In ordinary language the word rule is used in many different ways, and there is no unique definition which holds good for all of those uses. On the other hand, behavior analysts frequently restrict the concept and define rule technically as "description of contingencies", a definition first proposed by Skinner in 1969. However, an analysis of Skinner's texts suggests that they include, parallel to this restriction of the concept, an extension of the ordinary use of the term. The present paper discusses these two opposite movements of conceptual revision and argues that: (1) the extension of the concept may be interesting in that it may point to functional similarities between phenomena that have different names in everyday language, but may also obscure important differences; and (2) the technical restriction of the concept has produced conceptual confusions related to a referencialist view of language. It is concluded that the analysis of the different uses of the term rule in everyday language may contribute positively to theorizing on rule-governed behavior in the Experimental Analysis of Behavior.
experimental analysis of behavior; conceptual analysis; rule
