Figurative meaning and the semantics/pragmatics divide Aim: to - - PDF document

1 Figurative meaning and the semantics/pragmatics divide Aim: to explore the models of meaning and communication given by minimalism vs pragmaticism, outline the account of figurative uses of language in each model and then raise some questions for pragmaticism. §§1-3 introduce the two models, §§4-5 examine the notion of explicatures, §6 elaborates a little on the minimalist model of communication. 1) The Gricean model Total signification of an utterance (a) What is said

• Compositional linguistic meaning
• Plus:

 Disambiguation  Reference assignment

• Propositional

(a) = what is asserted (semantics) (b) What is implicated

• various forms of implicature:

conventional, generalised conversational, particularised conversational

• typically (a) + conversational maxims

allow hearer to infer (b)

• (b) = what is merely implied (pragmatics)

Objections to the Gricean model: i. Fails to match on-line processing, e.g. ignores ‘direct access metaphors’, etc. ii. (a) doesn’t fit with intuitive judgements of what a speaker asserts: asserted content is

• ften pragmatically enhanced content.

iii. Problems with the maxims. 2) The Minimalist model Total signification of an utterance (a) What the sentence means Compositional linguistic meaning

• Plus:

 Disambiguation  Reference assignment

• Propositional

(a) = what is literally expressed (semantics) (b) What the speaker means

• usually pragmatically enhanced
• includes both minor and major

alterations to (a)

• ften indeterminate/multiple

propositions

• includes figurative meaning

(b) = what is conveyed (pragmatics) (a) and (b) = different types of meaning, underpinned by a different kinds of cognitive process. Literal meaning is computationally recovered within the language faculty, non-literal meaning is recovered abductively, outside the language faculty. Objections to the Minimalist model: i. Words and structure alone always/often/sometimes fail to yield propositions. ii. Model lacks psychological reality. iii. Ignores an important divide between different types of speaker meaning.

2 3) The Pragmaticist model Total signification of an utterance (a) Linguistically encoded content

• Compositional linguistic

meaning

• Always/often/sometimes sub-

propositional. (a) = linguistic semantics (b) explicit content of the utterance: explicature

• proposition(s) the speaker directly

communicates

• an expansion or development of

(a), involving reference assignment and certain free pragmatic effects (FPEs) (b) = what is asserted (c) what is implicated

• further relevant

propositions hearers can infer on basis of (b) (c) = what is merely implied 3.1) What is an explicature? i) A pragmatically inferred development of logical form (where implicatures are wholly pragmatically derived); S&W 1986: 182, Carston 2009: 41 ii) What the speaker intends directly to communicate; S&W 1986: 183, Carston 2009: 36 iii) The first content hearers recover via relevance processing; S&W 1986:184-5 iv) The essential premise for inferring further (implicated) propositions; Carston 2009: 41 v) The proposition on which S’s utterance is judged true or false; Carston 2009: 36 A: ‘How was the party?’ B: ‘There was not enough drink and everyone left’ Explicature: there was not enough alcoholic drink to satisfy the people at [the party]i and so everyone who came to [the party]i left [the party]i early. Implicature: the party was no good. (Carston 2009: 35) Figurative uses: On the RT model there exists a continuum of loose uses, with minor alterations at one end of the scale and metaphor at the other end (or perhaps involving a special, meta-representational process; Carston). Irony is off the scale. “Robert is a computer” Explicature: Robert is a computer* Implicatures: Robert lacks feelings, processes information well… (Wilson 2011: 180) Question: how do we individuate explicatures? 4) Which free pragmatic effects are explicature-generating? 4.1) How many kinds of FPEs? Traditionally theorists have posited two different kinds of FPEs: (i) modulation and (ii) unarticulated constituents (UCs).  Why do we need both modulation and UCs? E.g. why treat the location in an utterance of ‘It’s raining’ as an unarticulated constituent as opposed to allowing the meaning of ‘rain’ to be modulated (broadened or narrowed)?  Carston and Hall 2012 reject a modulation treatment of the weather predicates, but they don’t give an argument for this.  Since it is unclear what the constraints on broadening/narrowing of senses are, it is unclear whether there remains any role for UCs to play on the pragmaticist model. Without UCs pragmaticism may come to seem more closely aligned to Travis’

• ccasionalism than it perhaps once did (undermining the idea of ‘developments of LF’?).

