The pragmatic basis of aphasia By Marc L. Schnitzer (review)

348LANGUAGE, VOLUME 67, NUMBER 2 (1991) The pragmatic basis of aphasia. By Marc L. Schnitzer. Hillsdale, NJ: Erlbaum , 1989. Pp. x, 213. Cloth $34.50. Reviewed by Harold Goodglass, Boston University School of Medicine The author of this book is a linguist who is well grounded in aphasiology, having contributed a number of well-known articles on linguistic aspects of aphasie speech—particularly morphosyntax and phonology. In this thoughtful book Schnitzer presents a point of view that deserves serious consideration, both with respect to its central thesis and with respect to his comments on the nature of language. Having said this much by way of establishing S's credentials , I will also summarize my opinion that his central thesis is a weak and nonexplanatory description of the phenomena that form the subject matter for the book. In 1978 S, in collaboration with Nicolas Linares at the Medical Campus of the University of Puerto Rico, set out to devise a clinically practical test of aphasia for speakers of Puerto Rican Spanish, with a parallel form in English. Their objective was to develop an instrument suitable for evaluating aphasie patients who were bilingual in those two languages. As often proves to be the case in the development of a clinical test, the full examination took much too long for practical application. S and Linares therefore restricted their data collection to the morphosyntax sections of the Spanish and English test batteries . The procedures for item selection and test validation were at best rough and ready, with little resemblance to the standards normally expected in clinical test development—for example, the use of a normative population of reasonable size that is similar in age and education to the experimental population. S and Linares tried out their morphosyntax battery on ten school children, aged 1 1 to 13, who were considered bilingual. Their goal was simply to eliminate or modify items that could not be passed by a sixth grader. Having modified the test to their satisfaction, they proceeded to administer it to aphasie subjects, recruited from various inpatient and outpatient sources. Of more than 100 candidate subjects referred to them, they were able to find only 21 who were actually aphasie and who were testable on at least one of their battery of 20 English and 16 Spanish subtests. These subjects varied widely in age, education, and severity and type of aphasia. Only four were able to take the whole battery; most patients had eight or fewer of the Spanish subtests. Of the 16 subjects who had some degree of bilingualism, only two had the full English battery and most had only a few subtests. With the handicap of unsystematic, scattered data, S is forced into an extremely rough and statistically unsound data analysis. One hundred twenty of the 189 pages of text in the book are devoted to a detailed presentation of the two test batteries, including the wording of the instructions to subjects, individual case summaries of the 21 patients along with their test scores, and statistical comparisons between a number of factors. Among the factors compared are modality of presentation and response (e.g. oral vs. written) and category of morphosyntactic task. Only a few of these REVIEWS349 comparisons are theoretically driven; these include pragmatic oddness vs. normality of sentences, transparency vs. opacity of information to be extracted from a sentence, and availability of alternative responses. Since theoretical considerations had not been systematically applied in the original selection of test items, some of the more interesting comparisons are based on a post-hoc analysis of item demands. For example, scattered among the test sentences are a few that are obviously pragmatically odd (e.g. La maestra fue empujada hacia el piso por el estudiante 'The teacher was pushed to the floor by the student'). These were later selected for a comparison with the much larger number of 'normal' sentences. S does not identify the items that he considered pragmatically odd. Many of the test sentences are strangely artificial in their construction and choice of vocabulary (e.g. The priest went to the store and the lawyer also did so), and it is unclear whether they were classified as pragmatically odd or...