Phrase Length and the Type and Severity of Aphasia

Abstract

Three major syndromes of aphasia are repeatedly described in the neurological literature, characterized chiefly by the nature of changes in the speech pattern. The terms Wernicke's aphasia, Broca's aphasia, and amnesic aphasia are generally understood, but have slipped out of popularity. Terminology based on the distinction between language input and output has become popular, but actually does not correspond well to the syndromes and raises false issues concerning them. Moreover, the chief distinguishing characteristics of these syndromes are elusive to objective testing and have been ignored in factor analytic studies. In this study, a series of rating scales was used in place of objective scores and each of the syndromes was defined by a profile of ratings. High reliability of diagnostic classification was obtained by this means, after impressionistic classification on the basis of a check list proved grossly unreliable. An objective index of length of uninterrupted word-sequences was found to produce a sharp dichotomy in the aphasic population, separating Broca's aphasics from Wernicke and amnesic aphasics. This objective measure was reliably reproduced by several of the rating scales, including one which explicity required a judgment as to the longest number of uninterrupted words in nonstereotyped sequences. Physical signs of localization — i.e., presence of hemiplegia — and operative findings corresponded to the differences in the speech profiles. Severity of aphasia must be taken into account, since the profile offers no differentiation among patients who have recovered to a level of mild or slight residual aphasia. Natural and artificial factors influencing the distribution of different types of aphasia must also be considered in interpreting test results from unselected aphasic groups. Such selective factors may produce artifactual correlations or else obscure fundamental relationships.