H. P. Grice

H. P. Grice took on various philosophical topics, including personal identity, perception and the senses, intention, reason, and value, but he was and is best known for his work on meaning and what he called the conversational 'implicatum.'

Grice's approach to the 'explicatum' and the 'implicatum' has had lasting impact on debates about such questions as which kind of meaning, linguistic or 'utterer's meaning,' is more fundamental and on how and where to draw the line between, to use Morris's terminology, semantics and pragmatics.

Most influential are his analysis of utterer's meaning and his account of the conversational implicatum.

Grice’s ingenious notion of reflexive intention is designed to capture what is distinctive about what we intend when we communicate, and his account of the conversational implicatum aims to explain how we can say one thing and manage to communicate something else.

Effectively repudiating Wittgenstein’s famous dictum, ‘Don’t look for the meaning, look for the use’, Grice’s emphasis on both meaning and use undermines certain influential claims made during the hey-day of Oxford's ordinary language philosophy, and has affected debates on a wide range of philosophical issues within and beyond philosophy of language.

Grice's legacy is encapsulated in such widely used phrases as ‘Griceian intention’, ‘Griceian maxims’, ‘Griceian pragmatics’, and ‘Griceian reasoning’.

To get an intuitive feel for Grice’s basic ideas and the philosophical motivations behind them, we should begin with some simple examples.

Compare what an utterer is likely to mean in uttering one or the other of these sentences:

(1) Lizzie felt lousy and ate some chicken soup.
(2) Lizzie ate some chicken soup and felt lousy.

A speaker is likely to use (1) to mean that Lizzie felt lousy and, then, in order to feel better, ate some chicken soup, but use (2) to mean that Lizzie ate some chicken soup and, because of something in the soup, later felt lousy.

How should this difference in meaning be explained?

The only relevant difference between the two sentences is the order of the two verb phrases.

That might be enough to explain the difference in meaning if "and" meant something like "and then, as a result," but it does not mean that.

In ‘Toledo is in Ohio and Memphis is in Tennessee’, the question of order does not even arise.

Grice held that the meaning of "and" is adequately captured by the truth-functional connective of formal logic, notwithstanding the fine-grained observations on everyday use made by the ordinary language philosophers of Oxford, notably J. L. Austin, who led the Play Group.

The Play Group's ‘linguistic botanising’ is a reaction to what they regarded as the excessive formalism characteristic of such philosophers as Frege, the early Wittgenstein, and the logical positivists, who abstracted from red-blooded speech to pursue their interest in language suitable for science.

Austin and his Play Group (or kindergarten) evidently accepted the later Wittgenstein’s view that, generally speaking, ‘the meaning of a word is its use in the language.'

Grice appreciates the importance of distinguishing them, although he granted that meaning is grounded in use, as evident from his endorsement of the analytic-synthetic distinction (Grice and Strawson, "In defence of a dogma.").

Ordinary language philosophers like Austin noted the intuitive differences between pairs of potential utterances like (1) and (2), and pointed out that systematic semantics seemed unable to account for the differences.

If ‘and’ is just logical conjunction, ‘A and B’ should be equivalent in meaning to ‘B and A’; but this doesn’t seem to be the case with pairs like (1) and (2).

By identifying more and more cases like this, a sceptical view of the applicability of formal methods to natural language was supported.

Grice argued that the differences found in cases like the one we’re considering could be explained by theories which held on to the idea that words make constant (and relatively easily capturable) contributions to determining the meanings of the sentences in which they appear, in large part by distinguishing between sentence meaning (the meaning determined by constant word meanings) and the utterer's or occasion meaning (the meaning got across by the utterer using the sentence he did in the particular conversational context in which he spoke).

Let’s look at how this works in the current case in more detail.

In the case of (1) and (2), however, the difference in linguistic meaning is minimal but the difference in use is considerable.

It is the fact that the utterer utters (1) rather than (2) or, as the case may be, (2) rather than (1), presumably with a certain intention, that makes the difference.

