Category invasion and Web data

The present paper proposes a theoretical approach to a language use that is considered exceptional to the rule. Specifically, under the title ‘category invasion’, this paper investigates a phenomenon in which neighboring categories with mutually distinct prototype centers deeply invade each other’s referential territory including the prototypical core. It attempts a theoretical description of the phenomenon by analyzing the critical notions of cognitive domain and dimension and explores its linguistic mechanism. Then, a carefully devised test using Web data detects an occurrence of invasion in the language of the Web. The study aims to show that invasion is a permanent and legitimate feature of language, to be treated on a par with such established notions as synonymy, overlap, and intersection rather than be dismissed as a simple result of bad use.