Fluency and Aphasia: A p r a g m a t i c reconsideration ABSTRACT Ihe fluency behaviour of fourteen aDh ' 0PS0MM1NG m a ClaSSificatory concept Vlotheid is by veertien afat' — d i s r a p , i 0 1 1 o r p r e s e r v a t i o n ' * * expressive out™,, „ S e r v a t l o n o f Auencv in anh^.v , . P -a ιwineia en ι frequently in the literature. I n I c M t i? t h " Pressive output which has for J e d the I ^ m e n s i o n of ex- of classificatory schemes in a p h a s l a W * ̂ n U m b e r for example, that at least 75% o f * ^ e t a i · ' 6 ^'aim, can be distinguished along th! H u e n ' T ^ a differentiation is made betweenTh^ E s s e n ^ U y , agrammatic) patient, who n ° " f l u e n t " ^ r i o r , o r halting and effortful w ^ P r 0 d U C t i ° n i s 4 99 / 4 2 38 48 47 37 43 44, j / 4 4 38 54 54 45 49 50. Figure 1 Linkage tree demonstrating degree of similarity between subjects on Fluency Scale V RESULTS Fro C i U f S a V a i l a b l e f 0 r C l u s t e r i " g sets of suitable for the present ^ * * * W ° U l d b e m ° s t l l n ^ e d i a g r a 7 d e n T n V e S t l g a t l 0 n W e r e <*™*>η of a field. FigufesT 2 ' η Ζ 0 8 ' 3 1 1 1 ^ C ° U n t e ^ r t ' * * cluster P r o g r a m r d i r T S e n t ί* ***** ( t r e e ) d i a 8 ^ , ^ s i s o f ^ l t i 1 U f r , f i e l d a c t i v e l y for the '"g of Ss is i n £ t S h 8 ^ ^ h n k a g e d i a g r a m ' ^ group- i e above a p a S ^ ^ T ^ ° f ^ ^ s e c t i o n of the Ζ S ' . T h e strength n f T J * d i a 8 0 ^ 1 line starting next to numerical values in rh , association is represented by t 0 1 0 0 a n d represent corrdat' ^ T ^ a r e S C a l e d 1 correlations from 0,00 to 1,00. Thus, the A . „ — " " " ιιυιπ υ,υυ to S U i d - ^ ^ e T y d s k r i f . J vir *immumkasieafivykings, Vol. 30, 1983 Figure 2 Dendrogram demonstrating degree of similarity between subjects on Fluency Scale V 6, 2 8 12 and 10) showed relatively distinct * I T distinctions will be considered under e a S t o S ^ u p ' A. APPROPRIATE GROUP ttXaiZLSz&SZ R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 01 2) 6 instances of the nonfluent behaviours identified by Table 2 oc- curred predictably in all Ss. S14 showed considerably fewer of these behaviours than other Ss in this group. In feet, both judges commented on her behaviours as being 'too fluent'. In many conversational turns there were no examples of non- fluencies at the expected junctures. This could perhaps be tied to a self-monitoring defect, which will be discussed later. This is illustrated by: S14 T: So you're doing a lot of work at home? P: Well I must. It's the only way. First of all it's easier for me to learn things as I do. ( ) *When they speak to me I ask them. It's a lovely feeling. It really is. It's difficult but it slowly slowly come right. Writing nice- ly. My writing's very well. It improved a lot I think so it makes a I feel so good then I start again then it's lovely. S3 showed, by contrast to S14, instances of revision, repetition and hesitation which, though judged as largely appropriate, were sometimes seen as interfering with communicative flow. This aspect was also noted in Si's sample. Her frequent use of 'er' and occasional word-finding difficulty were felt to in- terfere with the flow of the message though overall fluency was maintained through the discourse, e.g.'s: S3 T: You want to go and see the top man? P: Right right. So um about um He'll give us a letter an answer you see. About a month later had a chat this fellow on the telephone. 'Look what do you think?' He says no I'm sorry we can't do it but I'll send us a letter. He will send us a letter which is three months later. (Note revision or "high level amendment" to use Butter- worth's3 term.) SI P: My son is in er Escom but er it's in Germiston and er he has to er by six o'clock he must er he goes by car but actually he goes afterwards by er train, it's easier you know because er er not so expensive to go by car. S6 (A(b)) warrants special comment. His tie with the other Ss in this group was low (0,53). All aspects of nonfluency were observed by the judges to occur very frequently. This S's rapid rate of speech, however, provides the listener with an overall impression of fluency. Thus judgments of inappropriate were in part offset by this factor, e.g.: S6 T: So you hit your hand? P: Ja Ja Ja Ag you know Lovey. Um you see this one. This look now what ( ) T: The drink P: You see how look but if you ( ) like this one I'll never you'll do the same lovey. I never said you know I promise you I had enough. B. INAPPROPRIATE The severely restricted output of Ss 2 , 4 and 5 made judgments of fluency difficult in terms of the categories suggested on the fluency scale (Table 2). Thus these Ss are grouped together in the cluster analysis, the few possible opportunities for judg- ment indicating that repetitions, false starts and incomplete phrases hindered overall communicative flow, e.g.: S2 T: Why did they have collars round their necks? P: They