Forty years on reference: Devitt draws up the account in Reference and Beyond

Michael Devitt, Reference and Beyond: Essays in Philosophy of Language Oxford University Press

Seen from outside, the question of what a name refers to looks like hair-splitting.

It is not. The last fifty years of the philosophy of language have turned largely on that question — and the answer given to it has settled how knowledge, meaning and reality are to be understood.

Devitt's collection, published by Oxford University Press, is the account of those fifty years drawn up from one side of the argument.

What was the argument?#

The description theory. The view inherited from Frege and Russell, dominant in the middle of the twentieth century: a proper name is in fact a concealed description. When I say "Aristotle", what I mean is a bundle of descriptions — the pupil of Plato, the teacher of Alexander, the author of the Metaphysics.

John Searle defended the most developed form of the theory. A name answers not to one description but to a bundle of them; the object that satisfies a large enough portion of the bundle is the name's referent.

The causal theory. In the 1970s Saul Kripke brought the theory down in Naming and Necessity. His counter-example is famous. Most people know only one description of "Gödel": the man who proved the incompleteness of arithmetic. Suppose the proof was in fact Schmidt's and Gödel stole it. On the description theory, we ought then to be referring to Schmidt when we say "Gödel". But we are not. We are talking about Gödel.

Kripke's alternative: a name is fastened to an object at a moment of baptism, and that attachment travels along a causal chain handed from speaker to speaker. I may know false things about Gödel; the name is tied to him along the chain all the same.

Devitt is among the most persistent and most systematic defenders of this tradition.

What Devitt adds#

Two things set Devitt's position apart from the rest.

Naturalism. For Devitt the philosophy of language is not an a priori inquiry resting on intuitions but an empirical investigation of how language actually works. That is a strong thesis about philosophical method, and it cuts against most of the field.

Devitt has long argued, accordingly, that what philosophers call intuition is itself theory-laden, and that its value as evidence should therefore be handled with care.

Realism. Devitt's Realism and Truth sets the dispute in the philosophy of language on a metaphysical footing: the world exists independently of how we represent it. A theory of reference is the story of how we reach that independent world.

Put the two together and the result is a programme that is anti-descriptivist and opposed to semantic internalism at the same time.

The NDPR review#

The review in Notre Dame Philosophical Reviews finds the strength of Devitt's collection in its consistency: though the papers span forty years, the book pursues a single programme throughout.

The same consistency, the review notes, is also the book's most contestable feature. Devitt brings the same basic argument to bear on the more refined versions of descriptivism developed over the past two decades, and at certain points the reader may look for more give.

The Searle connection#

That the book should appear today is no accident, given the file on John Searle we are also carrying.

The most accomplished defender of the description theory Devitt has spent forty years opposing was Searle. With Searle's death, the appearance of the collected work of the other side closes one period of the argument.

There is an irony in this. Searle's best-known argument — the Chinese Room — holds that syntax is not enough for semantics. Devitt's theory of reference holds that meaning comes not from the descriptions in a speaker's head but from a causal link with the world.

Both agree that meaning is not in the head. On where it does come from, they part company entirely.

Who is it for?#

This is not an introductory book. It is for readers already at home in the central debates of the philosophy of language.

For graduate students working on the philosophy of language in Turkey it is nevertheless a valuable resource: following the forty-year course of an argument through the pen of a single writer gives a grasp of it that scattered reading of individual papers does not.

It also has an unexpected point of contact with the question set out in the file on Macit Gökberk we are running today — how philosophy is to be done in Turkish. A theory of reference bears directly on the question of what a concept goes on referring to as it is carried from one language into another.

Translation, after all, is a problem of reference.