Language from within (review)

Language from within is a large and ambitious work. In nearly 800 pages of text, Pieter Seuren covers a broad range of topics in syntax, semantics, and pragmatics, from both contemporary and historical perspectives. Many of the topics are familiar from S's extensive previous writings, but they are brought together here in what is intended to be a sketch of, or perhaps a set of pointers to, an overarching synthesis.

There are a number of broad themes running through the two volumes. Perhaps the broadest of these themes, introduced at the beginning of Vol. 1, is the need for both formalism and ecologism in the study of language. The ecologism espoused by S has both cognitive and social (or societal) components, making the overall title, 'language from within', perhaps slightly misleading in its apparent emphasis on the cognitive. S is, however, rightfully critical of those who espouse the view, either explicitly or implicitly, that ecologism is incompatible with formalism. Ecologism in its correct form should, according to S, be compatible both with those existing formalisms that apply (mainly) to the less context-dependent aspects of language (e.g. syntax) and with the mainly yet-to-be-developed formalisms for those aspects of language that tend to be the focus of current ecology-based approaches to language, and which proponents of those approaches sometimes claim cannot be formalized.

S further promotes two more specific themes. The first is that the ontology underlying natural language is not the one that underlies the standard modern approach to logic and its application to natural language. Throughout the two volumes, S takes issue with a number of authors and schools of thought. In this particular connection his bêtes noires are Bertrand Russell and W. V. Quine, with their extensionalist ontology. Alexius Meinong, whose ideas on existence versus subsistence (being) were first embraced and then rejected by Russell, is portrayed as a kind of antihero. A more subtle form of extensionalism, which S considers a dead end, is the Montagovian extensionalization of intension, in Richard Montague's version of possible-worlds semantics. The basic problem for extensionalism, on S's analysis, is the omission of the cognitive in accounts of natural language meaning. So, S suggests that Montague has things the wrong way around, and he calls for an intensionalization of extensions. S proposes that language is interpreted according to a functional principle, which he calls the principle of optimization of sense, truth, and actuality (OSTA), and which he claims is a version of a more general principle governing all of perception and cognition. The principle states that when there is a choice, utterances are interpreted so that they make sense (rather than nonsense), so that they are true (rather than false), and so that they refer to actual things rather than fictitious things, in that order of priority.

S finds recent discussions of extension and intension to be muddled. He argues that the traditional extension/intension distinction for predicates really does apply only to predicates. He claims that the generalization of this idea to terms and propositions, which is widespread in modern approaches to logic, is a mistake. He further argues that Gottlob Frege's notion that the sense of a proposition is the thought it expresses has merit, but its extensionalization in possible-worlds semantics does not.

The second theme is that the question of which of the many possible logics best maps onto human thought and language is an empirical issue. This theme is taken up in detail in Vol. 2, though the conclusions that S reaches have been presaged in his previous writings.

A crosscutting theme, again familiar from S's published work, is the (initially negative and later dismissive) treatment that generative semantics, or semantic syntax as S prefers to call it, has received in the last thirty-five years. S has, of course, continued to...