had something ( ) * Unintelligible utterance. Claire P e n n T: Oh so they had something done to them? P: No one the man who the man who was ( ) The manner of speaking in Ss 7 and 9 (B(a)) also seemed to account for judgments of their fluency being very similar Both of these Ss have a particularly slow rate of speech resulting in lengthy unfilled pauses. Other aspects of nonfluency in these Ss (e.g. interjection and repetitions) were felt to be appropriate and not a hindrance to communicative flow, e.g.: S7 T: Don't your grandchildren go to school? P: No they - they a small. They too small yet. They er er the er oldest one he does to s - - kindergarten you know and er the other one he can't walk yet. S9 T: Do you prefer some programmes to others on TV? P: No. Er - - er - - Sport - 1 like and er - - um er progra and er - - a few programmes a week. The remaining five Ss (B(b)) were all judged to.have many in- stances of inappropriate fluency behaviours in their samples (despite the fact that three of this group could be traditionally classified as 'fluent' Ss). All behaviours delineated by the Scale were noted in all Ss but the following were commonly felt to interfere specifically with communicative flow: inter- jections and repetitions (in four out of five Ss), pauses and word-finding difficulties (in all five Ss) and incomplete phrases and false starts (in three Ss). It appeared that the longer the conversational turn on the part of the S and the higher the frequency of non-fluent behaviours in such a turn the more likely it was that the S would be judged as being inap- propriate with regard to fluency. This was particularly noticeable for S12: T: She lives here? P: No in ei'near Durban. They've got a farm and er I feel there's something funny so I phoned her and er she said. Well she er I'm fine Honey am um I'm not so well but I'm fine but er did you get my letter? So I said No. Anyway um so I waited for the letter and er then the letter that worried me so much because she told me that um that I had flu. Broca's patients, Ss 10 and 11 had relatively impaired syntactic output in relation to other Ss in this group and their nonfluen- cies e.g. repetitions, revisions and interjections were felt in part to be not altogether successful attempts to compensate for this, e.g.: 510 T: How would you change a tyre? P: Well I don't to I would put er a a jack and then I would - put a - wheel up a little bit and then I would get the bolts er in the not in the bolts and - put it in under the - See? 511 T: What did you do? P: Um Talk. Hospital Um Um Um. Two time a week. T: You go there for therapy? P: Yes Um um um yes - um my hand, my leg. Word-finding behaviours seemed to affect fluency judgments of Ss 8 and 13. e.g.s: S8 T: Could you tell me how to change a tyre? P: Yes well 111 know what it is. Obviously and went on but it's hard hard to get something. T: What do you do? P: You know it's difficult - - Um - you know I told all forever in ( ) things like that. Beautiful well unfor- The South African Journal of Communication Disorders, Vol. 30 19S3 R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 01 2) Fluency and Aphasia: A Pragmatic Reconsideration tunately as it went down down and so I was getting useless. But how do I do it? S13 P: It's a sort of thing you wind up and a thing you get to play records not records but it's a records record sounds like a records something. T: A tape reorder? P: No no no not a tape recorder. Um - - T: Not a record player? P: No no no not a record. Not strictly a record player. It's for infants of the age 2 to 6. T: Is it one of those Fisher Price toys? P: Fisher Price. Fisher Price. Yes. Fisher Price. High Med Low BDAE Severity Rating Figure 3 Clustering of the subjects on Fluency Scale in relation to Appropriateness and severity ratings By way of a summary, in Figure 3 the main clusters of patients are plotted against overall degrees of appropriateness and the BDAE seventy rating for fluency ratings. For reasons describ- ed above, Ss 2, 4 and 5 were not included in this analysis. Of the remaining subjects, three were judged to have overall high levels of appropriate fluency, one showed a somewhat equivo- cal pattern and three By virtue of their severe agrammatism could not be effectively judged. Seven Ss were rated as having instances of inappropriate fluency either related to manner of production, length of turn or word-finding difficulties. Clu- sters suggest more cases of inappropriate nonfluency in the group of Ss as a whole than appropriate nonfluency and further that aphasia type does not necessarily correlate with judgments of appropriateness. Clear-cut distinctions were not found on this scale either With regard to appropriateness or severity. DISCUSSION The results obtained confirm to a large extent recent published findings with regard to the fluency disorder in aphasia, viz· that aphasics show many instances of dysfluency and that these dysfluencies are