This difference derives from the order in which the speaker utters the two clauses and the presumption that this reflects the order of events described (assuming there is an order).

Also, the speaker exploits the fact, and the hearer’s knowledge of the fact, that it is folk wisdom that chicken soup can make one feel better or, alternatively, that any food can be contaminated, to convey something that goes well beyond the linguistic meaning of the sentence itself.

Finally, notice that in uttering (1) or (2) the speaker could have canceled the implication about the order of events (and about what caused what) by adding, ‘but not in that order’.

So this implication cannot be a matter of linguistic meaning.

Here is another kind of case, illustrating Grice’s line of resistance to another ordinary language challenge to formal philosophy.

Reacting to traditional philosophical worries, about such questions as what it is for an action to be voluntary and what it is for things in the world to be as they appear, Austin and his followers thought there was something wrong with the following (a)-sentences, in contrast to the very similar (b)-sentences:

(3) a. Sam scratched his head voluntarily.
      b. Sam handed over his car keys voluntarily.

(4) a. There appears to be a computer screen before me.
      b. There appears to be a dagger before me.

According to Austin’s dictum, ‘No modification without aberration,' the trouble with the (a)-sentences is that they are normally inappropriate to utter, since with (3a) there is no reason to think that Sam scratched his head involuntarily and with (4a) there is no reason (for me) to question whether there is a computer screen before me.

However, surely this does not show that it is not true that Sam scratched his head voluntarily or that there appears to be a computer screen before me.

Rather, it shows that there is normally no point in making such statements. These examples illustrate how applying the distinction between linguistic meaning and language use can do justice both to traditional philosophical concerns and to niceties about everyday usage.

Here are two more examples. One might argue that believing implies not knowing because to utter, say, ‘I believe that bungee jumping is dangerous’ implies that one does not know this, or argue that trying implies effort or difficulty because one would not say ‘Harry tried to stand up’ if Harry stood up with ease.

However, as Grice would reply, what carries such an IMPLICATUM is not what one is saying but that one is saying it – rather than the stronger 'I know that bungee jumping is dangerous’ or ‘Harry stood up’. Grice similarly objected to certain ambiguity claims, such as that the word ‘or’ has both an inclusive and an exclusive sense, by pointing out that it is not what ‘or’ means but its use that can carry the implication of exclusivity.

Grice’s Modified Occam's Razor, ‘Senses are not to be multiplied beyond necessity’ (1978/1969, p. 47), cut back on the growing philosophical conflation of language use with linguistic meaning. It has since helped linguists appreciate the importance of separating the domains of semantics and pragmatics.

In ‘Meaning’ Grice contrasts ‘natural’ meaning with the sort of meaning involved in language and communication. Whereas smoke means fire by virtue of being naturally correlated with fire, the word ‘smoke’ means smoke by virtue of being conventionally correlated with smoke. And, whereas smoke is correlated with fire in the sense of indicating the presence of fire, obviously the word ‘smoke’ is not correlated with the presence of smoke. It is a conventional means for talking about smoke, whether or not smoke is present.

Its meaning is a matter of convention, since it could just as well have meant something else, and some other word could just as well have meant smoke. It means smoke because, and only because, speakers normally use it to mean that and expect others who use it to mean that as well. Suppose a speaker utters the sentence, ‘I smell smoke’, using the pronoun ‘I’ to refer to herself, the verb ‘smell’ (in the present tense) for olfactory sensing, and the noun ‘smoke’ to refer to smoke, presumably some smoke in her vicinity. Presumably she means that she then smells smoke. Importantly, this is not a direct consequence of what her words mean, and of how their meanings are put together into the meaning of the sentence, given its grammatical structure. After all, she could have been speaking figuratively, and have meant, say in response to an explanation her husband just gave her, that she suspects that he is trying to divert her attention.

So it is one thing for words to mean something and another for a speaker to mean something, even if it is the same thing, in uttering those words.

