Kathleen Baynes - US grants

The primary goal of this project is to contribute to the modeling of the processes involved in auditory comprehension, through a better understanding of the deficit in auditory comprehension that has been associated with "agrammatic" aphasia. Proposed manipulations are intended to more sharply define the impaired components of auditory comprehension, to investigate dissociations and associations in comprehension abilities and to determine whether any of the dissociations now demonstrated in the productive abilities of "agrammatic" aphasics are mirrored in components of comprehension. Specifically, the possibility that difficulty processing the negative morpheme in "pure definitional negatives" may be associated with difficulty processing other semantic phenomena will be investigated to determine if there are theoretically interesting sets of semantically defined functions that are at risk in aphasia. Furthermore, if these problems are related to a change in semantic access procedures, it may be possible to observe consistent relations between them and the failure to show normal priming in lexical decision tasks. This project also begins to examine the auditory comprehension of a variety of sentence types associated with so-called "agrammatic" aphasia across a broad spectrum of patients to determine 1) whether the same dysfunction uniformly effects all implicated sentence types and 2) whether the difficulty is specific to any particular aphasic sub-group. Finally, this project initiates a study of the recovery of auditory comprehension beyond the period usually associated with spontaneous recovery. Observed improvements in patients at a point more than two years post-stroke suggest more systematic observation of progress would be useful. In pursuing these goals, a battery of clinical and experimental tests will be developed to allow adequate description of patients both to observe progress and to select patients for appropriate experimental studies. Proposed work will include both group studies and case studies as dictated by the hypotheses under consideration and the profiles of the patients available.

The primary aim of text processing research is to specify the nature of readers' discourse representations and the processes that construct them. Readers represent a text in the form of a discourse model, a representation of the entities (e.g., people, objects, places) and events in a text and the relations among them. A discourse model is constructed incrementally; incoming language is analyzed against a background of knowledge structures consisting of representations of recent text input and general world knowledge. Much of the research in text comprehension has been aimed at understanding how and when readers access relevant world knowledge and integrate this knowledge into their developing discourse model. Relatively little attention has been devoted to understanding how the discourse model is neurally instantiated. One of our goals in this project is to investigate the right hemisphere's role in discourse processing using the theoretical concepts and paradigms that are standard in psycholinguistic and text-processing research. Our second goal is to examine hypotheses linking the right hemisphere's role in discourse processing to the nature of its lexical-semantic system. One hypothesis suggests that the right hemisphere represents words differently than does the left. More diffuse word representations in the right than in the left hemisphere supports priming of distantly-related concepts. These concepts support many of the inferences that are necessary to construct a coherent discourse model. A second hypothesis suggests that the right hemisphere processes words differently than does the left. Slower activation and decay in the right hemisphere relative to the left may be important in the elaboration and reinterpretation of the discourse model. The right hemisphere's role in discourse processing will be investigated among normal young adults, right- and left- hemisphere lesioned patients, and callosotomy patients. Discriminating between these hypotheses has important implications for understanding a number of disparate phenomena, including differences in semantic priming in the left and right hemispheres and differences in comprehension deficits that arise from left and right hemisphere damage.

memory; neural information processing; semantics; phonology; psycholinguistics; speech; clinical research; human subject;