As one phase in a study of ‘intelligent system’ utilization for understanding/facilitating the interaction between a patient and a psychotherapist, contrasting models of their verbal exchanges are examined critically. The two emergent models studied are the essentially Speech Act Theory model as used by William Labov, et al. and our own initial formulation of a Discourse-Analytic model. Extent of computational operationalizability, degree of fine-grainedness of analysis, and other properties of the two developing models are examined. Focus and precision are provided for the comparison by using each model vis-a-vis the same body of primary data. The prospective usefulness of such an intelligent systems capability both as a clinical aid and for research on the therapeutic process is clarified in the discussion of the results of using each of the contrasting models. To aid in assessing the feasibility of realizations of the models in intelligent systems programs, the issue of tacitness/explicitness in their present form is directly addressed.
展开▼
