putting them in the right order) the linguistic data, which is to say that something is wrong with the auditory processing mechanisms. The last step, which is the most difficult of all, requires that the subject’s memory be intact and, therefore, demands an extremely high level of cortical functions. It is this level of information-processing that reveals the grammatical properties at the conceptual level most. Before we report the results of neurolinguistic testings given to L.K., we believe a general description of his neuropsychological conditions is in order, as we also ran a few neuropsychological tests including the following: Reading, Writing, Facial Recognition, Repetition, Drawing, Calculation, Geometric Figure Recognition, Finger Recognition. The findings are summarized below: His reading ability was limited to single characters. For instance, niu ( q) ‘cattle’, ya (t@ ) ‘duck’, chi ($$$ ) ‘chicken’, kou ( 36~) ‘dog’ were all read perfectly. However, when reading involved sentences of three or more characters each, the speed with which he read them slowed down as the number of characters increased. For instance, he read shei yao chic tim hua ($$3Z$@ g% ) ‘Who is to answer the phone’ and wo yao huo sia ch’ii ( W!$& 7;s) ‘I want to live on’ extremely slowly; when it came to nan hai la nii hai te t’ou fa (J$Jg j$ && mjj$ g ) ‘A boy pulled a girl’s hair’, he not only read it very slowly, he had to repeat the sentence from the beginning whenever he stumbled over a word in order to complete the reading process. This defect in reading seems to suggest that his memory span could not sustain the interruption of an error, an interesting point for speculation to which we shall return later. For the time being, and as a result of his reading problems, he was rendered dyslexic. His writing, on the other hand, was extremely poor. For instance, he could write his own name but was unable to write anything else to dictation. He could also write his own home address but failed to write down the above sentences to dictation. He could recognize a face by identifying its components but could not properly identify fingers (such as left index or right thumb) because of his left-right disorientation. Repetition (i.e., mimicry) of single words was possible (the same as his reading ability, which we believe was no coincidence) but became an impossible task when the target sentence was three or more than three-word long. For instance, he was completely lost when asked to repeat siang chiao hen hao ch’ih ( BE jE&et> ‘Bananas are very delicious’ which was said to him. But the interesting point is that
Agrammatism
and Conduction
Aphasia:
A Chinese
Case
215
when told to repeat siuo hai k’u le ( /J\gTT) ‘The child is crying’, he repeated it as siung chzizo hen hao ch’ih which he had failed to repeat a couple of minutes ago, a phenomenon of (auditory) sensory perseveration known as palinacousis. Again, this is another significant manifestation of his memory impairment. Drawing of a clock proved to be too difficult; nor was he able to draw the shape of a house. Calculation, as expected, was extremely poor. For instance, on the first day of his test, about one week post onset, he could only add 1 t 1 as 2; deduction, multiplication, not to mention division were all out of the question. However, on the second day, he was able to add 2 t 4 as 6 but could not add 5 + 1. In contrast, he did fairly well on his geometric figure recognition. What do the findings of the above neuropsychological tests tell us about L.K.‘s language behavior (i.e., his neurolinguistic condition)? A straightforward answer is “nothing,” unless we look further into the impairment of his grammatical properties. Since we tested him by following the procedure for neurolinguistic purposes, the results can now be provided here so that the impairment of his grammatical properties may also be ascertained. First Step: Visual and Auditory (1)
(2)
Model Phrase:
hsihuan( gg&)
1st trial:
‘to like’ L.K. did it ahight but read it to himself while copying it (i.e., arranging the cards).
Model Phrase: 1st trial:
(3)
Model Phrase: 1st trial:
(4)
Model Sentence: 1st trial:
(5)
Model Phrase: 1st trial:
(6)
Stimulations
Model Sentence: 1st trial: 2nd trial: 3rd trial:
tian shi ( tfijj ) ‘television’ L.K. did it with no problem but, again, read it to himself while copying it. k’an tian shi ( d%iiLH ) ‘to watch TV L.K. had no difficulty with this but, again, read it to himself while copying it. hsi huan k’an tian shih (B& $~Ef$m ‘(I) like watching TV Again, he performed the task without aloud to himself.
difficulty
siang chiao ( &j$$> ‘bananas’ No problem, but with his own verbal reinforcement. siang chiao hen hao ch’ih ($@!&f$J ‘Bananas are very delicious’ zi8@93% h%#swti %~RZ&-oz
by reading it
216
Journal of Neurolinguistics,
Volume
2, Number 2 (1986)
As was mentioned further above, this first step was intended to find out if any deficits of his grammatical ability existed. Note first that as the results show L.K. could copy certain sentences but not certain others: In the case of hsi huan k’un tian shih, he succeeded in copying the model sentence owing probably to the fact that it was gradually built up for him (from two characters to five); his errors showed through in the caseof sting chiao hen hao ch’ih, although he did well when copying one of its immediate constituents, viz., siung chiao. We shall return to this point for a discussion later. Note next that in each of the three (erroneous) trials above he kept the last phrase hue ch’ih (@ %) ‘delicious’ intact while varying the order of the first three characters; the interesting point to observe is that even though @% is incorrectly arranged, the order &a which is the reverse of the correct one he successfully performed earlier was kept constant throughout. L.K. tried three times to match (6), while reading the sentence aloud to himself, and yet failed in each time with no grammatical improvement from one trial to another. (‘1)
Model Sentence: 1st trial:
(8)
Model Sentence: 1st trial: 2nd trial:
(9)
Model Sentence: 1st trial: 2nd trial:
niu rsui ch’ih ts’ao (+ge@) ‘Cattle are feeding on grass’ L.K. failed to do this, although paraphasically) to himself.
