Protocol analysis and the "silent dog" method of analyzing the impact of self-generated rules.

Within the cognitive literature, verbal protocols of cognitive events are plagued by difficult questions of unconsciousness, completeness, reactivity, and validity. In this paper we argue that these concerns apply with much less force or not at all when protocol analysis is used to determine whether a given instance of behavior is governed by self-generated rules. When adequate controls are used, some patterns of results allow this question to be answered unambiguously and in a manner untouched by the philosophical hurdles encountered within the cognitive literature on protocol analysis. We argue that in at least some circumstances a slightly modified version of protocol analysis allows us to know, in a functional sense, what a person was thinking. Protocol analysis can be very useful to behavior analysts who are interested in determining whether task-relevant behavior is controlled by self-generated rules or is purely contingency shaped.