3 4.2) Which FPEs generate explicatures?  One of the on-going challenges to any account which wants a special class of free pragmatic effects concerns how to restrain them: what stops ‘snow is white’ directly expressing the proposition ‘snow is white & 2+2=4’?  In the past, theorists have appealed to: the Availability test (Recanati 1989: 309-10), truth-evaluation tests (Recanati 2004: 15, Noveck et al), the Scope test, and the mechanism of relevance (Carston 2002). However all face problems (see Carston and Hall 2012).  For Hall 2008, Carston & Hall 2012 the ‘ultimate arbiter’ of whether or not a pragmatic enrichment contributes to the proposition literally expressed is “the derivational distinction between local and global pragmatic inference” – an effect which modifies a subpart of the linguistically encoded meaning counts as part of the explicature, one which operates on fully propositional forms contributes to implicatures.  However, it’s not clear that locality will work as some local enrichments apparently capture implicature content. A: Do you want to have dinner? B: I’m going to the cinema. How should B’s utterance content be modulated?

• Narrowed from GOING-TO-THE-CINEMA to GOING-TO-THE-CINEMA-

TONIGHT

• Narrowed from GOING-TO-THE-CINEMA to GOING-TO-THE-CINEMA-

AT-A-TIME-THAT-MAKES-HAVING-DINNER-WITH-A-IMPOSSIBLE  Both of these are local effects, but only on the first will it be an implicature of what B says that she cannot have dinner with A (on the second it looks like something she directly asserts).  Perhaps the problems with constraining FPEs is instructive. Perhaps we should simply reject the idea that there is a special category of ‘pragmatic developments of logical form’ (i.e. explicatures) which play an inferential role in recovery of other propositions. 5) Do we need the notion of an explicature? i) Explicatures (as developments of LF) need not be psychologically real for speakers. Thoughts are just as underdetermined as utterances (if S utters ‘pass me the red pen’ I think she need not, contra Wilson 2011: 181, have internally specified that she wants the pen that writes in red, the one that contains red ink, the one that is red on the lid, the one that says ‘red’ on it, etc; compare ‘I want to travel to London’.) ii) Explicatures need not be psychologically real for the audience: all hearers may consciously entertain is ‘implicature’ content. iii) As theorists explicatures can play a role in a rational reconstruction of a route from literal meaning to conveyed content, but why should we think hearers must or even typically do follow this route, either explicitly or implicitly? iv) Soliciting judgements of truth/falsity for utterance content is inappropriate, it influences the phenomena it is supposed to be uncovering (‘quantum effect’): Under what circumstances is B’s utterance above true?  there wasn’t enough alcoholic drink and so everyone at the party left early But: Is the utterance true in a situation where there was a lot of wine at the party but it was held in a locked cabinet? And is it required that everyone left early in the evening or early for a party?  there wasn’t enough easily available alcoholic drink and everyone at the party left after

• ne hour

4 But: Is it true in a situation where there was plenty of crème de menthe available at the party, or where those hosting the party didn’t leave?  there wasn’t enough available and attractive alcoholic drink and everyone who came to the party as the result of an invitation left early. In asking these questions we prompt the audience to sharpen the original content of the utterance in various ways, but these moves reflect decisions about how to sharpen content not an uncovering of material which is already present.  Conclusion: there may not be a workable notion of ‘a development of logical form’ here (§4), nor do we need the notion of an explicature to explain what is in the speaker’s mind, or the hearer’s mind, or to explain our judgements of T/F. Perhaps, then, the traditional picture where we just have what the sentence means and what the speaker means, where speaker meaning can diverge to a greater or lesser extent from sentence meaning, is to be preferred. 6) Minimalism and communicated content Socio-linguistic structures allow a direct move from literal meaning to communicated content (it’s not understanding a language which requires understanding a way of life but understanding a speaker).  Assertion can go either with what is literally expressed (legal discourse: ‘use a firearm’) or what is pragmatically conveyed.  Speakers can demur from any content attributed on the non-literal side without contradiction, but they do not have final veto on what is communicated (consider libel cases).  There is a threshold of tolerance within which differences between the propositional content assigned by speaker and hearers as communicated content can simply be

• ignored. The boundaries of tolerance are set in context, depending in part on the kind of

conversational exchange in which agents are engaged.  If a proposition risks falling outside the contextually determined limits of tolerance, there are two options: (i) the speaker/hearer puts her interpretation on the record (‘are you saying…?’/‘I mean that…’) and interlocutors negotiate over the content to be added to the conversational record or (ii) the potential divergence between speaker and hearer’s take on the communicated content goes unnoticed and communication fails.  It is in this sense that all communicated content is ‘an expedition abroad’, no longer part

• f linguistic understanding per se but instead relying on a culturally entrenched network
• f beliefs and practices. Metaphor may be a more exotic trip but all non-literal content is

a journey beyond linguistic understanding.

e.g.n.borg@reading.ac.uk June 2013