ABSTRACT
There is one incontrovertible fact on which this paper is based: there are mental processes, largely outside conscious awareness, that underlie speaking and understanding a natural langu~ge,- A generation ago this assertion might have been considered a dangerous speculation; today, behaviorists and phenomenologists apart, it seems like a truism. Several of its implications, however, are controversial, and I shall deal with some of them that concern what is known as "procedural semantics"-an approach to the psychology of meaning based originally on computer programming, and which accordingly emphasizes the process of interpretation.
