Ernie LePore

Most linguists think that there are infinitely many sentences, that languages are productive and systematic. Maybe the most remarkable achievement of our lives is that we learn this thing with infinite power. But the whole thing hangs on those sentences being built up out of their components, which are words. So it’s not even clear what one of the more striking theses in the development of linguistics over the last half century signifies or means without an account of the atoms, so to speak, out of which we build these things

We take it that Brandom’s sense of the geography is that our way of proceeding is more or less the first and his is more or less the second. But we think this way of describing the situation is both unclear and misleading, and we want to have this out right at the start. Our problem is that we don’t know what “you start with” means either in formulations like “you start with the content of words and proceed to the content of sentences” or in formulations like “you start with the content of sentences and you proceed to the content of words.” Brandom’s official view seems to be that he’s talking about explanatory priorities (see the preceding quote); but we think that can’t really be what he has in mind, and we can’t find any alternative interpretation that seems plausible. Speaking just for ourselves, we’re inclined towards a relatively pragmatic view of explanation; what explanation we should “start with” depends, inter alia, on what it’s an explanation of and whom it’s an explanation for. But, in any case, we would have thought that explanatory priority is of more than heuristic interest only if it reflects a priority of some other kind: ontological, semantical, psychological, or whatever. In talking about what one “starts with”, Brandom must be claiming more than

Following Aristotle (who himself was following Parmenides), philosophers have appealed to the distributional reflexes of expressions in determining their semantic status, and ultimately, the nature of the extra-linguistic world. This methodology has been practiced throughout the history of philosophy; it was clarified and made popular by the likes of Zeno Vendler and J.L. Austin, and is realized today in the toolbox of linguistically minded philosophers. Studying the syntax of natural language was fueled by the belief that there is a conceptually tight connection between the syntax of our language and its semantics, and the belief that there is a similarly tight connection between the semantics of our language and metaphysical facts about the world. We are less confident than our colleagues about the relation syntax has to semantics and metaphysics. In particular, we do not believe that the current status of theoretical syntax (or semantics or metaphysics) provides much support for either of the above two beliefs. We will illustrate our view with a case study regarding the status of complex demonstratives. We will show that a recent and particularly subtle syntactically based argument for the semantic/metaphysical status of complex demonstratives does not in fact show what semantic category complex demonstratives are..

It's an achievement of the last couple of decades that people who work in linguistic semantics and people who work in the philosophy of language have arrived at a friendly, de facto agreement as to their respective job descriptions. The terms of this agreement are that the semanticists do the work and the philosophers do the worrying. The semanticists try to construct actual theories of meaning (or truth theories, or model theories, or whatever) for one or another kind of expression in one or another natural language; for example, they try to figure out how the temperature could be rising compatibly with the substitutivity of identicals. The philosophers, by contrast, keep an eye on the large, foundational issues, such as: what's the relation between sense and denotation; what's the relation between thought and language; whether translation is determinate; and whether life is like a fountain. Every now and then the philosophers and the semanticists are supposed to get together and compare notes on their respective progress. Or lack thereof.

Compositionality is the idea that the meanings of complex expressions (or concepts) are constructed from the meanings of the less complex expressions (or concepts) that are their constituents.1 Over the last few years, we have just about convinced ourselves that compositionality is the sovereign test for theories of lexical meaning.2 So hard is this test to pass, we think, that it filters out practically all of the theories of lexical meaning that are current in either philosophy or cognitive science. Among the casualties are, for example, the theory that lexical meanings are statistical structures (like stereotypes); the theory that the meaning of a word is its use; the theory that knowing the meaning of (at least some) words requires having a recognitional capacity for (at least some) of the things that it applies to; and the theory that knowing the meaning of a word requires knowing criteria for applying it. Indeed, we think that only two theories of the lexicon survive the compositionality constraint: viz., the theory that all lexical meanings are primitive and the theory that some lexical meanings are primitive and the rest are definitions. So compositionality does a lot of work in lexical semantics, according to our lights.

