Implicature

Implicature is a pragmatic relation distinct from the semantic relation of logical entailment, but the two are frequently confused. The implicatures of a sentence go beyond what the sentence logically entails. H. P. Grice was the first to identify and explain implicature. For a further description of the concept, see the Wikipedea definition of implicature. Grice identified and explicated two kinds of implicature: conversational and conventional. Grice claimed that ordinary conversation is governed by some general rules or maxims: 1) be informative, 2) be truthful, 3) be relevant, 4) be clear. According to Grice these four principles amount to one overarching principle "Be helpful." Because conversation is governed by these norms, we can deduce information not just from a statement's content, but, from its conversational implicature--that is, what it communicates beyond the mere assertion. How does conversational implicature work? Let's look at an example: I ask a friend about what my blind date is like. She responds, "He has a nice personality." The content of her statement is only that my date has a nice personality. But, from that statement I might deduce information that is not asserted or entailed, that is, that my date is not very attractive. My friend has violated Gricean norm of informativeness to great effect. Of course, my friend can also cancel the implicature by saying, "He's also really handsome." The feature of cancelability is common to all implicatures and serves as a means of distinguishing them from logical implications. Conventional implicaure functions in a slightly different way. Conventional implicatures, again, go beyond the entailments of a sentence, but, arise from the meaning of the sentence, not conversational norms. For instance, the sentence, "She was fat, but pretty," does not entail contrast between fatness and prettiness. But, due to the meaning of the word "but," the sentence conventionally implicates that there is such a contrast.

Grice and The Equvalence Thesis

H. P. Grice maintained that A⊃C is logically equivalent to A→C. This equivalence thesis has been subject to putative counterexamples. Such counterexamples exploit the fact that according to the equivalence thesis, all conditionals with false antecedents and all conditionals with true consequents are going to come out true. For instance, "If the president isn't named George, then there are no people in the world named George," and "Bill Clinton served 100 years in the military if he never served in the military," are both true by the equivalence thesis. But, these conditionals seem absurd.

Grice uses the concept of implicature to explain why conditionals of the above sort appear ridiculous, despite being true. Conditionals like those mentioned above violate the norms of brevity and perspicuity, and are, therefore, unassertable. It is ridiculous to say, "Bill Clinton served 100 years in the military if he never served in the military," if you can simply say, "Bill Clinton never served in the military."

Problems With the Analysis

Frank Jackson points to some serious flaws in the Gricean defense of the equivalence thesis. Jackson contends that his objections defeat Grice's analysis. Let's look at the following three issues:

1. I might want to assert A→C when I am fairly confident that A is false. Consider the indicative, "If I am struck by lightning 16 times today, I will die." The conditional seems assertable, but it violates the Gricean maxim of brevity.
2. Similarly, I might want to assert A→C when I am fairly confident that C is true. Consider the indicative, "If you are hideously disfigured, I will love you." Again, this seems assertable, but by Grice's lights, it ought to look as bad as, "I have 1000 children if I don't have any children."
3. Grice's theory gives us no means of distinguishing between logically equivalent sentences. Consider the following two sentences, "I will do my assignment, but if I don't, fairies will dance," and, "I will do my assignment, but if I don't, I'll fail the course." If the equivalence thesis is correct, then these two sentences must be logically equivalent. They both take the form A&(¬A⊃B). On Grice's view, these sentences ought to be on a par in terms of their assertablility (or lack thereof). But intuitively, we want to make distinctions between them.

Grice's use of the concept of conversational implicature to defend the equivalence thesis fails in light of Jackson's criticisms. Nonetheless, we continue explain many linguistic phenomena in terms of conversational and conventional implicature.