This project aims to clarify some logical and computational aspects of the natural language meaning system. A central focus of research will be the general architecture of the grammar and the place ofsemantics within that general architecture. The traditional model here is that of the *pipe line*, a strictly sequential approach in which the interpretation component takes its input from the output of syntax. In contrast, we will investigate the logical and computational consequences of two strictly parallel models which we call the description approach and the combinatorial approach. Additionally, logical and computational models will be developed that throw light upon one of the most important foundational problems of semantics, the problem of intensionality.