Herbert Paul Grice was born in Birmingham on 15 March 1913 and died in Richmond, California on 28 August 1988. He began his formal philosophical studies with W.F.R. Hardie at Corpus Christi College, Oxford. Later he held various positions at Oxford, and served in the Royal Navy during World War II. In 1967 he moved to California, taking a position at the University of California, Berkeley. He retired from Berkeley in 1979, but continued to teach afterwards until 1986.
Grice is best known for his work on meaning, in particular for his work on the pragmatic dimensions of meaning. He gave an analysis of what it is for a person to mean something by his or her words and gestures, where that might depart from what the words literally or standardly mean, and offered a theory of the phenomenon of conversational implicature. He wrote on reasons and reasoning, perception, value, justice and happiness, sometimes engaging with classical philosophers. Grice can be characterized as a systematic philosopher. He appealed to the notion of a speaker’s intentions in communication to explain meaning, and further explored the role of intentions in reasoning. His interest in the pragmatic elements of meaning seems to have been sparked by his reactions to some applications of ordinary language philosophy, a movement in which he participated early in his career.
Those works that have had the greatest impact include his early article ‘Meaning’ published in 1957 (drafted in 1948), and his William James lecture series, ‘Logic and Conversation’ given at Harvard University in 1967. One part of the lecture series was published under the same title in 1975. Other parts were published as ‘Utterer’s Meaning, Sentence-Meaning, and Word-Meaning’ in 1968, and ‘Utterer’s Meaning and Intentions’ in 1969. In 1989 many of Grice’s works, including a revised version of the original William James lecture series, and a seminal paper on perception, were published as Studies in the Way of Words. Posthumous works include Aspects of Reason (2001) and The Conception of Value (1991). Grice published under two different names: H.P. Grice in his early work, and Paul Grice in his later work.
In his early article ‘Meaning’ Grice distinguishes between natural and non-natural signs. He then turns to an analysis of speaker’s meaning (or utterer’s meaning, as he called it). An example of a natural sign is the smoke of an accidental fire. Another example is the spots that result from measles. What the spots mean – viz. measles – is not the sort of thing that is true or false and it would be unreasonable to infer that some person means anything by producing measles spots. Non-natural signs, on the other hand, have the sort of meaning attributed to the ringing of some bells when someone says ‘those three rings mean the bus is full’. Upon hearing the bells it would be perfectly reasonable to infer that someone (the bus driver) means something by the three rings, and it is perfectly possible for the ringing of the bells to mean the bus is full while in fact the bus is empty – the bus driver can in that way mislead the public.
Grice’s analysis of what it is for a person to mean something by an utterance (ringing a bell, making a hand gesture, giving voice to a sentence) adverts to intentions and other psychological states of the utterer. The thought that intentions might be involved somehow in determining what a person means is a natural one. Speakers have many intentions when speaking with another person. One of these intentions determines which proposition the speaker means. What Grice did was to show how hearers are able to isolate this intention. On Grice’s analysis it is true to say that someone meant something by uttering X just in case it is true that this person had an intention of a special kind to produce a certain effect in the hearer. Crucially, the kind of intention in question is an intention to produce an effect by means of the audience’s recognition of the speaker’s intention.
The kind of intended effect varies with the kind of speech act the speaker is performing. For example (on Grice’s 1957 analysis), if a speaker A is performing the act of making a statement, the intended effect is for the audience to believe what is stated. Later, Grice modified this analysis: the intended effect is for the audience to believe that the speaker believes what is stated. By contrast, if the speaker is issuing a command, the intended effect is not to get the audience to believe anything, but to get the audience to do something (Grice’s 1957 analysis) or to get the audience mentally geared up for action, so that he forms an intention to do something (1969 analysis).
It is important to keep in mind that though a speaker might intend all kinds of things in producing an utterance, not all of what a speaker intends is included in the specification of what he has said or meant. Some of what he intends he may positively not want the speaker to recognize. Only those effects that he wants the audience to recognize as intended effects enter into the specification of what the audience means. As mentioned, part of the relevant speaker’s intention is that the effects come about because of the audience’s recognition of the speaker’s intention. The effects must come about by a particular causal route, one involving the audience’s ability to employ his recognition of the speaker’s intention as a reason for entering into a change of psychological state. The importance of this last point came to the fore after Grice responded to a number of putative counter-examples to his analysis of speaker meaning. Cases can be contrived in which a speaker intends to induce a belief in the audience by way of the audience’s recognition of his intention, but still has not said something (or said that which, according to Grice’s analysis, he is supposed to have said). Often the way the counter-examples work is by the invention of a quirky causal route to the intended effect, so that the audience’s recognition of the speaker’s intention does not enter in as a reason in the right sort of way for the audience to enter into the intended psychological state.