he again tried to read it (but
the shih shei fe mao tsu (%@@&%3) ‘Whose hat is this?’ ~~%4izHf!J ilsb%%mw nei ke shih wo fe shou piao ( jJ$Qj$%!&@~) ‘That is my wrist watch 3lHlet&~@m 6rnBR%&@
. Observe now that (7) is the shortest of the three sentences above, yet L.K. failed to copy it, even though both visual and auditory stimuli were provided. The failure means he gave up, after playing with the word cards for a while, rather than making grammatical errors. In contrast, (8) and (9) were tried on but he got (8) right only on the second trial while he got worse on the second trial for (9); this could be explained by the sheer difference in length between (8) and (9) one being six-character long and the other, seven. But such an explanation cannot account for his failure on (7). We shah return to this point further below. (10)
Model Sentence: 1st trial: 2nd trial:
nU ren pu shih NO the (&h;F;&j$;$) ‘Women are not weaklings’ fiims.l&h 2zT#awA
Agrammntim
and Conduction
Aphasia:
A Chinese Case
217
In this case, L.K.‘s grammatical errors got worse, as he did both trials extremely poorly; not only was he unable to construct the target sentence using the appropriate word cards, when the auditory stimulus was also provided once in the first trial, the result came out wrong too, when the auditory stimulus was provided twice in the second trial. It seems then that the length of a sentence may have something to do with L.K.‘s difficulty of grammar, in spite of (7) alluded to above. To this lengthdependent hypothesis of agrammatism, we shall return later. Second Step: Visual Stimulation only (11)
Model Sentence: 1st trial: 2nd trial:
(12)
Model Sentence: 1st trial: 2ndtrial:
(13)
Model Sentence: 1st trial: 2nd trial:
(14)
Model Sentence: 1st trial: 2nd trial:
(15)
Model Sentence: 1st trial:
(16)
Model Sentence: 1st trial:
niu fsui ch’ih fs’uo (+aem ‘Cattle are feeding on grass’ zz the ski/~ shei fe muo tsfd (z g%&J@F) ‘Whose hat is this? ~gfp& . nei kishih “, te shou piao (W@@~~~$% ‘That is my wrist watch
)
iiiz$zEE sheiyuo chic tima burr (;@@&a%) ‘Who is to answer the phone’ #i%%ZS% s?@iass~ woyuo hue sti ch’ii (%%%T%) ‘I want to live on’ 3%&m& nun hui Iu nii hui te t’ou fu (%~?&a&%@@%) ‘A boy pulled a girl’s hair iwzE.frms%
Observe that (11) is the same as (7), (12) the same as (8), and (13) the same as (9). This is because we are now in the second step of the procedure where L.K. was given only the visual stimulus without any auditory reinforcement from the experimenter, so as to see if the absence of the auditory stimulus would make any difference in his grammatical ability that is impaired. Although we did anticipate that it would make a difference, we were surprised to note the difference between (11) and (7). For one thing, L.K. simply gave up in the first step of the procedure on (7), as was mentioned above, but he tried on (11) twice in the second step, albeit wrong on both trials. For another thing, it should be pointed out that the ungrammaticality in the two trials of (11) lies in the position of the word tsai (& ) ‘aspect maker’ (which is a function word, rather than a content word) and that
218
Journal
of Neurolinguistics,
Volume 2,
Number
2 (1986)
the two ungrammaticalsentencesalso differ from eachother in terms of the function word. This pattern of errors resemblesthe errors he made on (6) above where the function word hen (@ ) ‘very’ (an adverb) was misplaced. We shall have more to say in this respect. Observe next that (14) (15), and (16) were the sentencesL.K. wasaskedto read earlier; he read slowly, as we noted above. But in this second step, except for (14) which he erred on the first trial, (15) and (16) were all correctly copied without any auditory reinforcement on the first trial. This raisesan interesting point; that is, if L.K. could read a sentence(no matter how slowly) without any mistake, he could probably copy it correctly, a significant point that will be discussedlater. Third Step: Only Auditory Stimulus wasprovided (17)
(18)
(19)
(20)
Model Sentence:
siang chiao
1st trial: 2nd trial: 3rd trial:
‘Baranas are very delicious’ %$5RB&3oz %?i%EwJz #iG!SCfkFE
hen hao ch ‘ih ( %%8?
6%
Model Sentence: the shih shei te moo tsu ( &&$&J@$) ‘Whose hat is this?’ 1st trial: BZf$&TiYJ 2nd trial: ~gi&fgl@~ Model Sentence nei ke shih wo te shou piao (#$#j&&&JZJ%$ ) ‘That is my wrist watch’ 1st trial: #$@j (L.K. gave up after the first two characters) 2nd trial: $Qm (He gave up againfor the rest) Model Sentence:
nii ren pu shih rue the (&A&
1st trial:
‘Women are not weaklings’ L.K. could not perform this task and simply gave up.
g$fjjj%)
Note that (17) is the same as (6), except for the procedure in that (I 7) was tested without the visual stimulation. Of interest, then, is the fact that the grammatical errors in both casesare quite similar with the exception of the second trial in each step; that is, the adverb hen (jR) wasmisplacedin (6) but was correctly placed in (17). But also of great interest is the non-improvement of (18) and (19) which had already been tested, respectively, in the first step (8) and (9) and in the second step (12) and (I 3). That L.K. also gave up on (20) which is the sameas(10) but which he did poorly for the first step of testing is, too, a very significant point of how his grammatical ability was impaired. We shall return to this point in the discussionlater.
