Abstract

Toulmin’s (1958) model of argument was employed in the analysis of verbal protocols obtained during the solving of ill-structured problems. The participants were experts in the domain under study. For the analysis the Toulmin model was extended in order to enable description of lines of argument found in protocols as long as 10 paragraphs. Results included: (1) That while the protocol was comprised of a large number of specific arguments, the analysis provided for tracing a solver’s line of argument. (2) On occasion datum and backing were difficult to distinguish. (3) Warrants essentially were not stated, although substantial backing was provided. However, as perhaps would be expected, the Toulmin model did not provide for delineation of components of the problem-solving process. A second analysis assuming a “higher level” problem-solving structure and a “lower level” argument structure produced an integrated problem-solving – argumentation structure depicting how reasoning is used in relation to particular task goals. Finally, at a more general level, problem solving was considered as a classical rhetorical structure.