A propositional perspective on context effects in human associative learning

Propositional models of associative learning postulate that the behavioral impact of regularities in the presence of two events is mediated by the formation of propositions about the relation between these events. Because the mere statistical contingency between events often does not provide enough information to infer the nature of the relation between those events (e.g., whether one event is a cause or an effect of the other event), it is likely that people will take into account relational information that is provided by the context when forming propositions about the relation between events. Hence, propositional models predict that contextual cues which provide relational information can moderate associative learning. The present paper provides a brief review of several studies that support this prediction. These findings contribute not only to the cognitive literature on the mental mechanisms that mediate associative learning but also to the functional literature on associative learning by providing novel evidence for arbitrarily applicable relational responding. Vice versa, functional research on relational responding can provide a new source of information for the development of cognitive theories of associative learning. This article is part of a Special Issue entitled: SQAB 2013.