Agrammatism
Fourth Step: No Stimulation (21)
Target Sentence: 1st trial: 2nd trial: 3rd trial: 4th trial:
and Conduftion
Aphasia:
A Chinese
Case
219
provided
fu chiu i ch’i lui ch’ung ke ( ~%-&$DE@ ‘We all come to sing a song together’ I%&~E%-A --RR*il!xW %AR&%n-%As?-&!JEW
)
Here, L.K. was simply given a set of word cards and instructed to construct a sentenceout of the cards in his hands;there wasno model sentenceto copy from nor was there any auditory input for him. The results were disastrous,even though he had four trials. DISCUSSION Several things have been pointed out in the CASE DESCRIPTION above which now require our closer attention. These are briefly recapitulated here for easy reference: A. Difficulties with memory capacity. B. Build-up and grammatical errors. C. Length-dependent hypothesis. D. Length-dependent hypothesis and the nature of function words. E. Reading and copying. F. The complexity of sentences. For the first point, we mentioned that L.K. seemedto have difficulties with his memory capacity, because in his reading of a long sentence (consisting actually of eight characters) his memory span could not sustain the interruption of an error. We would like to attribute this deficit to the phenomenon of perseveration alluded to earlier. According to the current view in the literature, this condition is a kind of auditory perseveration called palinacousisthat can be said to be reflected in the oral mode (i.e., speech). But we suspectthat L.K. probably had both palinacousis and palinopsia(i.e., visual perservation). In the case of palinacousis,we think that it can be thought of asfollows: A set of auditory neurons is excited by a stimulus but the excitation cannot be momentarily inhibited in order to give way to the excitation to be caused by a different stimulus. As a result, the previous impulsesperseverate. In the caseof L.K., recall that he heard the sentencesiung chiuo hen hao ch’ih which he could not repeat but when he heard siuo hai k’u le, and was supposedto repeat it, an act that required the inhibition of the previously excited auditory neurons first, in order to get the next excitation for the production of siao hai k’u le. it came out as siung
220
Journal
of Neuroliiguistics,
Volume
2, Number
2 (1986)
chiao hen hao ch’ih; this perseveration means that L.K. lacked the rapid inhibition to suppress the excitation of the auditory neurons for siang chiao hen hao ch’ih,-so when the new stimulus siao hai kit le was given, it failed to excite his auditory neurons and, therefore, the previous impulses caused by siang chiao hen hao ch’ih perseverated. Although his palinacousis is a form of memory impairment, it has no bearing on his reading problem. We would like to propose that the fact that he read each sentence very slowly was in effect caused by the lack of rapid inhibition in his brain due to the infarction, in the parieto-occipital region, except that in the case of his reading problem, the lack of rapid inhibition was not so severe as to cause actual perseveration. In other words, he had to wait for the excitation of each set of neurons to be inhibited each time before he could move on to read from one character to the next. And this hypothesis is plausible because when he stumbled over a word, which was read erroneously by him, he knew it was wrong but he had no way of going back to the previous word on the ground that the excitation of the neurons for the visually perceived word (which was erroneous) could not be inhibited right away. That is why he had to go back to the beginning of nun hai la nii hai te t’ou fa each time he made a mistake in reading in order to complete the reading process. It is this deficit that we suspected that L.K. also had palinopsia. This explanation can be further confirmed in the first step of our testings in which he read every phrase and sentence to himself while copying it. That is, he had to read to himself what he saw for two purposes: (i) He wanted to use his feedback to make sure that he was copying (i.e., arranging) the word cards right, although the task required that he need only use his visual and manual mode (the latter being bare minimum); and (ii) since in the first step both visual and auditory stimulation were provided, his feedback could very well serve as a reinforcement of the auditory stimulation to doublecheck what he heard. The above explanation leads us to the second point whereby L.K. performed better when there were build-ups for his task. Recall that (1) through (4) of the first step in the procedure were presented to him in such a way that the ultimate sentence hsi huan k’an tian shih was broken up in small “chunks” (in linguistics, they are called Immediate Constituents) first and then “spoon-fed to him,” that is, starting with a two-character construction (because we knew he could read single characters well) and then gradually adding an immediate constituent to make up a larger construction for him to copy, until reaching the ultimate sentence. Thus, he did them perfectly, albeit still reading each stimulus to himself. But when this “spoon-feeding” process was not taken, that is, when we jumped from siang chiao in (5) with which he had no problem, to (6) by adding three characters all of a sudden, his grammatical errors surfaced. Why? The following is our explanation. We believe that his palinacousis and palinopsia had something to do with the
Agrammatism
and Conduction
Aphasia:
A Chinese
Case
221
