Cross-linguistic Comprehension Variation in Aphasia: New Results and Hints for a New Analysis

Grodzinsky, Yosef (2006) Cross-linguistic Comprehension Variation in Aphasia: New Results and Hints for a New Analysis. [Clinical Aphasiology Paper]

PDF 187.pdf Download (115kB)

Abstract

I. Stability and Variability in Aphasic Patients’ Comprehension I.A. Stability The stability of sentence comprehension in aphasia has long been a subject of debate. A clear view is important here: an understanding of the underlying pathology would be hindered by unstable performance patterns, which in turn would render the design of treatment and evaluation protocols impossible. In an attempt to assess the degree of variability in Broca’s aphasia, a retrospective quantitative literature survey was recently conducted, from which raw scores of a large group of Broca’s aphasic patients were culled. The resulting database contains 233 data points, constructed from comprehension scores of 69 Broca’s aphasic patients who received 5934 “semantically reversible” sentences of 14 different syntactic types (Drai & Grodzinsky, 2006). Analyses of this large-scale database with techniques that build on the beta-statistic indicate that cross-linguistic and cross-laboratory results give a stable picture once the correct method for partitioning impaired from preserved abilities is introduced. Of several candidate deficit analyses considered, one revealed robust structure: It appeals to the syntactic concept of Movement as the locus of comprehension deficit in Broca’s aphasia, in line with the Trace-Deletion Hypothesis (TDH). While a robust syntactic Movement effect is evident, further empirical investigations of the database uncover new sources of variability. These are the topic of the current presentation. I.B. Cross-Linguistic Comprehension Variability: Active and Passive Central to the above view are comprehension results of English experiments on active and passive sentences. Comprehension of active (–Movement) sentences is far superior to that of passive (+Movement), which seems to be at chance. However, a further analysis of the Drai-Grodzinsky database reveals an odd, persistent, and previously unnoticed variability that transcends the ±Movement distinction: While –Movement active sentences are universally comprehended at above-chance levels, cross-linguistic results of experiments that use +Movement passives present a highly variable picture. An array of data from 8 languages (Dutch, English, German, Hebrew, Italian, Japanese, Korean and Spanish) results in a picture that has the following qualitative character (in parentheses – group performance level per sentence type in a given language group):

Actions (login required)

View Item