Grice held that speaker meaning is more basic than word meaning, on the grounds that the meanings of words ultimately come down to what people mean in uttering them (Grice, 6 1968). But what is it for a speaker to mean something, whether in uttering words or in doing something else? It is not merely to cause some effect on one’s audience. There are lots of ways of doing that. At the very least it must be intentional. But there are different ways in which one can intend to produce some effect on others, and most of them do not involve communication.

In communicating something, one intends one’s audience to identify what effect one is trying to produce in them. Indeed, as Grice (1957) argued, one intends to produce that effect precisely by way of their recognizing one’s intention to produce it. This is the gist of Grice’s ingenious idea that the special sort of intention involved in meaning something, in trying to communicate something, is in a certain sense reflexive. Think about what is involved in communicating. You have a certain thought and you wish to ‘get it across’ to someone. So your intention to convey it must be overt. Your intention will not be communicative if you intend the hearer to think a certain thing without thinking you intend them to think it. Suppose, for example, you want the person to think you are modest, you make some self-deprecating remarks. In this case your intention will not be fulfilled if it is recognized. The person will not think that you are modest but merely that you want them to think that. And even if your intention is overt, that doesn’t by itself make it communicative. For example, by turning up the heater you could intend to get your guest to think that it is time to leave, but you would not have communicated that. If the person comes to think that it is time to leave, their awareness of this need not depend on their recognition of your intention for them to think that. If the room is hot enough, your intention will be fulfilled independently of their recognizing it. Such examples suggested to Grice that for an intention to be communicative, it must be overt in a specific sort of way. His idea was that communicative intentions are intentionally 7 overt and that this feature plays a special role in their fulfillment. That is, in trying to communicate something to others by saying something, a speaker intends the audience to recognize that intention partly by taking into account that they are so intended.

Because this is part of what the speaker intends, communicative intentions are distinctively selfreferential or ‘reflexive’. A speaker means something by his utterance only if he intends his utterance ‘to produce some effect in an audience by means of the recognition of this intention’ (Grice, 1957/1989, p. 220). However, not just any sort of effect will do.

If you are communicating something to someone, communicative success does not require that respond as you wish, such as to believe you, obey you, or forgive you.

As Searle pointed out, these are perlocutionary effects (Austin, 1962, 101ff.), the production of which goes beyond communication (1969, p. 47). It is enough, as Strawson had argued, for the utterance to be understood (1964, p. 459). For that the hearer must identify the attitude (including its content) the speaker is expressing – believing, intending, regretting, etc. – whether or not the speaker actually possesses that attitude. This suggests that for a speaker to mean something is to intend just such an effect. If, as Bach and Harnish proposed, ‘to express an attitude is reflexively to intend the hearer to take one's utterance as reason to think one has that attitude’ (1979, p. 15), communicating successfully, being understood, consists in having the expressed attitude recognized. It does not require the hearer to respond in any further way. For a communicative intention, ‘its fulfillment consists in its recognition’ (ibid.). This is the flip side of Grice’s idea that communicative intentions are inherently reflexive.