Besides speaker meaning, there is the question of what it is for marks, gestures or sounds themselves to mean something on an occasion, and across occasions of use. Grice held that speaker meaning was the fundamental notion, and the meaning of an utterance type the derivative notion. Initially Grice proposed that in the case of statement making, an utterance X timelessly means that p if and only if what people generally intend to effect by X is the belief that p. Later Grice explicitly acknowledged that an audience often can and will exploit the standard meaning of the utterance type in forming his beliefs about what the speaker intends on a specific occasion (and that this is something the speaker would take into account in forming his communicative intentions). He nevertheless continued to maintain that speaker meaning is the fundamental notion. One may wonder whether the fact that an utterance has a standard meaning is ever necessary for the reasonableness of the speaker’s belief that his utterance will produce the intended effect on an occasion, and if so, whether this does damage to Grice’s claim concerning the priority of speaker meaning. Further worries arise for Grice’s overall explanatory strategy for the central case of sentence meaning, where the utterance type is a sentence type, rather than, say, a hand gesture. The meanings of sentences are arguably composed of, or determined by, the meanings of their parts, and speakers can, in theory at least, recover the meanings of sentences that would never be uttered under any conditions. Grice explores some of the issues in ‘Utterer’s Meaning, Sentence-Meaning and Word-Meaning’.
A speaker may exploit the standard or conventional meaning of a sentence to say something using that sentence, and he may, in a different way, exploit other conventions, to implicate something that is unsaid. In ‘Logic and Conversation’ Grice distinguishes between conventional implications of what is said, for example its formal entailments, and conversational implicatures of the saying of what is said. One of Grice’s examples of implicature is that of a man who is asked how some third person is getting along in his new job, and who responds: ‘Quite well, I think; he likes his colleagues, and he hasn’t been to prison yet’, implicating perhaps that the man is dishonest and tends to get into trouble.
Conversational implicatures are possible because there are certain general principles that govern conversations. Grice proposes that a ‘Cooperative Principle’ governs conversations, to wit ‘Make your conversational contribution such as is required at the stage at which it occurs, by the accepted purpose or direction of the talk exchange in which you are engaged.’ From this general principle, more specific maxims emerge, such as ‘avoid ambiguity’ or ‘make your contribution as informative as required’ or ‘be relevant’. Grice’s idea is that when the speaker openly flouts such maxims, he is exploiting the maxim conversationally to implicate something. What is implicated is something that the audience could work out by reference to the maxims that are flouted if he assumes that the speaker is only flouting the maxims because he is trying to implicate something. For the example given earlier, the maxim ‘be relevant’ is flouted, yielding the implication that dishonesty is a concern regarding the person discussed.
Underlying conversational principles like these are operative in all kinds of conversational contexts, including philosophical ones. In the context of a discussion of sense data, in ‘The Causal Theory of Perception’, Grice points to pragmatic implications of remarks like ‘The apples look red.’ If uttered in a context in which the lighting is poor, or other unusual conditions, then such a remark would be informative and relevant as a description of how the apples look under those conditions. On the other hand, if uttered in perfectly normal conditions, normal-for-viewing-apples conditions, the statement would need an explanation. ‘Why is the speaker stating the obvious?’, the audience might ask himself, searching for the implicature. According to the view that Grice is here disputing -a view put forward by some ordinary language philosophers of Grice’s time – in the second sort of case the speaker would be guilty of misusing words. His statement would be neither true nor false, but meaningless. For Grice, the statement could be true, though trivial, and not maximally informative. The mistake of his interlocutor is to treat the pragmatic presupposition – that the speaker has some reason to be doubtful about the apples – as part of the statement’s meaning. Grice says that this pragmatic presupposition is not a part of the meaning of the expression any more than ‘the speaker believes it is raining’ is part of the meaning of the statement ‘it is raining’. The oddity of the statement shows that something is being implicated, not that the statement is meaningless.
Grice’s work on conversational implicature gave rise to a whole new sub-area of linguistics. He succeeded in putting the phenomenon of speaker meaning into the foreground in philosophical studies of meaning. His programme for explaining the meanings of conventional signs, and in particular the meanings of linguistic items, in terms of intentions in the minds of speakers, on the other hand, faces obstacles. Even so, the idea of a two-stage reduction of linguistic meaning to speaker’s meaning, and speaker’s meaning to intentions and beliefs of the participants in the exchange, informs the work of many philosophers of mind and cognitive scientists fifteen years after his death. If speaker’s meaning itself is to be analysed in terms of a speaker’s and an audience’s intentions and beliefs, that brings with it an ontological commitment to intentions, and other representational psychological states. Philosophers today working within a Gricean framework must face the question whether such notions are empirically respectable. Grice himself was liberal about these things; he was concerned to discover the purely philosophical grounds of rationality, not with the shape that an empirically well-grounded theory of mind must take. A second fundamental concern for those working within Grice’s framework is the question whether language is indispensable to thought in some way, perhaps as the very medium of thought. If this is the case then it looks like the intentions and beliefs in terms of which linguistic meaning is supposed to be explained are not after all more fundamental than language. In any case, Grice’s general take on language, according to which it at root involves people attempting to affect other people in certain ways, is distinctive and rich in its consequences.
‘Logic and Conversation’, in Peter Cole and Jerry Morgan (eds), Syntax andSemantics, vol. 3, Speech Acts (New York, 1975), pp. 41–58 ; and in Donald Davidson and Gilbert Harman (eds), The Logic of Grammar (Encino, 1975), pp. 64–75.