Conversational implicatures as meant or communicated?

At this point it is important to reiterate that in introducing the notion of conversational implicature, Grice was proposing a normative apparatus, that is, a set of principled expectations, by which speakers could make what is implicated available to hearers (and in that sense speaker-meant or intended), rather than outlining how what is implicated is communicated per se (Saul 2002a). The Cooperative Principle and conversational maxims do not, therefore, refer to a set of rules for how speakers should speak, as they have sometimes been mistakenly interpreted. Instead, Grice was implicitly advancing a distinction between what a speaker is meaning versus what he or she is communicating, given that to mean nn something is for the speaker to have a particular set of intentions, while to communicate something depends in large part on the hearer’s interpretation (including interpretations of putative intentions on the part of the speaker). Some scholars have argued that the Cooperative Principle and conversational maxims can be used by hearers to figure out what has been implicated by the speaker.

However, Grice specifically formulated the Cooperative Principle and conversational maxims from the speaker’s perspective, as we saw above, and so they were evidently aimed at accounting for how speakers can make available meanings nn that are not said but rather implicated, rather than explaining the cognitive processes by which hearers infer such meanings nn.

This is also evident from the distinction Grice made between implicature (cf. implying) and implicatum (cf. what is implied) (Grice [1975]1989: 24). The term implicature was initially coined to refer to the process of intentionally meaning nn something in addition to what is said, and so concerned only the speaker’s meaning representation (i.e. what the speaker thinks he or she is implicating). The term implicatum, on the other hand, was used to refer to the “product” of this process, and thus involved the meaning representations of both speakers and hearers (i.e. what the hearer thinks the speaker has implicated, and what the speaker thinks he or she has implicated to the hearer). A conversational implicature (as opposed to a conversational implicatum), therefore, refers only to what the speaker intends to implicate, and in that sense involves making a particular meaning representation beyond what is said available to the hearer. It doesn’t encompass what is actually implicated to the hearer, or what the speaker thinks has been implicated to the hearer.

In sum, Grice arguably only focused on how a speaker makes available conversational implicatures (a kind of pragmatic meaning representation), but not on how they might be specifically understood (i.e. what is implied, namely, the conversational implicatum in question).3 In this way Grice accounted for the way in which speakers can be sure about what they intend to mean, but what they are taken to implicate by others is not necessarily consistent with those intentions. We will return to discuss the question of how hearers (and indeed speakers) reach an understanding of what has been implicated (i.e. conversational implicatum) in actual communicative interaction. However, at this point we now move to consider how Grice’s classic distinction between what is said and what is implicated has been fleshed out and altered in various ways. Most importantly, we will see that Grice’s assumption that what is said is primarily a semantic notion that only requires minimal pragmatic input is a highly questionable one.