Commenting on the notion of an intention ‘to produce an effect in an audience by means of the recognition of this intention’, Grice remarked, ‘this seems to involve a reflexive paradox, but it does not really do so’ (1957/1989, p. 219). It seems to because the intention is self-referential. Indeed, this air of paradox may seem to imbue the hearer’s inference with circularity, inasmuch as the hearer is to identify the speaker’s intention partly on the supposition that he is intended to. Is there anything paradoxical about this? It might seem paradoxical if one confuses iterative intentions with reflexive ones, as indeed Grice himself seems to have done. Indeed, earlier in the very paragraph just quoted from, Grice gives an alternate formulation, requiring that a speaker ‘must also intend his utterance to be recognized as so intended’ (1957/1989, p. 219). And, in his later attempt to improve upon his earlier formulation, Grice (1969) abandons reflexive intentions in favor of iterative intentions. Still later, in ‘Meaning revisited’, he rejects that idea and suggests that what is needed instead is the absence of ‘sneaky intentions’ (1982/1989, p. 302). However, Gilbert Harman thinks that Grice should have stuck with his original idea: Grice himself originally states his analysis as involving a self-referential intention … but, because of worries about what he calls ‘self-reflective paradox’, he goes on to restate the analysis as involving a series of intentions, each about the preceding one. This turns out to lead to tremendous complexity in the resulting theory. Much of this complexity is artificial and due to Grice’s refusal to stick with the original analysis and its appeal to a self-referential intention. (Harman, 1986, pp. 87-88) A reflexive intention is not a series of intentions, each referring to the previous one. A speaker does not have an intention to convey something and a further intention that the first 9 be recognized, which itself must be accompanied by an intention that it in turn be recognized, and so on ad infinitum. Grice’s move to iterative intentions led to increasingly complex formulations beginning with Strawson’s (1964), followed by Grice’s (1969) own, and culminating with Schiffer’s (1972), each prompted by counterexamples to the previous one. However, as Harman pointed out, sticking with self-referential intentions avoids this complexity and the threat of an infinite regress.

Bach (1987) argued similarly, responding also to other concerns, due to Recanati (1986), about reflexive intentions. The semblance of reflexive paradox arises from Grice’s key phrase ‘by means of the recognition of this intention’. This might suggest that to understand the speaker the hearer must engage in some sort of circular reasoning. It sounds as though the hearer must already know what the speaker’s communicative intention is in order to recognize it. However, that misconstrues what the hearer has to take into account in order to recognize the speaker’s intention. The hearer does not infer that the speaker means a certain thing from the premise that the speaker intends to convey that very thing. Rather, operating on the presumption that the speaker, like any speaker, intends to communicate something or other, the hearer takes this general fact, not the content of the specific intention, into account in order to identify that intention. 3. Conversational Implicature Grice is best known, especially in linguistics, for his theory of conversational implicature. The basic idea was not new. What was new, aside from Grice’s name for it, was his account of it, which incorporated his account of speaker meaning, along with his application of it to various philosophical questions. A speaker implicates something if what she means is distinct from what she says. In general, though not always, it is also distinct from what is 10 implied by what she says. That is why Grice (1967) uses the verb ‘implicate’ rather than ‘imply’ and the neologism ‘implicature’ rather than ‘implication’. For example, suppose you are asked about a dinner you had at an expensive restaurant, and you reply, ‘The meal didn’t give me food poisoning’. Saying this implicates that it was mediocre at best. However, what you said obviously does not imply this.

After all, even excellent meals need not cause food poisoning. Besides, as Grice points out, conversational implicatures are cancelable. You could have added, ‘I don’t mean to suggest that the meal wasn’t great’. In fact, there are circumstances in which the implicature would not arise in the first place, say if there had been a recent outbreak of E. coli. According to Grice, most implicatures are also nondetachable, in the sense that the speaker could have said what she said in a different way and still implicated the same thing. The exception, as discussed below, is the case where how the speaker says what she says, the wording or the pronunciation, plays an essential role. How can a speaker implicate something that is not implied by what she says? She can do this by exploiting the fact that the hearer presumes her to be cooperative, in particular, to be speaking truthfully, informatively, relevantly, and otherwise appropriately (from now on we will use ‘she’ for the speaker and ‘he’ for the hearer). If taking the utterance at face value is incompatible with this presumption, the hearer, still relying on this presumption, must find some plausible candidate for what else the speaker could have meant. In the above example, where the speaker is asked to evaluate a certain dinner, he must figure out what she meant, relying on the presumption that she intended it to be an accurate, informative, and appropriate answer to the question. In effect what the hearer does is, on the presumption that the speaker is being cooperative, to find a plausible explanation for why she said what she said. 11 3.1 The Cooperative Principle and the Maxims of Conversation Grice codified these ideas by formulating an overarching Cooperative Principle and four sets of subordinate maxims of conversation (Grice, 1975/1989, pp. 26-27):

COOPERATIVE PRINCIPLE:

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.

