Language

Peter Hanks has written an immensely thought-provoking book, of great originality and vast ambition. At the heart lies a novel conception of content, based on rejecting one of the most deeply entrenched distinctions in analytic philosophy: the content-force distinction. Much space will be needed here for presenting and discussing the basic ideas due to their originality. Their application to outstanding problems in the philosophy of language can therefore be mentioned only briefly. H himself sums them up as follows:

Over the course of this book I have advanced an account of the semantic contents of proper names, a solution to Frege’s puzzle about identity sentences, a theory of empty names and true negative existentials, a semantic analysis of propositional attitude reports, a solution to Kripke’s puzzle about belief, an approach to the semantic contents of indexicals and demonstratives, a solution to the problem of de se belief, a proposal about the contents of interrogatives and imperatives, and, in outline, a new taxonomy of speech acts. This is rather a lot. It is a testament to the power of the general conception of content in which these views are framed that it lends itself so easily to solving these problems. But none of these views is as important to me as this general conception of content.

(205)

It is indeed a lot, but we shall focus on what is indicated as most important.

H adamantly rejects the content-force distinction, overwhelmingly accepted in analytic philosophy since Frege 1918. Frege argued as follows: when you assert a conditional p → q, you do not assert the antecedent, p. If the assertive force is part of the content of an asserted proposition, then it is part of p as it is asserted on its own, but not part of p as antecedent in the conditional. Consider the modus ponens inference: p; p → q; hence, q. On the assumption that force is part of content, the content of p in the minor premise and in the major premise are not the same, so there is equivocation, and the inference is not valid. But modus ponens is valid. Hence force is separate from content.

That we must separate force from content has practically been an axiom in speech-act theory ever since, and in the philosophy of language generally. H correctly distinguishes between two content-force distinctions (18–19). The constitutive distinction says that there is nothing assertive in the content of an assertion, nothing interrogative in the content of a question, and nothing imperative in the content of an order. The taxonomic distinction says that utterances with different forces share the same truth-conditional content: the contents of an assertion that p and of a question whether p are the same.

H rejects both versions. His main argument against the former is that the classical conception of mind-independent propositions runs into problems. More precisely, it runs into the problem of ‘the unity of the proposition’, a problem with a history going back to Plato. The major issue here is explaining how propositions have truth conditions (42–43). According to H, we need such an explanation since we need to understand why judgments have truth conditions, and on the classical picture, they do thanks to having propositional content. Therefore, saying that it is just a primitive fact about propositions that they have truth conditions is unsatisfactory. Bertrand Russell’s conception of a proposition as an ordered pair does not work, since an ordered pair does not say anything (a common objection). Frege’s conception of how the sense of a name saturates the sense of a predicate does not explain why the result has truth conditions (a less common objection; 49).

According to H, we must reverse the order of explanation: propositions get their truth conditions from our acts (mental and linguistic) of joining, for example, a name reference and a property expression in an act of predication. A similar line has been taken by Scott Soames (e.g. in 2014), and H...