grammatical errors he produced, when copying (6). But we also believe that there was something deeper that was wrong which was at the conceptual level, the two conditions (palinacousis and palinopsia) being impairments of the process from conceptualization to production. This deeper problem is what we call agrammatism which is a form of impairment at the conceptual level - in linguistics, we may linken it to the content plane rather than the conceptual level which is more familiar to cognitive psychologists or psycholinguists. But this problem - agrammatism will not become clear until the entire procedure of our testings is fully explained. So, for the time being, let us say that due to his memory impairment which at times manifested itself in the forms of palinacousis and palinopsia L.K. could not process without any build-ups (i.e., rehearsal) a phrase or sentence that was more than three-character long, because it exceeded his memory span and the excessive materials began to interfere with his retrieval process. In the case of his problem with (6), notice that he had no difficulty with (5), siung chiao, which fell in line with his memory capacity, but when three more characters were added to it, becoming (6) without, say, the intermittent rehearsals, like hao ch’ih ‘delicious’ and hen hao ch’ih ‘very delicious’, he got it wrong three times, even at the expense of siang chiao which he had had no problem with as (5). But notice also that in (6), as we mentioned earlier, he kept hao ch’ih right throughout, although he erred on sting chiao by reversing it to chiao siang and kept the wrong order constant. This trade-off is what we meant by intereference. That is, since his memory capacity could handle two-character phrases well, when (6) was presented to him, he was preoccupied with the new information, which was the added materials, hen hao ch’ih, but lost control of the old information, which was siang chiao, although he had had it right in (5) but then he could only handle two-character phrases well, so he had hao ch’ih (which is the head of the phrase, hen hao ch’ih) while losing control of hen (which is the modifier of the phrase, hen hao ch’ih) by mixing it with chiao Gang which he also erred. It is this error, chiao siang, that perseverated; that is, since hen hao ch’ih got in the way and interfered with his processing of siang chiao which came out as chiao siang, he could no longer inhibit the excitation of the neurons for the erroneous chiao siang which served as evidence for his palinopsia, even though he tried to read each trial as siang chiao hen hao ch ‘ih. The above discussion seems to point to the fact that L.K. had problem with length which is the third topic for this section. But when (7) is taken into consideration, it immediately becomes clear that his deficits, i.e.. agrammatism, were much more complicated than just the mere explanation relying on a lengthdependent hypothesis. Recall that we have pointed out that (7) is shorter than (8) and (9) or even (6), for that matter, which we just discussed in great detail. What is important here, then, is not the length involved but the complexity of rhe
222
Journal
of Neurolinguistics,
Volume
2, Number
2 (1986)
syntactico-semantic properties which characterize Chinese grammar. Note first that unlike (6) where there is no verb involved, (7) has a co-verb, tsui, and a verb-object construction, ch’ih ts’ao ‘(lit.) to eat grass’. It is this combination, co-verb + verb-object construction, that posed one of the greatest difficulties for L.K., as will be evidenced by (11) later. For the time being, bear in mind that the length-dependent hypothesis could account for only part of his problems, but that this hypothesis is basedon his obvious deficits of palinopsiaand palinacousispointed out earlier. Meanwhile, let us move on to (8) and (9). Here, ostensibly the length-dependent hypothesis seemsto prevail, because(9) is longer than (8) and L.K. got both trials wrong for (9) but at least got the second trial right for (8). Otherwise, the two sentencesare structured in exactly the same way, asmay be shown in the figure below: (8)
the
shih
shei
(9)
(nei ke)
shih
wo
te
Pronoun
Particle
Demonstrative Noun Coupla I (Dem.NominalPhrase)
te
,,1,,.,
ma0 shou
Nominal
tsu
piao Phrase
I EndocentricConstruction
Subject I
Predicate I Sent&e
Figure 2. Sentence(8) and Sentence(9) analyzed in terms of ICs. But notice that in both sentences,the errors L.K. made were almost identical; that is, in (8) the shih shei te mao tsu he moved the possessiveparticle te to the very end, an error that he repeated (or perseverated) in (9) on both trials, although he was able to correct this error in the second trial of (8). However, there was an additional grammatical error in (9) which L.K. did not commit in (8). This error is the placement of shih ‘copula’ in the first trial of (9), an error that he wasable to correct on the second trial of (9). In other words, while L.K. made only one grammatical error on the first trial of (8) (which he corrected on the second trial), he made two errors on the first trial of (9) but corrected only one of them on the secondtrial. This is where the fourth topic can be raised. Here, we are now concerned with the interrelationships of length and the nature
Agrammatism
and Condqxion
Aphasia:
A Chinese
Case
223
of function words. Granted that the length-dependent hypothesis could explain in part the difference of easiness between (8) and (9) as far as L.K.‘s grammatical errors were concerned, it could not fully explain why he erred repeatedly on te and shih later and not on something else, like shou piao ‘wrist watch’ (which is a noun phrase) or the position of the and nei (which are demonstratives). So, there has got to be something else besides the length factor, the control of which L.K. seemed to have lost, something that has to do with the complexity of function words. A function word in Mandarin Chinese is a morpheme that marks an aspect, like tsui in (7) mentioned above, or the relationship between one noun (or pronoun) and another noun (or noun phrase, like te) or the linkage between subject and predicate, viz., shih ‘the copula’, and so on. It seems to us then that L.K. also had difficulties with the syntactico-semantic properties of function words, as these words play not only structural role in Chinese syntax but also reveal subtle differences in meaning depending on where they occur. The following sentences, though they were not tested on L.K., would be totally impossible for him to handle: (21)
(22)
wo kei ni two
fan hao ma? (T&$i$?ylp#$E~N~
?)