QUALITY: Try to make your contribution one that is true.
1. Do not say what you believe to be false.
 2. Do not say that for which you lack evidence.

QUANTITY:
1. Make your contribution as informative as is required (for the current purposes of the exchange).
2. Do not make your contribution more informative than is required.

RELATION: Be relevant.

MODUS: Be perspicuous [sic]
1. Avoid obscurity of expression.
2. Avoid ambiguity.
3. Be brief. (Avoid unnecessary prolixity [sic])
4. Be orderly.

We could dwell on the precise meanings of these maxims, on whether they are adequately formulated, and on whether the list can be simplified (for discussion see Harnish, 1976). One might wonder, for example, about what happens when applying different maxims gives 12 different candidates for what a speaker might be implicating.

A common objection to Grice’s account is that it is not adequately predictive and, indeed, that different social situations or even cultural differences make for different norms. All these issues have been raised, as if an adequate account must explain precisely how a hearer figures out what the speaker is implicating and, for that matter, how a speaker, in choosing what to say, manages to anticipate how the hearer is going to do this. Surely, though, this is expecting too much of a philosophical theory. These difficult issues belong to psychology (of the future). Philosophically, the important point is that, whatever the details, speakers commonly succeed in conveying things that are distinct from or go beyond what they say and that, unless communication is a kind of telepathy, there must be rational constraints on speakers’ communicative intentions and on hearers’ inferences about them. As the examples below will illustrate, Grice’s maxims point to the sorts of considerations that speakers and hearers take into account if communication is to succeed, as surely it often does. With this mind Bach and Harnish suggest that the maxims are better viewed as presumptions, which the hearer relies on to guide his inference to what the speaker means (1979, pp. 62-65). So, when a presumption seems not to be in force, he seeks an interpretation of the speaker’s utterance such that it does apply after all.

Indeed, Bach and Harnish propose to replace Grice’s CP, the vague and rather unrealistic Cooperative Principle, with a Communicative Presumption: when people speak, presumably they do so with identifiable communicative intentions (1979, pp. 12-15). The following examples illustrate how hearers accommodate (to borrow a term from Lewis, 1979) apparent violations of different maxims (or, if you will, apparent suspensions of different presumptions). If a speaker says something that is obviously false, thereby flouting the first maxim of quality, she could well mean something else. For example, with 13 (5) she might mean the opposite of what she says, with (6) something less extreme, and with (7) something more down to earth. (5) The Aristocrats is the most wholesome movie I’ve ever seen. (6) I had so much fun I could have played D&D for a month without ever sleeping. (7) He bungee-jumped from 85% approval down to 40%, up to 60%, and down to 15%. In these cases, of irony, hyperbole, and metaphor, respectively, it should be evident what a speaker is likely to mean. It is not what she says or, as Grice puts it, ‘makes as if to say’. Notice, however, that not all quality implicatures depend on the obvious falsity of the literal content. John Donne’s famous ‘No man is an island’ is a case in point.

Quantity implicatures are typified by the fact that speakers convey information not just by what they say but also by the stronger things they don't say. Consider these examples: (8) Barry tried hard to lift the 300-lb barbell. (9) He thought he was strong enough to lift it. (10) He had lifted the 250-lb barbell three times. (11) Barry finished his workout with a swim or a run. Keeping in mind that speakers, not sentences, implicate things, we have to imagine uttering such sentences or hearing them uttered in particular contexts in order to get clear examples of implicatures, quantity implicatures in this case. In uttering (8) you might implicate that Barry failed to lift the 300-lb barbell. Otherwise, you would have said that he succeeded. Similarly, with (9) you might implicate that he wasn’t sure that he could lift it. With (10) you would probably implicate that he didn’t lift the 250-lb barbell more than three times. And, finally, in uttering (11) you would implicate that Barry went for either a swim or a run and not both and that you do not know which. In the case of (8) and (9), what the speaker implicates can be figured out on the presumption that if she was in a position to give 14 stronger or more specific information, she would have. With (10), the presumption is that the speaker is in a position to know how many times Barry lifted the 250-lb barbell, whereas with (11) the presumption is just the opposite, since if she knew whether Barry went for a swim or a run (or both), she would have said so.

