GRICE’S THEORY OF CONVERSATIONAL IMPLICATURE

INTRODCTION

In this article, we shall be considering some more concepts associated with Grice’s theory of Conversational Implicature and how they enable us to understand better how speakers and hearers are able to communicate effectively. We shall also consider some reactions by scholars to the Gricean notion of implicature.

Types of Conversational Implicature

Grice identifies two types of implicature (i) generalized conversational implicature (ii) particularized conversational implicature. Generalized conversational implicature occurs irrespective of the context. For example:

(i)                Some Zambians are Muslims

(ii)              5 litres of fuel starts my engine

The two statements above give rise to the same generalized implicatures regardless of the context they occur. And they remain implicatures rather than entailment because in statement 

(i) some Zambians are Muslims, it is clear that the statement may be denied. The implicature is that not all Zambians are Muslims; in fact we have more Muslims in the north than there are in the south. Statement 

(ii) 5 litres of fuel starts my engine, may as well be denied because the statement didn’t say that 5 litres is all my engine needs to start. The engine actually requires 70 litres. The case of the generalised implicature is that the same inference,

(iii) not all Zambians are Muslims and that my engine needs more than 5 litres to run, are the most likely irrespective of the context.

However, statement (i) above may give rise to other forms of implicatures which depends on the context. For example some Zambians are Muslims, while some are Christians; some are neither Muslims nor Christians, some are traditional religionists etc. Similarly for statement (ii) someone might even conclude that less than 5 litres may start my engine or more etc. Because these implicatures depend on the context of use, Grice calls them ‘particularized implicature.’ A particularized implicature is different from the generalized implicature that is associated with words like some since they are the inferences we need to make as they relate to some particular contexts.

You will recall that one of the conversational maxims is relation or relevance, i.e. make your contribution relevant to the conversation/context. If all implicatures were particularized, one can reasonably argue that the maxim of Relation (relevance) is enough to account for all implicatures, because the implicature would be what the addressee has to assume to render the utterance relevant to the context. 

But generalized conversational implicature does not rely on how relevant an utterance is to a context, rather on quantity (maxim of Quantity) and manner (maxim of Manner). When a speaker uses the word ‘some’ it is because s/he is not in position to use the word ‘all’ and is therefore taken to imply ‘not all’ by the Maxim of Quantity. This is explained in the figure below:

Zufferey, S., Moeschler, J., & Reboul, A. (2019). Generalized Conversational Implicatures: Gricean, Neo-Gricean and Post-Gricean Pragmatics. In Implicatures (Key Topics in Semantics and Pragmatics, pp. 111-140). Cambridge: Cambridge University Press. doi:10.1017/9781316410875.006

The above data gives rise to what is known as ‘scalar implicature.’ According to Gazdar (1979) implicature therefore operates with scales, so that one scale would include ‘some’ and ‘all’ and another ‘do brilliantly’ and ‘make progress.’ What this means is that if you take any item on a scale, the items above or below it is automatically excluded. In other words you cannot choose ‘some’ and ‘all’ at the some time. It you choose ‘do’ you automatically exclude ‘make’. Gazdar gave other hypothetical scales as <certain…probable…possible><and…or> and <must.. .may.. ,might> (Grundy, 2000). This explains why you are not likely to get confused if I ask you:

(iii)    Would you like Coke or Fanta?

Of course you know that I’m not asking you to choose both. My choice of ‘or’ has excluded the possibility of ‘and’ so you’re sure I’m saying it’s either Coke or Fanta and not both. By the notion of scalar implicature, because ‘or’ is on the scale below ‘and’ a speaker selecting ‘or’ (as I have done) would be implying ‘not and.’ Thus either Coke or Fanta or both is an entailment and either Coke or Fanta but not both is an implicature (Grundy, 2000). If you listen to people converse, you will notice that they apply the notion of these scales without even realizing it.

Non-Conversational Implicature

Another term for describing ‘non-conversational implicature’ is ‘conventional implicature’ which according to Levinson 1983:127 is the ‘non-truth conditional inferences that are not derived from superordinate pragmatic principles like the maxims, but are simply attached by convention to particular lexical items or expressions.’ Let me illustrate with this advert by the First Bank of Zambia:

(iv)    First Bank: truly the first.

The lexical item ‘truly’ is a conventional implicature because (i) the word ‘truly’ has a general meaning or entailment, i.e. certainly, in actual fact, or factually (ii) the word also conveys an implied meaning or implicature like this may be contested, or this is not really true. So the word ‘truly’ is conventional because it is closely associated with particular lexical items, i.e. ‘the first.’ While it is the first bank in Zambia, the second idea of ‘first’ is the implicature though conventional because we can’t really say that it is ‘truly’ the first or not.

Explicature

The term “explicature” was first used by Sperber and Wilson in their book titled: Relevance: Communication and Cognition, 2^nd edition, (1995). They argue that the single principle of relevance is enough to explain the process of utterance interpretation and understanding.

They replaced the Gricean notion of implicature (a non-conventional meaning recovered by making some inferences) with a two-stage process in which the hearer recovers first an explicature which is an inference or series of inferences that enrich the under-determined form of the utterance to a full propositional form, and then an implicature – an inference which provides the hearer/reader with the most relevant interpretation of the utterance. We can then say that explicature is an enrichment of an original utterance to a fully elaborated propositional form. Look at the following examples:

(i)                First Bank: truly the first

(ii)              Limca: 1^st for taste,

The above adverts may be explicated as follows: (i) First Bank is the first indigenous Zambian Bank, and has proved to be the first in terms of its highest standard of customer service delivery (ii) Limca soft drink is first among all others for taste. Sperber and Wilson believe that Gricean conversational implicature leaves the addressee with too many probabilities and therefore propose a Relevance theory that goes beyond these probabilities to enable addressees to be sure that they have recovered the most relevant of a all possible set of inferences. 

If an addressee is able to recover the explicature of a proposition, it becomes easier for him/her to make the right inference. Sperber and Wilson also identified a “higher level explicature” which seeks to reveal the propositional attitude of the speaker to his/her utterance. In other words, the speech acts description for the utterance. This means that even where an utterance is explicit enough (may be associated with an explicature) there is still a higher level explicature which the addressee needs to recover. Speech acts are therefore treated as attitude to propositions rather than as actions.

It is argued that explicature (i.e. inference/series of inferences that enrich/elaborate the under-determined form…to a full propositional form) are motivated by the indeterminacy of language . This indeterminacy is as a result of the economy of expression which characterizes natural language. A lot of expressions may represent or mean other things which require inferential process to be able to arrive at their full interpretation. Even where utterances with straightforward grammatical relations are made, there may still be some possible semantic relations that may be inferred with different uses of the expression. So utterances require some degree of enriching or fleshing for the most relevant inference to be made about their meaning.

CONCLUSION

Grice’s conversational implicature stems from the fact that people communicate meaning or assumptions explicitly (clearly, overtly or plainly) or implicitly (totally completely though often indirectly). What is said or entailed (what our words mean) is distinguished from what is implicated (what we mean by using some particular words). So implicatures are simply a matter of more being communicated than said. For a conversational implicature to take place it is assumed that some cooperative principles are in operation, which of course are usually violated.