‘1s it alright if I cook for you?’ [(lit.) I give you make rice alright Question Marker?] wo kei ni two
te fan hao ma?
(We
###~&&il~
?)
‘How is the rice that I cooked for you?’ [(lit.) I give you make Possessive Particle rice alright Question Marker?]
(23)
(24)
wo kci ni tefhn
two
hao le ma? (~,J#J~#&~~~?)
‘Have you cooked the rice that I gave you?’ [(lit.) I give you Possessive Particle rice cook alright Aspect Marker Question Marker?] wo tsuo kei ni te fan hao ma?
(s#
~Jx &J&
&I!!$
?)
‘How is the rice that I cooked and gave to you?’ [(lit.) I make and give you PossessiveParticle rice alright Question Marker?]
(25)
wo two
fbn kei ni hao ma? (~&!&$$@&I!$
?)
‘Is it alright if I cook rice and give it to you?’ [(lit.) I make rice give you alright Question Marker?]
Observe that (21) and (25) differ only in the positions of kei ni ‘to give you’ and tsuo fan ‘to cook rice (actually, to cook a meal)’ The. English glossesare intended to give the subtle differences in meaningsbetween them. Likewise, the difference between (22) and (24) lies in the position of kei ni te ‘that I give you’, which is interpolated between tsuo ‘to make’ and fan ‘rice’ in (24). but is partially overlapped by fsuo fun in (22). Again, the English glossesoffer the subtle differences in meaningsbetween (22) and (24) where there is no increaseof words. But when the position of kei ni te is completely separatedfrom ts~o fan (whose order is now reversed asfan tsuo) coupled with the addition of the aspect marker le, as in (23)
224
Journal
of Neurolinguistics,
Volume
2, Number
2 (1986)
the meaing as shown by the English gloss changes drastically. Note that in those examples, we have been concerned with the word order (of function words like DUO [a co-verb in (24) but a main verb in the rest], kei [a co-verb in (21) (22) but a main verb in the rest] , and te [a possessive particle]) in relation to ni ‘you’ (a content word) and fan ‘rice’ (another content word) and the role played by le ‘aspect marker’ and ma ‘question marker’, both being function words, in the determination of meanings in the sentences. Note further that while in Mandarin the question marker remains constant (in one position) the English equivalents change the interrogative structure three times, from Copula-Pronoun to AuxiliarPronoun and then to How-Copula, owing probably to the fact that in Mandarin the changes in meaning (and this is true of all Chinese languages) are determined not so much by the changes of lexical items (as in English) as by the changes in positions of the lexical items already employed. And this is the crux of Chinese grammar and we believe that L.K. has lost control of some of these properties, as may be demonstrated by his grammatical errors in (8) and (9) a loss that will be further evidenced when we come to Step Two, Step Three, and Step Four. In the second step, for instance, when the auditory reinforcement was cut off, he managed to copy (11) [which is the same as (7) but which he failed to do in the first stepJ yet he had difficulty with the co-verb tsai that also functions as an aspect marker. Notice that he kept niu ch’ih ts’ao ‘cattle eat grass’ but did not know where to place the function word tsai in the sentence, even though the visual stimulation was provided, that is, he had a model to copy from, albeit without the auditory reinforcement from the experimenter. Since (11) is shorter than (12) and the rest, we cannot rely on the length-dependent hypothesis to explain this kind of difficulty. We would, therefore, like to propose that in addition to palinacousis and palinopsia, on the basis of which we formulated the length-dependent hypothesis, L.K. had the disintegration of the grammatical properties of Mandarin to a certain degree, especially that part of Chinese grammatical properties which has to do with the complex interrelations function words and word order have in determining the subtleties of meanings. In the case of (1 l), we believe L.K. probably knew approximately what it meant, because he got niu ch’ih ts’ao right on both trials, which must have meant something to him; what he failed to determine was, therefore, the subtle difference in meaning between niu ch’ih ts’ao, which he understood and copied correctly, and niu tsai ch’ih ts’ao ‘cattle are feeding on grass’, which he could not get at, because he lost control of where tsai was supposed to go. In Luria’s terminology, then, he might have linked it to a disorder of the predicative function and to the disintegration of sentence schemata (Maruszewski 1975: 108). What is of interest in this second step is the observation that in (12) and (13) L-K. repeated the same grammatical error that he committed in (8) and (9) respectively. That is, he lost control of where to place the function word te in both
Agmmmptism
and Conduction
Aphasia:
A Chinese
Case
225
(12) and (13) and of where the copula shih was supposed to go in (13), although there is a slight difference between the second trial of (9) and the second trial of (13) regarding the position of fe, even though in both trials it was erroneously placed. Recall then that (8) and (9) were tested in the first step while (12) and (13) were tested in the second step, which is to say that the former had both the visual and auditory stimulations but the latter had only the visual stimulation. We believe that the difference just mentioned, coupled with the erroneous placement of shih, is attributable to the difference between Step One and Step Two of the procedure. In other words, we believe that L.K. got worse when the auditory reinforcement was cut off, as may be evidenced by his errors in (13) which were worse than his errors in (9) although the errors in both (12) and (8) were exactly the same. To support the above claim, which is that as far as L.K. was concerned the addition of an auditory stimulation from the experimenter was not redundant but a significant reinforcement to help him copy the model sentences, we can cite his error in (14) which had not been tested on him (in Step 1). In this task, he made the error of placing the main verb chic ‘to connect’ (@) at the beginning. Here, the typical grammatical properties consist of a cluster of two verbs, yao. chic ‘(lit.) wanting to connect’ (Kg), whose order cannot be reversed. It is this syntacticosemantic property that L.K. apparently lost, albeit temporarily, because he restored it on the second trial (by again relying on his own feedback). But the ability to restore the grammatical property may be attributable to the fact that (14), though not tested in Step 1 of the procedure, had been used for testing his reading. This leads us to the fifth point (E) about reading and copying with respect to L.K.