Crossing the associative/inferential divide: ad hoc concepts and the inferential power of schemata

How do we construct ad hoc concepts, especially those characterised by emergent properties? A reasonable hypothesis, suggested both in psychology and in pragmatics (Relevance Theory), is that some sort of inferential processing must be involved. I argue that this inferential processing can be accounted for in associative terms. My argument is based on the notion of inference as associative pattern completion based on schemata, with schemata being conceived in turn as patterns of concepts and their relationships. The possible role of working memory in inferential processes of this sort is also addressed.