greatly influenced by word posi- tion/grammatical function. ~ that there is not always a correlation between clinical form of aphasia and type of fluency - that the measurement and rating of fluency is not a simple matter v - that the examination of fluency in aphasic speech may offer some insight into the mechanisms of normal speech production3·6 Die Suid-Akaanse Tdskrif Kommunikasieafwykings, Vol. 30, 1983 With the exception of S 14, all the subjects in the present study demonstrated many instances of dysfluency in their expressive language samples. This confirms a large body of literature which discusses dysfluency phenomena in aphasia 2 "· The relative dearth of research into aphasia in relation to normal nonfluencies make such results difficult to interpret. In spite of this, the present data tends to confirm suggestions of workers such as Butterworth3 that the type and locus of dysfluencies in aphasia are similar to those in normals. Though the frequency of such behaviours was not directly measured in the present study it was this aspect which ap- peared to influence judges' qualitative decisions. With regard to type of dysfluency observed, judges noted in- stances of all behaviours delineated on the scale, viz· interjec- tions, repetitions, revisions, incomplete phrases, false starts and word-finding difficulties. These occurred at predictable junctures in terms of the literature on normal dysfluency phenomena, e.g. at syntactic boundaries, before content words and after an error in production.* The feet that most aphasic subjects by definition have greater difficulty in formulating verbal output would lend support to the idea that associated dysfluency would be higher. This is clearly not the case in S14 who conforms to the classic description of jargon aphasia of- fered by Kinsbourne and Warrington,'3 viz: "the jargon aphasic speaks in a copious flow uninterrupted by hesitation and correction". Unlike the patient described by Butter worth,3 S14 showed no instances of hesitation prior to the production of neologisms, nor indeed at expected junc- tures. This provides some support for the 'disinhibition' ex- planation advanced for this type of aphasia. Essentially it refers to a lack of self-monitoring such subjects have over their own output. Acknowledging that an analysis of hesitations and associated non-fluencies may provide an indicator of how the speaker monitors his production, the nature of this type of aphasic deficit is thus apparent. It became clear during the rating of this scale that the assess- ment of fluency, particularly with regard to appropriateness is a very complex matter. Judges' verbal and written justifica- tions were often not specifically related to the frequency locus or type of dysfluency per se but on the overall impression gain- ed. A measurement of fluency control inter alia involves aspects such as number of words, speaking time and speaking rate (which in themselves comprise variables such as phona tion rate, number and duration of silent pauses)/ The present study did not include rate as part of the fluency scale (cf. Yairi et al.). Rather this was seen as a paralinguistic behaviour to be measured on another scale. Similarly, hesitation analysis was subsumed broadly under 'pauses'. In the sense that global rather than quantitative decisions were required in the present study, this somewhat restricted taxonomy of fluency aspects is acceptable. It does, however, suggest that ratings may have been made on the basis of behaviours other than those delineated by the scale. Finally, and of particular relevance, the results on the Fluency Scale showed that there was not always a correlation between the clinical form of aphasia and type of fluency rating This supports the findings of Deloche et a l / who showed that when certain aspects of fluency are considered, no clearcut differentiations in terms of aphasic types suggest themselves. R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 01 2) The results of the present study suggest that some dysfluencies actually enhance the communicative process since they reflect underlying linguistic processing on the aphasic's part and ap- parently shadow normal strategies. In other words they appear to be an indication to the listener that the aphasic is attempting to retrieve the word, is self-monitoring and is employing (however inefficiently) search strategies which reflect com- municative motivation. It is only when such behaviours occur with high frequency or at unexpected junctures that they are judged inappropriate. The fact that these behaviours occur in all but one of the subjects studied casts some doubt onto the question of the utility of the fluency/non-fluency dichotomy which has been used to differentiate aphasics. This point will now be expanded. IMPLICATIONS The writer would suggest on the basis of the above results that the dichotomy of "fluent" versus "nonfluent" aphasia is a potentially misleading one. If this terminological distinction is used to separate patients on the basis of expressive output alone, and if the definitions in- volved are specified clearly and precisely, then it does have the advantage of reducing terminological confusion.16 However, as research on both normal and abnormal speakers suggests, fluency is a very complex phenomenon comprising a number of component parts and is hence difficult to measure. Deloche et al.