Robert S. Wachal’s research while affiliated with University of Iowa and other places

Thirty-eight measures of lexical diversity are obtained (with the aid of a computer) from the free speech of 20 aphasic and 20 normal control subjects. After statistical analyses of significance and redundancy, 18 variables are retained for a later empirical derivation of types of aphasia. Eight of the variables are sufficiently significant to discriminate the aphasic group from the normal group.

This paper presents the background, rationale, and examples for a comprehensive psycholinguistic study of free speech samples obtained from 50 adult aphasics in comparison with those obtained from 50 adult normal speakers. Procedures developed for the recording, transcription, and coding of the speech samples and the analysis of the data with regard to measures of lexical diversity, morphological complexity and grammatical form-class usage, generative-syntactic complexity, sentence length, abnormal productions and of paralinguistic features (rate of speech, pauses, pause fillers, vocal gestures, etc.) are described. First results of a pilot study with 20 aphasic and 20 normal speakers are reported. Hypotheses of the project include differences between the linguistic performance of normal and aphasic speakers on a combination of variables as well as the existence of specific types of aphasia which can eventually be related to medical findings.

A system of 20 Fortran and PL/1 programs, developed for an analysis of aphasic and normal speech transcripts, is described in detail. The programs aid in lexical, grammatical, paralinguistic and statistical analyses, as well as in data preparation and correction. They can also be used in schizophrenic and other kinds of pathological language and are adaptable to the analysis of written-language samples and the investigation of authorship and style. (Author/FWB)