Tacit integration and referential structure in the language comprehension of aphasics and normals.

Aphasics, brain-damaged patients with no language deficit, neurologically intact elderly subjects, and university undergraduates matched pictures to sentences having compelling tacit implications (e.g., the sentence The fox grabs the hen strongly invites one to assume that the fox will eat the hen). All groups made, for the same sentences, qualitatively similar referential errors consisting in choosing a tacit implication picture. Two auxiliary experiments using the same target sentences in other interpretive situations permitted ruling out the possibility that these errors were due to the putative intrinsic semantic properties of the sentences, showing that the sentences which were most liable to elicit integrative error varied from task to task. These results are interpreted within the conceptual framework which posits that reliable directions for interpretation are couched by the speaker in the very structure of his utterances (the utterance's referential structure) providing the hearer with means to restructure the relevant personal knowledge integrated into the interpretive process in accordance with the speaker's communicative intent. The determination of the referential structure (RSD) of utterances thus seems critical to their correct or, more precisely, conventional interpretation, and, along with the tacit integration of relevant sources of personal knowledge, constitutes the principal cognitive device enabling us to understand each other. But this device appears to be easily corruptible. It is suggested that many errors made by aphasics in language interpretation are due to a failure to follow all referential instructions, but that qualitatively similar failures also occur in normal subjects, though to a lessen degree. Language interpretation is a fallible process and aphasic errors provide remarkable clues for the understanding of its subtle referential mechanisms.