5 Situations, Constraints and Channels (Update of Chapter 4) Edwin Mares ∗ , Jeremy Seligman † , Greg Restall ‡ ∗ Department of Philosophy, Victoria University of Wellington, P.O. Box 600, Wellington, New Zealand, E-mail: Edwin.Mares@vuw.ac.nz † Department of Philosophy, The University of Auckland, Private Bag 92019, Auckland 1142, New Zealand, E-mail: j.seligman@auckland.ac.nz ‡ Department of Philosophy, School of Philosophy, Anthropology and Social Inquiry, The University of Melbourne, Old Quad, Parkville, Victoria 3010, Australia, E-mail: restall@unimelb.edu.au 5.1 From Situation Semantics to Situation Theory Central to the project of situation semantics was the goal of a ‘relational’ theory of meaning, which would explain meaning in terms of the relationship between situa- tions containing meaningful entities or actions such as utterances and situations they are about (Barwise and Perry, 1983). The contrast is primarily with the once dominant view of Davidson, Montague, and many others following Tarski’s seminal ideas about semantics, according to which the meaning of a declarative statement (at least) is to be understood in terms of the conditions under which it is true. There are some difficulties in making this contrast clear. After all, truth-conditional theories of meaning typically also involve a theory of reference, which is concerned specifically with the relation- ship between words and things, and the relation of reference is present in all attempts to produce a situation-based semantics. Likewise, truth has been studied within a situ- ation theoretic framework, most notably by Barwise and Etchemendy (1987), in their treatment of the Liar paradox. Dressing up a truth-conditional account of meaning in the notation of situations and infons is a futile exercise. The important difference is in the theoretical status of semantic vocabulary. For Tarski, semantics involves a clear separation between syntax and semantics, and this separation has been honoured by most of his followers. By contrast, situation seman- tics aimed to do without this separation, taking reference, for example, to be a relation like all others, and with no special theoretical status. That ‘Jon’ refers to Jon is just a fact to be modelled as refers, ‘Jon’, Jon; 1. A consequence of representing semantic facts in the object language is that there is no need for a hierarchical theory of mean- ing, on which the meaning of an expression is some unitary theoretical entity, such as a truth-condition, derived from its more basic semantic properties. Instead, the many facets of meaning can be left unassembled and ready to be used for whatever purpose is required. Handbook of Logic and Language. DOI: 10.1016/B978-0-444-53726-3.00005-0 c  2011 Elsevier B.V. All rights reserved.