‘s agrammatism. Note that all three sentences, (14), (15) and (16), had been employed for the purpose of testing whether or not L.K. could read sentences of three or more characters each, and that as we have already mentioned he did read them but very slowly. We also thought of palinopsia as a reason for his reading problem. We wish to claim now that the fact that he had been shown the three sentences and could read them, albeit extremely slowly, helped his task in the second step, because he arranged all of them correctly, even though with a problem initially on (14). How can we explain this kind of facilitation neurolinguistically? Well, we believe that although L.K. did have palinopsia which hindered his reading, and we also speculated that in line with Luria’s thinking L.K. probably had a disorder of the predicative function and a disintegration of sentence schemata, the visual perseveration may have helped him, serving as a rehearsal, to restore some of the disordered predicative function in his long-term memory, at least temporarily within the duration of his testings, and therefore helped him recognize the sentence schemata of (14) (15) and (16), so much so that when those model sentences were presented to him for copying, albeit without the auditory reinforcement, he could
226
Joumd
of Neurolinguistics,
Volume
2, Number
2 (1986)
recognize them and thus had no difficulty - except for the first trial of (14) - of copying them. Further evidence of whether or not reading, in which palinopsia occurs, and copying of visual stimuli, in which motor skills are minimized but which require some sort of mental activities, can facilitate each other is obviously needed before any conclusive statement can be made. For the time being, however, we feel comfortable to say that (i), L.K.‘s disastrous performance of (lo), which had not been shown to him for reading, and (ii) the ease with which he performed (I), (2), (3), and (4) because hsi hum k’un tian shih ‘(I) like watching TV’ was gradually built up for him, a step comparable to rehearsal, are sufficient evidence to at least point to the possibility that reading, no matter how slowly the speed, does serve as a rehearsal if conducted prior to sentence copying (or construction) to a task that requires an aphasic patient who has agrammatism to perform sentence constructions. The most vivid demonstration that L.K. had a severe case of agrammatism in the sense we defined it may be observed in his performances of (17), (18), (19), and (20) for the third step and (21) for the fourth step. Recall that in the third step there was no visual stimulation given; only auditory stimuli were provided. Observe now that here his scores for (17) and (18) were almost identical to those for (6) and (12), respectively, which is to say that his deficits were at the conceptual level rather than at the production level, the reason being that his grammatical errors did not seem to improve much whether he was tested in the first, second, or third step. Notice that (6), which was tested in the first step, and (17), which was tested in the third step, using the model sentence siang chino hen hao ch’ih, produced more or less the same errors except for their respective second trial. Is this a coincidence or the result of carry-over, viz., perseveration? We do not think that either one could explain satisfactorily the identical results; rather, we think that the best explanation is to say that his agrammatism was a central disorder bilaterally represented at the conceptual level, because his motor skills for production through the manual mode were minimized. This theoretical construct is confirmed by the fact that the model sentence the shih shei fe mao tsu ‘Whose hat is this?’ was tested three times in (8) for the first step, in (12) for the second step, and in (18) for the third step, and yet each time he made the same mistake on the first trial and then corrected it on the second trial. We should further point out that when the model sentence nii ren pu shih mo the ‘Women are not weaklings’ was tested in (IO) for the first step, the results were disastrous; when il was tested for the third step in (20), L.K. simply gave up. This is further evidence of his central deficits at the conceptual level, because nii ren pu shih rue the was not rehearsed in his reading. Better still, mention should be made that the model sentence nei ke shih wo te shou piao ‘That is my wrist watch’ was also tested lhree times in (9) for the first step, in (13) for the second step, and in (19) for the third step. and yet his results got steadily worse from one testing to the
Agcammatism
and Conduction
Aphasia:
A Chinese
Case
227
next; he eventually gave up for the third step, after completing only the first two characters. The most interesting point concerning L.K.‘s agrammatism - and we expect this to be true of all Chinese agrammatics - is that when no stimulation was provided, except for a set of word cards which any normal subject could arrange into a grammatical sentence, for example, ta chia i ch’i lai ch’ang ke ‘We all come to sing a song together’, L.K. was completely lost, even though he attempted to do so four times. Notice that in this test the only phrasehe got right throughout wasch’ang ke ‘to sing a song’ (which is a verb-object construction); the rest, except for i ch’i ‘together’ which was correct on the fourth trial, were all so battered that nothing sensiblecould be made out of them. This leavesus with the final point (D) where we must now be concerned with the complexity of sentence structures and agrammatism. To explain this point, let us recapitualate the model sentencesthat were usedin the testings. O(4)
hsi huan
k’an tian shih
SC& 1 [build-ups]
(6)
(7)
**@)
niu tsai ch’ih tsbo BE% B [repeated as (11) for the second
*(lo)
**(14)
‘Bananas
delicious’
are very
‘Cattle
are feeding
on grass’
step] ‘Whose
hat is this?’