A certain metaphysical thesis about meaning that we'll call Informational Role Semantics (IRS) is accepted practically universally in linguistics, philosophy and the cognitive sciences: the meaning (or content, or `sense') of a linguistic expression1 is constituted, at least in part, by at least some of its inferential relations. This idea is hard to state precisely, both because notions like metaphysical constitution are moot and, more importantly, because different versions of IRS take different views on whether there are constituents of meaning other than inferential role, and on which of the inferences an expression occurs in are meaning constitutive. Some of these issues will presently concern us; but for now it will do just to gesture towards such familiar claims as that: it's part and parcel of dog meaning dog2 that the inference from x is a dog to x is an animal is valid; it's part and parcel of boil meaning boil that the inference from x boiled y to y boiled is valid; it's part and parcel of kill meaning kill that the inference from x killed y to y died is valid; and so on. (See Cruse, Ch. 1 and passim.) IRS brings in its train a constellation of ancillary doctrines. Presumably, for example, if an inference is constitutive of the meaning of a word, then learning the word involves learning that the inference holds. If dog means dog because dog ---> animal is valid, then knowing that dog ---> animal is valid is part and parcel of knowing what the word dog means; and, similarly, learning that x boiled y ---> y boiled is part and parcel of learning what boil means, and so forth. IRS constrains grammatical theories. The semantic lexicon of a language is supposed to make explicit whatever one has to know to understand the lexical expressions of the language, so IRS implies that meaning constitutive inferences are part of the semantic lexical entries for items that have them. Lexical entries are thus typically complex objects (`bundles of inferences') according to standard interpretations of IRS..

It matters to a number of projects whether monomorphemic lexical items (‘boy’, ‘cat’, ‘give’, ‘break’, etc.) have internal linguistic structure. (Call the theory that they do the Decomposition Hypothesis (DC).) The cognitive science consensus is, overwhelmingly, that DC is true; for example, that there is a level of grammar at which ‘breaktr’ has the structure ‘cause to breakint’ and so forth. We find this consensus surprising since, as far as we can tell, there is practically no evidence to support it. (For example, there is no psychological evidence that you can’t have a word that expresses the concept BREAKTR unless you have the concept CAUSE. But there ought to be if CAUSE is a constituent of BREAKTR) This isn’t, of course, to say that there are no prima facie arguments at all for DC. The best one’s we’ve heard are the Impossible Word Arguments (IWA). That being so we’re very interested in whether IWAs are, in fact, sound.

It’s been, for some time now, a pet thesis of ours that compositionality is the key constraint on theories of linguistic content. On the one hand, we’re convinced by the usual arguments that the compositionality of natural languages1 explains how L-speakers can understand any of the indefinitely many expressions that belong to L. 2 And, on the other hand, we claim that compositionality excludes all “pragmatist” 3 accounts of content; hence, practically all of the theories of meaning that have been floated by philosophers and cognitive scientists for the last fifty years or so. A number of objections to our claim have been suggested to us, but none that we find persuasive (see, for example, the discussions of the “uniformity principle” and of “reverse compositionality” in Fodor and Lepore 2002). These objections have a common thread: they all grant that mental and linguistic content are compositional but challenge the thesis that compositionality is incompatible with semantic pragmatism. In this paper, we want to consider an objection of a fundamentally different kind, namely, that it doesn’t matter whether compositionality excludes semantic pragmatism because compositionality isn’t true; the content of an expression supervenes not on its linguistic structure4 alone but on its linguistic structure together with the context of its tokening. 5..

The idea that quotidian, middle-level concepts typically have internal structure -- definitional, statistical, or whatever -- plays a central role in practically every current approach to cognition. Correspondingly, the idea that words that express quotidian, middle-level concepts have complex representations "at the semantic level" is recurrent in linguistics; it's the defining thesis of what is often called "lexical semantics," and it unites the generative and interpretive traditions of grammatical analysis. Recently, Hale and Keyser (1993) have provided a budget of sophisticated and persuasive arguments for the claim that `denominal' verbs are typically derived from phrases containing the corresponding nouns: `singvtr' is supposed to come from something like DO A SONG; `saddlevtr' is supposed to come from something like PUT A SADDLE ON; `shelvevtr' is supposed to come from something like PUT ON A SHELF, and so forth.1 We think these are among the most persuasive arguments for lexical decomposition in the linguistics literature. Still, this paper is going to claim that they are finally unconvincing. In Part 1, we will show that there are quite serious arguments of a familiar kind against the decompositional analyses that Hale and Keyser (henceforth, HK) propose; in Part 2 we'll show that the arguments that HK offer in favor of their analyses are flawed.