Relevance implicatures can also be cases of conveying information by saying one thing and leaving something else out. Grice’s two best-known examples are of this type: (12) There is a garage around the corner. [said in response to ‘I am out of petrol’] (13) He is punctual, and his handwriting is excellent. [the entire body of a letter of recommendation] An utterance of (12) is relevant, and a rational speaker would intend it as such, only if the speaker means also that the garage is open and has petrol for sale. So the hearer is to reason accordingly. (13) is rather different, on account of the speaker’s reason for not being more explicit. In this case, the writer intends the reader to figure out that if she had anything more positive to say about the candidate, she would have said it. Manner implicatures are probably the least common of the lot. They exploit not just the speaker’s saying a certain thing but her saying it in a certain way. For that reason, they are exceptions to Grice’s nondetachability test. Obviously, if there are different ways of saying the same thing and how the speaker says it affects what the hearer is likely to take into account in figuring out what the speaker means, the implicature is not detachable.

 The following examples illustrate this. (14) You have prepared what closely resembles a meal of outstanding quality. (15) I would like to see more of you. 15 Imagine a culinary instructor uttering the longwinded (14). Her intention would likely be to convey that the meal is not nearly as good as it appears. A speaker of (15) could exploit its ambiguity to convey something besides wanting to spend more time with the hearer. It should be understood that Grice does not suppose that speakers consciously exploit the maxims or that hearers consciously take them into account. However, this raises the interesting question of just what is involved psychologically in the process of communication. This includes not only how hearers manage to figure out what speakers mean given that they say what they say, but also how speakers choose what to say and thereby manage to make evident what they mean, even when they do not make it explicit. Grice did not address the latter question, and his account of implicature is commonly misconstrued as an answer to the former. 3.2 Common Misunderstandings Numerous common misconceptions about conversational implicature have arisen, based on misreading ‘Logic and Conversation’ or on reading things into it. Space does not permit highlighting more than a few (for more see Bach, 2006) or documenting instances of them. The following two are perhaps the most important. First of all, the maxims (or presumptions) do not determine implicatures but, rather, faciliate their communication. They are considerations that speakers implicitly intend hearers to, and hearers do, take into account to figure out – determine in the inferential sense – what the speaker is implicating. Since that is a matter of speaker meaning, it is the speaker’s communicative intention that determines, in the constitutive sense of that word, what is implicated. Moreover, because the maxims figure in Grice’s account of conversational implicature, that is, of how implicatures get conveyed, it is commonly but 16 mistakenly supposed that the maxims apply only to implicatures. In fact, they apply equally to completely literal utterances, where the speaker means just what she says. After all, the hearer still has to infer this. It is thus wrong to suppose that the maxims come into play only where linguistic meaning leaves off and speaker meaning and extralinguistic, contextual information take over (for more on context and what it does and doesn’t determine, see Bach, 2005). Another misconception is that linguistic expressions (sentences or even individual words) can implicate things. Speakers do. To be sure, certain expressions are characteristically used (by speakers) to implicate things. When this occurs we have what Grice calls generalized conversational implicatures (as opposed to particularized ones). These have been investigated in great depth by Stephen Levinson (2000), who thinks they give rise to an intermediate level of meaning. In fact, they give rise to an intermediate kind of inference, but Levinson confusedly thinks of GCIs as inferences. A related misunderstanding leads to the objection that Grice’s account misrepresents the inference process involved in recognizing implicatures and that it cannot account for the putative phenomenon of pragmatic intrusion, as exemplified by embedded implicatures.

