Kent Bach

I’ve known about conversational implicature a lot longer than I’ve known Larry. In 1967 I read Grice’s “Logical and Conversation” in mimeograph, shortly after his William James lectures, and I read its precursor “(Implication),” section III of “The Causal Theory of Perception”, well before that. And I’ve thought, read, and written about implicature off and on ever since. Nevertheless, I know a lot less about it than Larry does, and that’s not even taking into account everything he has uncovered about what was said on the subject long before Grice, even centuries before. So, now that I’ve betrayed my ignorance, I’ll display my insolence. I’m going to identify the most pervasive and pernicious misconceptions about implicature that I’ve noticed over the years.

Language is used to express thoughts and to represent aspects of the world. What thought a sentence expresses depends on what the sentence means, and how it represents the world also depends on what it means. Moreover, it is ultimately arbitrary, a matter of convention, that the words of a language mean what they do. So it might seem that what they mean is a matter of how they are used. However, they need not be used in accordance with their literal meanings. One can speak nonliterally, and convey something other than what the sentence means ('The look on his face spoke volumes'), or speak indirectly, and convey something more than what the sentence means ('I wonder if you know the time'). Linguistic communication requires knowledge of linguistic meaning, on the part of both the speaker and his audience, but it requires extralinguistic knowledge as well.

Puzzles about sentences containing expressions of certain sorts, such as predicates of personal taste, epistemic modals, and ‘know’, have spawned families of views that go by the names of Contextualism and Relativism. In the case of predicates of personal taste, which I will be focusing on, contextualist views say that the contents of sentences like “Uni is delicious” and “The Aristocrats is hilarious” vary somehow with the context of utterance. Such a sentence semantically expresses different propositions in different contexts, depending on what standard or perspective (or whose standard or perspective) is implicitly adverted to. According to relativist views, the propositional content of such a sentence is fixed, but what it takes for that proposition to be true varies somehow with the context, depending on the relevant standard or perspective. I will argue that such views are neither well supported by the data nor well motivated by the puzzles themselves. Even so, there is an element of truth in each. I will sketch an alternative view, dubbed Radical Invariantism, according to which the appearance of context sensitivity is illusory. Rather than impute either kind of context sensitivity to these sentences or to their contents, Radical Invariantism says that these sentences are distinguished by what they don’t do. Because they are not explicitly relativized, they leave a certain semantic slack. They fall short of fully expressing a proposition, instead expressing merely a “propositional radical.” We can explain away the appearance of semantic context sensitivity pragmatically, by taking into account facts about how, and under what conditions, speakers who use or encounter these sentences manage to pick up the slack. This can occur in either of two ways. Speakers either take a certain standard or perspective as understood, or else they treat the sentence as if it expresses a standard- or perspective-independent proposition even though it does not..

Once upon a time it was assumed that speaking literally and directly is the norm and that speaking nonliterally or indirectly is the exception. The assumption was that normally what a speaker means can be read off of the meaning of the sentence he utters, and that departures from this, if not uncommon, are at least easily distinguished from normal utterances and explainable along Gricean lines. The departures were thought to be limited to obvious cases like figurative speech and conversational implicature. However, people have come to appreciate that the meaning of a typical sentence, at least one we are at all likely to use, is impoverished, at least relative to what we are likely to mean in uttering it. In other words, what a speaker normally means in uttering a sentence, even without speaking figuratively or obliquely, is an enriched version of what could be predicted from the meaning of the sentence alone. This can be because the sentence expresses a “minimal” (or “skeletal”) proposition or even because it fails to express a complete proposition at all.1 Indeed, it is now a platitude that linguistic meaning generally underdetermines speaker meaning. That is, generally what a speaker means in uttering a sentence, even if the sentence is devoid of ambiguity, vagueness, or indexicality, goes beyond what the sentence means. The question is what to make of this Contextualist Platitude, as I’ll call it. It may be a truism, but does it require a radical revision of the older conception of the relation between what sentences mean and what speakers mean in uttering them? Does it lead to a major modification, or perhaps even outright rejection, of the semantic-pragmatic distinction? I think..