« point out that traditional aphasic classification schemes often incorporate different aspects into their defini- tions of fluency. Even on the fluency scale in this study which incorporates components traditionally associated with an analysis of fluen- cy, clear-cut distinctions in terms of appropriateness could not be consistently made between nonfluent and fluent Ss Thus many of the so-called 'fluent' Ss were judged to be inap- propriate along these dimensions. A problem of definition arises here. The components of the fluency scale are certainly more conventional than the broad definitions applied by aphasiologists and hence may well have been too restricted to derive differences among patients. Notwithstanding this, the writer would suggest that the terms nonfluent/fluent have many potential pitfalls in terms of defini- tion and specification and that possibly a distinction should be drawn between 'fluency' as characterised in traditional terms and incorporating prosodic features such as rate, hesitations and repetitions, and between fluency in grammatical, lexical and semantic terms. If indeed there is a correlation between clinical forms of aphasia and these separate types of fluency then the distinction is possibly worth retaining. But as the results of the present study suggest, from both terminological and conceptual viewpoints, the use of a fluent/nonfluent dichotomy in the characterisation of aphasic language prob- ibly warrants reconsideration. ACKNOWLEDGEMENTS rhe writer gratefully acknowledges the help and guidance of ler supervisor, Prof. M. L. Aron, Head, Dept. of Speech Claire Penn Pathology and Audiology, University of the Witwatersrand. For financial assistance, the writer is indebted to the Human Sciences Research Council and to the Senate and Council Research Committees of the University of the Witwatersrand REFERENCES 1. Berger, H. & Sinoff, A. Aspects of cohesion, tense and pronoun usage in the discourse of the older language- impaired child. S.A.J. Comm. Dis., 1978, 25, 3-16 2. Brown, C. S. & Cullinan, W. L. Word-retrieval difficulty and disfluent speech in adult anomic speakers. J Speech Hear. Res., 1981, 24, 358-365. 3. Butterworth, B. Hesitation and the production of verbal paraphasias and neologisms in jargon aphasia. Brain Lang., 1979, 8, 133-161. 4. Cohen, J. Weighted Kappa: Nominal scale agreement with provision for scaled disagreement or partial credit Psychol Bull. 1968, 70, 213-220. 5. Dalton, P. & Hardcastle, W. J. Disorders of Fluency. Lon- don: Edward Arnold, 1977. 6. Deloche, G., Jean-Louis, J. & Seron, X. Study of the tem- poral variables in the spontaneous speech of five aphasic patients in two situations, interview and description Brain Lang., 1979 8, 241-250. 7. Dixon, W. J. & Brown, Μ. B. MBDP - 79 Los Angeles· U California Press, 1979. 8. Goldman-Eisler, F. Psycholinguistics London and New York: Academic Press. 1968. 9. Goodglass, H. & Kaplan, E. The Assessment of Aphasia an£ Related Disorders. Philadelphia: Lea and Febiger, 10. Holland, A. L. Communicative abilities in daily living Baltimore: University Park Press, 1980. 11. Holland, A. L. Observing functional communication of aphasic adults. J. Speech Hear Dis., 1982, 47, 50-56. 12. Kertesz, A. Aphasia and Associated Disorders. New York: Grune & Stratton, 1979. 13. Kinsbourne, M. & Warrington, E. U. Jargon Aphasia Neuropsychologia, 1963, I, 27-37. 14. Morris, R„ Blashfield, R. & Satz, P. Neuropsychology and cluster analysis: Potentials and problems. J. Clinical Neuropsychology, 1981, 3, 79-99. 15. Quinting, G. Hesitation phenomena in adult aphasic and normal speech. The Hague: Mouton, 1971 16. Wagenaar, R„ Snow, C. & Prins, R. Spontaneous speech of aphasic patients: A psycholinguistic analysis. Brain Lang, 1975, 2, 281-303. 17. Yairi, E. & Clifton, N. F. Disfluent speech behaviour of preschool children, high school senior and geriatric per- sons. J. Speech Hear. Res., 1972, 15, 714-719 18. Yairi, E„ Gintavitas, J. & Avent, J. R. Disfluent speech associated with brain damage. Brain Lang 1981 14 49-56. ' ' The South African Journal of Communication Disorders Vol. 30 1983 R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 01 2) Philips Hearing Aid Services A Division of S.A. Philips (Pty) Ltd. PHILIPS HEARING AID SERVICES E E . ? , f f i 0 ? i ° ° 5 C a v e n d i s h C h a m b e r s , 183 ^gPP^Street, P.O. Box 3069, JOHANNESBURG. PHILIPS Die Suid-Afrikaanse 7wrw · „ •yasmj w Kommunikasieafwykings, Vol. 30, 1983 R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 01 2)