PURPOSE AND THE PROBLEM OF ASSOCIATIVE SELECTIVITY
Wallace A. Russell
UNIVERSITY OF MINNESOTA
Since Gibson ( 1941) critical review of the concept of set, it has been apparent that the term has meant many things to many people. Confusion over the usage of the word has been so deep that in some circles it has become equivalent to an admission of ignorance to suggest that an observed bit of behavior was due to the subject's set. Nevertheless, the inevitability with which the term, or one of its many variants, intrudes itself into otherwise meaningful discussions attests to the reality of at least some of the effects which it has been called in to explain. In the area of verbal learning and verbal behavior, scarcely an investigator has failed to mention the possibility that sets, intentions, selector mechanisms, response biases, Aufgaben, motives, purposes, or what have you, might drastically affect performance in his experimental task. In symposia such as this one, it seems that one current of discussion inevitably eddies about the manner in which the use of language is selectively directed, how certain verbal structures are "tuned," or "primed," how certain "schemata" are adopted, or how particular "supraordinate categories" are activated.
There are wide individual differences among investigators with respect to their willingness to introduce special constructs to account for the behavioral selectivity which is usually attributed to set. Most often, however, the attempt is made when the usual operation of recognized associative principles fails to occur. It may be a matter of stimulus selectivity, in which a subject responds to one aspect of a situation and not to another even though it is known that the ignored stimulus elements have in the past been capable of eliciting some response. There may be response selectivity, that is, an instance in which a particular stimulus evokes one response rather than another with which it has been strongly associated. More recently, studies at the University of Minnesota ( Marin