Philosophy of Language - Overview of the Class
Two ways to look at this material: in terms of its historical development, and by topic. We will try to combine the two.
I. Historical Development.
The 20th century is really where the philosophy of language begins. Earlier philosophers sometimes talked about language, and it is possible to revisionistically interpret them as concerned about language even when they didn't realize it. But self-consciously taking language as a subject matter for philosophy, in a way that regards it as of fundamental importance, is basically a 20th-century phenomenon.
A historical precursor: John Locke devotes an entire book of his Essay Concerning Human Understanding to language ("Of Words"). However, Locke views words as essentially just labels for ideas which exist prior to and independently of the words which label them. He thinks of philosophy as primarily concerned with ideas and their relations. Words are important only because we can fail to communicate when different people attach the same word to different ideas, or different words to the same idea.
For much of the 20th century, many philosophers thought of language as much more important to philosophy than this. Many philosophers regarded all philosophical issues as essentially issues about language, so that in a sense the philosophy of language was all of philosophy.
In 1967 Richard Rorty published a landmark anthology of twentieth-century philosophy entitled The Linguistic Turn. He begins the book with a collection of "classic statements of the thesis that philosophical questions are questions of language." That is the thesis that is characteristic of the linguistic turn, which in turn characterized much of 20th century philosophy. (We might think of the 20th century as the rise and fall of the idea that all legitimate philosophical questions are questions of language.)
Rorty distinguished between two strands in the linguistic turn: ordinary-language philosophy and ideal-language philosophy.
A. Ordinary-language philosophy held that we need to carefully understand how people ordinarily use language, and that most philosophical issues arise from too simplistic an understanding of the way language is actually used. (Wittgenstein: philosophical perplexities arise when language "goes on holiday," when we attempt to use words outside the contexts in which they are meaningful.)
B. Ideal-language philosophy held that ordinary language itself is often inherently misleading, and that the best way to think about philosophical issues is to construct a better language, one which will not suffer from the ambiguity, vagueness, and misleading implications of ordinary language.
Both strands tended to be skeptical about the value of much of historical philosophy.
We will look at the writings of several of the most influential philosophers of the twentieth century, notably Bertrand Russell, Ludwig Wittgenstein, A. J. Ayer, J. L. Austin, and Saul Kripke. Of these figures, Russell and Ayer could be regarded as "ideal-language philosophers," while Wittgenstein and Austin are ordinary-language philosophers. Kripke does not fall into either category, and indeed rejects the assumption that philosophical issues are all at bottom linguistic. His book Naming and Necessity marks an important turning point, a sort of counter-revolution against the linguistic turn.
II. Topical: Syntax, Semantics, and Pragmatics
(For a good introduction to these categories, see the Introduction to A. P. Martinich, ed., The Philosophy of Language, Fifth Edition, Oxford University Press, 2008.)
'Syntax' is more or less synonymous with 'grammar', though philosophers often use the term more broadly to refer to any characteristics of a sentence that don't involve semantics. Thus, while a linguist would distinguish between phonology and syntax, philosophers may treat phonology (and orthography) as "syntactic".
Semantics is the study of the meanings of linguistic expressions (as opposed to their sound, spelling, etc.). Of course, 'meaning' is a notoriously vague and ambiguous term; many different kinds of meaning are part of semantics. Among the semantic notions we will make use of are these:
reference or extension: the object or set of objects to which an expression applies. (Actually it might be better to regard the reference of an expression on a particular occasion of use as pragmatic rather than semantic. More on this later.)
truth and falsity (sometimes these are regarded as the extensions of declarative sentences)
meaning or intension (what determines the extension of an expression; often regarded as a function from possible worlds to extensions)
what a competent user of an expression must know (alas, although this is a very important concept, there is no term that unambiguously expresses it. The word 'meaning' itself probably comes closest, but unfortunately 'meaning' can also mean intension, sometimes is used to mean 'extension', and sometimes is used to mean still other things.)
Pragmatics has to do with context-dependent features of language. There are at least two rather different varieties of pragmatic notions.
The term 'semantics' tends to be restricted to properties of sentences that remain constant as long as the same language is being spoken, while pragmatic values vary from context to context. For instance, the expression 'local fauna' has the same meaning regardless of when I use it, but its extension varies from one context to another, so we may say that the extension of 'local fauna' is pragmatically determined. (By contrast, the extension of 'dog' does not vary from one context of use to another.)
Pragmatics also includes things people can do with words or sentences that go beyond the literal meaning of the expressions involved. For instance, the sentence "he always shows up for class on time" means exactly what it says. But consider the following context of use: someone asks me, "is so-and-so a good student?" and, after a long pause, I utter the sentence above. In this context, I am very likely using the sentence to convey the information that the student really is not all that good, even though I have not literally said this. One way to put the difference is to say that the literal meaning of the sentence is semantically encoded by the sentence, while the information that the student is not very good is instead pragmatically imparted (cf. Nathan Salmon, Frege's Puzzle).
One way to see the difference between syntax, semantics, and pragmatics is to think about various kinds of linguistic deviance. A sentence can be pragmatically deviant without being semantically or syntactically deviant, and it can be semantically deviant without being syntactically deviant. Consider the following examples:
Easy please to John. (syntactically deviant, i.e. ungrammatical.)
The present king of France is bald. (syntactically OK, but semantically problematic because the subject term has no referent.)
Colorless green ideas sleep furiously. (syntactically OK, but seems semantically messed up in that it is meaningless or at least has an incoherent meaning.)
Plato was the discoverer of uranium. (syntactically OK, no problems with reference, perfectly meaningful, but false.)
Suzy is one of the best two students in the class. Said in a letter of recommendation; in fact, unmentioned in the letter, there are only two students in the class. Syntactically fine, semantically OK (terms all refer, sentence is meaningful, and sentence is even true), but pragmatically deviant in the sense that it will communicate something false: it's misleading.