IT wAS IN the spring of 1976 that I began developing the conception of language structure that has come to be known as cognitive grammar. At that time theoretical linguistics was languishing in the hard-fought but ultimately rather pointless debate between proponents of "generative" vs. "interpretive semantics," which pivoted on the still-contentious issue of whether syntax represents a distinct and autonomous component of the linguistic system. Though hardly a neutral bystander, I felt that

more »

... ther camp was attacking the basic conceptual problems that needed to be resolved before that issue could be examined in a meaningful way. So one day I sat down to do just that, assuming (with the immodesty befitting a linguistic theorist) that my own efforts to formulate a natural and viable theory were unlikely to yield a product very much worse than those currently available. My strategy was to start anew by ignoring all extant theories as such, without however forgetting the factual knowledge and certain specific insights on which they were based (i.e. to raze the buildings but salvage some of the lumber). Within a few years, I had constructed a basic conceptual framework (then labelled space grammar) that no doubt seemed equally bizarre to linguists of all theoretical persuasions. This reaction was attributable in no small measure to superficial reasons (terminology; notation; the heavy use of diagrams), but was also due to the radical nature of the theory's central claims (the symbolic nature of grammar; the notional definability of basic grammatical classes) and the unfamiliarity of the world view it manifested. As a consequence it tended to be ignored in the busy theoretical marketplace. The original framework has changed very little. As I and my students have probed more deeply and applied it to more languages and a vastly greater range of grammatical phenomena, it has undergone refinement and substantial elaboration, but nothing in the way of basic revision. What has changed to some degree is the prevailing intellectual context. Over the same period, many other linguists have also been responding, in their individual ways, to the imperative of their own insight and the perceived inadequacies of the vii