for the 2nd step and as (18)
for the 3rd step]
nei ke shih wo te shou piao ‘That is my wrist watch’ %BS wKJ?F ek! [repeated as (13) for the 2nd step and as (19) for the 3rd step] nii ren pu shih rue the *A&aFE423 [repeated as (20) for the third sheiyao
chic tian hua
q% 42 )g t [also O(15)
TV’
step]
the shih shei te mao tsu gg &$Jl@ 3 as (12)
‘(I) like watching
@A
siang chiao hen hao ch’ih z$ i% G! PJ ti [repeated as (17) for the third
[repeated (9)
t
‘Women
are not weaklings’
step] ‘Who is to answer
the phone’
$3
used for his reading]
wo vao hue sia ch’ii
‘I want
to live on’
pulled
aG32vF & [also used for his reading]
“(16)
nalt hoi la rlii hoi te t’ou fti %f%E-k&F-d@% [also used for his reading]
‘A boy
721)
ta chia i c/r ‘i lai C/I Zng kc A Q-&?R “?I a
‘We all come
a girl’s hair’
to sing a song togcthcr’
228
Journal of Neuroliiguistics,
Volume 2, Number 2 (1986)
Of the sentencesemployed two L.K. did poorly, which are starred, viz., (10) and (21); but he had no problem with three, i.e., (4), (15), and (16), which are circled; two were erred and then corrected, viz., (8) and (14), which are doublestarred; the rest, namely, (6), (7), and (9) were erred but could not be corrected by him. What does all this tell us about L.K.‘s agrammatism? Put differently, is there any correlation between his deficits and the grammatical properties of Chinese? The answeris a straightforward “Yes.” First, we would like to suggest that the more complex sentenceswere more difficult for him to handle. The sentencesthat he failed completely are (10) and (21) whosestructures are presentedbelow: (10)
nii Nominal ,
,
ren
PU
Phrase ,
“9;:,1,%ula
shih ,
Subject
the
NO
NOT
,
Phr,ase
Predicate Sentence
(21)
ta
chia
Nominal ‘r
Phrase
i Numberal ‘....,I
Subject
ch’i
lai
ch ‘ang
Verb
Co-Verb
Verb
,
ke
[erb;kre.r_““In
Predicate
I
Sentence
Figure 3. Sentence(10) and Sentence(21) analyzed in terms of ICs. The grammatical complexity of (10) is seenin the verbal phrase which is made up of the copula shih (g ) and the negative pu (q), both being function words; they are extremely productive in that the former can be concatenated with
Agrammatism
and Conduction
Aphasia:
A Chinese
Case
229
practically any noun or nominal phrase while the latter, with any verb, not just the copula. Perhaps, as a result of or even owing to such common concatenations, L.K. kept pu shih ($$S) ‘are not’ together on both trials of (lo), even though he failed to place it in the appropriate position. Put differently, the patient now had to reckon with the position of a phrase, viz., Negative + Copula, in the sentence, rather than that of a word, viz., Copula alone which he also had some trouble with in (8) and (9), other things being equal. It is probably this complexity that caused L.K. to make more grammatical errors in (10) and to give up on (20). In the case of (21), it should be noted that the sentence is even much more complex than (10) in terms of both length and structure, as can be seen from Figure 2. Note that (21) has a cluster of verbs in the predicate coupled with an adverbial phrase i ch’i (-&Z) ‘together’ which makes the predication a highly complex structure. If (10) was too complex for L.K. to handle with both visual or auditory stimulation (i.e., help), it is only natural that he could not construct the target sentence expected of him in (21) without any stimulation. If we take (10) and (21) to be the upper end of complexity in the sentences recapitulated above, we can say that the phrases of each sentence, such as siung chiao ‘bananas’ and tian shih ‘TV’ which are nominal constructions, or hsi huan ‘like to’ in (4) which was tested as (I), represent the lower end of complexity for L.K.‘s tasks. When such complex sentences were rehearsed, as in (4), (16) or to some extent (8) and (14) his tasks seemed greatly facilitated. CONCLUSION In this article, we have been discussing a case of agrammatism in Chinese whereby the paient, L.K., was diagnosed to have conduction aphasia. Because of the symptoms of grammatical errors in his sentence constructions, we were convinced that L.K. had agrammatism, a disorder of the schema of sentence construction at the conceptual level which is believed to be centrally and bilaterally represented. We then proceeded to examine the extent of his agrammatism by employing a method that is based on the written symbols which followed a procedure consisting of four steps. In so doing, the following findings were obtained. First, we found that L.K.‘s agrammatism was length-dependent in that he made more grammatical errors as the target sentences became longer for him to handle. However, this finding must be qualified. That is, his agrammatism was also sensitive to two factors: (a) the grammatical complexity of the sentence concerned; and (b) whether or not the sentence to be tested was gradually built up for him or shown to him suddenly. In the former, we found that even when the sentence being tested on him was shorter, he would have difficulty, as in niu tsai ch’ich tsao, if its complexity was great enough due to its grammatical structure of function word +
230
Journal
of Ncurolinguistics,
Volume
2, Number
2 (1986)
