The Linguist is a languages magazine for professional linguists, translators, interpreters, language professionals, language teachers, trainers, students and academics with articles on translation, interpreting, business, government, technology
Issue link: https://thelinguist.uberflip.com/i/691128
JUNE/JULY 2016 The Linguist 23 thelinguist.uberflip.com FEATURES context-sensitivity of language: the fact that linguistic meaning depends in all sorts of ways on features drawn from a context of utterance. In recent philosophy of language, this worry about context-sensitivity has taken on a new form and, to many, it delivers a decisive blow to Formalist approaches. Context: a blow to Formalism? So what exactly do we mean when we say that linguistic meaning depends on contextual factors? If we had asked this question in the 1970s, the philosopher's answer would have focused on demonstratives (words such as 'this' and 'that'), indexicals (e.g. 'I', 'now', 'today') and tense markers. For these items, it is unarguable, even for a Formalist, that some appeal to a context of utterance must be made in order to recover their content. So, for instance, in order to recover the proposition expressed by 'I am happy', one needs to find out from the context of utterance who the speaker is and when they are speaking; only then can one arrive at a content which is truth-evaluable (i.e. which can be assessed as holding or not). Yet it is this property of truth-evaluability which traditionally has been the hallmark of semantic content (the kind of literal linguistic content Formalists aim to model), so these expressions reveal that even Formalists must admit some role for contextual factors in determining semantic content. This kind of context-sensitivity, however, has been thought relatively unthreatening to the Formalist project, because it looks as if one can devise fairly simple formal rules that specify exactly when contextual features must be looked to (i.e. when one of the key words appears) and which features to look to (e.g. 'I' refers to the speaker). 1 Though this point has been much debated in recent work, the context-sensitivity exhibited by these kinds of items seems rule-governed in a way that is potentially amenable to being captured within a Formalist approach. A second way in which meaning can depend on context was brought to prominence by Paul Grice. In one of his classic examples, Grice imagines a professor who is asked to write a reference for her student, Smith, who is applying for a philosophy job. She writes only 'Smith has nice handwriting'. As Grice notes, such an exchange conveys a wealth of information (such as Smith lacks philosophical talent) that we do not countenance as part of the semantic content of the sentence. There is undoubtedly, then, a rich role for context to play in specifying the content a speaker manages to convey by an utterance. Yet the Formalist's assumption is that this kind of contextual influence is best viewed as part of a post-semantic, pragmatic speaker-meaning, rather than contributing to an account of linguistic meaning in and of itself. Again, then, the Formalist project has been held to be compatible with this kind of context-sensitivity. Recently, however, so-called Contextualists have argued for the rejection of Formalism on the grounds that contextual influences on meaning cannot be limited to reference fixing for demonstratives, indexicals and tense markers, and a post-semantic determination of a speaker's meaning. They argue that the propositional content literally expressed by many, most or even all utterances of natural language can only be recovered with rich contextual input. Consider the following: 1 Jill is ready. 2 Jill is a sailor. 3 The apple is red. 4 It is raining. None of these sentences contain standard indexicals or demonstratives, but still the Contextualist argues that we need to make a rich appeal to a context of utterance in order to work out what proposition the speaker expresses. In 1), we cannot arrive at a truth-evaluable proposition until we know what Jill is ready for. An utterance of 2) might be true in one context (when talking about hobbies, say) even though it is false in another (e.g. when asking about paid occupations), even though the facts of the matter about Jill and her sailing remain unchanged. In 3), we need to know whether the apple is being said to be red on the outside, the inside, red all over, red-to-degree-n or n+1, etc. Finally, although no syntactic element in 4) apparently marks a location, the speaker standardly conveys that it is raining at some specific place. These phrases, then, although they do not appear context-sensitive on the surface, apparently reveal, on closer inspection, that the truth-evaluable content they express depends in complex ways on the context in which they are uttered. The Contextualist concludes that these phrases are semantically context-sensitive and, furthermore, what goes for these phrases goes for all or almost all other sentences. Finally, the Contextualist contends that this more covert context-sensitivity is not susceptible to analysis by simple formal rules: there is no simple rule telling us when we will need to look to a context of utterance, nor which bit we should look to. For instance, a speaker uttering 4) might intend to refer to their own location, another location, or even to no location at all (as when simply looking at a precipitation recording device). According to Contextualists, then, these kinds of cases, together with the pervasive nature of metaphor and polysemy, reveal that it is not possible to specify an appropriate truth-evaluable content for linguistic items without paying full and careful attention to the context in which those items are produced. Semantic content is underspecified by the formal features of sentences, they conclude, so Formalism is fundamentally compromised. SCIENTIFIC HOPE Could neuroscience hold the answers that have so far eluded the philosophers? It is not a debate that linguists can afford to ignore as a mere turf war in the philosophy of language