The Structural Prominence Hypothesis (SPH) proposed in Friederici and Gorrell (1998) has been claimed to find support from cross-linguistic data and to be empirically superior to linear strategies. In this article we evaluate the SPH against data of thematic-role-reversal comprehension of Chinese aphasics using Mandarin Chinese and Cantonese. We focus on the comprehension of Chinese relativized constructions for the reason that linear position and structural prominence diverge in these structures. The two types of strategies thus make different predictions about thematic reversal comprehension with respect to these structures. An examination of the performance patterns of Chinese aphasics from three different studies (Su & Law, 1993; Law & Leung, 1998a, 1998b) reveals a tendency on the part of the patients to make use of heuristics that takes into consideration the linear order of lexical NPs, not their structural prominence. Furthermore, their strategy does not necessarily make reference to verb position.
Copyright 2000 Academic Press.