verb construction; in the latter, however, the difficulty could be overcome, if and when the sentence had been “rehearsed” through build-ups, as in (4) hsi huan k’an tian shih, or reading, as in (16) nan hai la nii hai fe f ‘ou fa, whereby his agrammatism was somehow suppressed. We have speculated on this phenomenon as the result of facilitation in his memory capacity, a phenomenon that is to some extent related to the next finding which is his palinopsia and palinacousis. We found that L.K. had the auditory perseveration in that he was supposed to repeat siao hai k’u le ‘The child is crying’ but instead he repeated it as Gang chiao hen hao ch’ih ‘Because are very delicious’ which he had failed to do so a few minutes earlier. We also found that he read every sentence very slowly as if he had to think for a while before he could continue from one word to the next. We came to the conclusion that this phenomenon was due to his visual perseveration because he lacked the ability of rapid inhibition as part of his memory impairment. In spite of this form of impairment, he could overcome it if and only if a stimulus had been “rehearsed” ahead of time, through reading or some other means, as in (14) sheiyao chic tian hua and (8) the shih shei te mao tsu which he could correct on a second However, this facilitation through rehearsal to overcome his memory trial. impairment was rather limited, as may be evidenced by his failures on (6) (7) and (9) because of the complexities of these sentences. Our third finding is that of the four steps in the procedure the fourth step was most difficult for L.K. We believe that this is true of all agrammatics regardless of their aphasia types. To some extent, it could also be said that the first step, when both visual and auditory stimulations were provided, was the easiest of all. The second and third steps, where only one stimulation, either visual or auditory, was provided, were such that we think the second step was slightly easier for him, which is to say that the auditory mode was more damaged than the visual mode. This is supported by the fact that while his palinacousis manifested itself vividly, his palinopsia was only implicitly evident. In light of the aforementioned findings, we feel quite comfortable and justified to imply that the concept of agrammatism is now in need of a new definition. This was presented by Peng eC al. in the 8th ICU Conference on Neurolinguistics, November 22, 1986 entitled “Agrammatism Redefined.” For the time being, we would like to conclude that L.K.‘s agrammatism was a form of impairment that was not so much due to the breakdown of the grammatical system in his LTM as due to the inaccessibility of the grammatical information needed; the concept of “inaccessibility” in the present conext may be taken to mean the problem of retrieval of stored information, rather than the breakdown of the stored information itself. In this sense, then, the lack of rapid inhibition mentioned above and the concept of inaccessibility form a cause and effect relationship. To corroborate this hypothesis, we ran a test which required only the activation
Agammatism
and Condiiction
Aphasia:
A Chinese
Case
231
of L.K.‘s STM without having to retrieve any information from his LTM. We asked him to copy Japanese Kana (three lines of the syilabaries, both kata kana and hira gana) which he saw for the first time in his life and, of course, could not read by using the second step of the procedure. Although it took him 40 seconds or so for each line, with one minor exception he did the rest correctly. This test supports our third finding above and suggests that his agrammatism stemmed from the impairment of the retrieval system of his LTM and that his STM was relatively intact which could be activated visually or auditorily to somehow make the best of the impaired system for retireving grammatical information - no matter how imperfect the retrieved information was - from his LTM. However, we should add that our patient was more often than not unaware of the ungrammaticality of the sentences he either copied or constructed and appeared content with the erroneous results. This is not tantamount to saying that he had anasognosia; rather, it only meant that he was not aware of each erroneous instance he produced on the spot, because he knew something was seriously wrong with his language behavior, and said so to us, but only in general terms. Finally, we should point out that agrammatism in conduction aphasia is most likely not the same in neurolinguistic terms as that found in other language disorders, say, transcortical (motor) aphasia, where speech may be labored and nonfluent and the patients lack an active attitude and do not initiate speech (Maruszewski 1975: 112). While there may be aphasiologists and/or neuropsychologists who wish to make a distinction between motor agrammatism and paragrammatism (which is conceptual in nature, cf. Goodglass 1973: 187) or sensory agrammatism (in the senses Pick used them (1931) we now wish to call this distinction into question by proposing to re-examine the nature of grammar in aphasic speech and its disturbances. For instance, we have begun to suspect whether agrammatic speech in a motor aphasic represents the sparing of the conceptual and loss of the linguistic phase of speech (Goodglass 1973: 187). Put differently, we are now of the opinion that even in a nonfluent (motor) aphasic if agrammatism shows through the patient may also have conceptual problems, especially when the language involved is an isolating language, such as Chinese, and therefore that the distinction of agrammatism and paragrammatism may just turn out to be unnecessary (or even superfluous) in that case. REFERENCES
Critchley, Macdonald 1970 Aphasiulogy and Other Aspects of Language,London: Edward Arnold. Goodglass,Harold and Sheila Blumstein 1973 Psycholinguistics and Aphasia, Baltimore: The Johns Hopkins
232
Jownd
of Neurolinguisticq
Volume
2, Number
2 (1986)
University Press. Maruszewski, Mariusz 1975 Language Communication and the Brain: A Neuropsychological Study, Warsaw : PWN - Polish Scientific Publishers. (and the Hague: Mouton).