Here is a typical formulation of that objection: Grice’s account makes implicature dependent on a prior determination of ‘the said’.

The said in turn depends on disambiguation, indexical resolution, reference fixing, not to mention ellipsis unpacking and generality narrowing. But each of these processes, which are prerequisites to determining the proposition expressed, may themselves depend crucially on processes that look indistinguishable from implicatures. Thus what is said seems both to 17 determine and to be determined by implicature.

Let us call this Grice’s circle. (Levinson, 2000, p. 186)

This objection is based on confusing the two sorts of determination mentioned above. In the relevant, constitutive sense, what is said neither determines nor is determined by what is implicated. Levinson and many others misconstrue Grice’s account as a psychological model of the hearer’s inference, indeed one according to which the hearer must ascertain what the speaker says before figuring out what the speaker implicating (see Bach, 2001, pp. 24-25, and Saul, 2002). But that is not how Grice intended his account. He required that ‘the presence of a conversational implicature be capable of being worked out’ (1975/1989, p. 31), but he did not say that it must be, much less in what sequence. The confusion just mentioned leads to the widespread misconception, evident from an extensive literature, that some implicatures are ‘embedded’, as with utterances of sentences like these: (16) It is better to get married and get pregnant than to get pregnant and get married. (17) Michelle thinks that Barack has two children. Since the two infinitival clauses of (16) are semantically equivalent, a speaker is likely to implicate that what is better is getting married and then getting pregnant. With (17) the implicature is not that Barack has exactly two children but that Michelle thinks that. In fact, such examples illustrate merely that the process of figuring out what is implicated does not require first ascertaining what is said. They do not show that the implicature is embedded in anything. Indeed, since speakers implicate, it does not even make sense to say that some implicatures are embedded. 18 3.3 Between Saying and Implicating With his contrast between saying and implicating, Grice allowed both for cases in which the speaker means what she says and something else as well and for ones in which the speaker says one thing and means something else instead. He counted both as kinds of implicature, although the latter might better be described as speaking figuratively. And, of course, a speaker can say something and mean just that. Grice seems to have overlooked an intermediate phenomenon, one that has been investigated by many others (Sperber and Wilson, 1986, Bach, 1994, Carston, 2002, and Recanati, 2004).

As they have observed, there are many sentences whose standard uses go beyond their meanings (even with references fixed and ambiguities resolved) but are not implicatures or figurative uses either. One way in which this can occur is when what the speaker means is a more elaborate proposition than what is expressly said, as with a likely utterance of (18): (18) Sidney and Sally are engaged. The speaker is likely to mean that they are engaged to each other, even though she does not make the last part explicit. Clearly that element is cancelable, since she could have added ‘but not to each other’ to her utterance of (18) without contradicting herself. Similarly, someone utter (2), (2) Lizzie ate some chicken soup and felt lousy. would likely mean that Lizzie ate some chicken soup and as a result felt lousy. Again, the inexplicit part is cancelable, for the speaker could have uttered (2) and added, ‘not that she felt lousy as a result’. In both cases it is not the linguistic meaning of the uttered sentences but the fact that the speaker said what she said, presumably with maximal relevant informativeness, as per the first maxim of quantity, that provides the hearer with reason to think that the speaker intended to convey something more expansive. 19 In other cases, what the speaker says is not merely less expansive than what she means but does not explicitly comprise a proposition. Suppose a boy has just finished his ice cream and blurts out (19), (19) I want more. He can’t just mean that he wants more, full stop. Rather, he probably means that he wants more ice cream. And if his mother yells back, (20) You’ve had enough. she means that he has had enough ice cream. In both cases the sentence falls short of fully expressing a proposition – it is semantically incomplete. Yet what the speaker means is a complete proposition.

Sentences like (19) and (20) appear to violate the grammar school dictum that a sentence, unlike a mere word, phrase, or ‘sentence fragment’, always expresses a ‘complete thought’. As with (17) and (18), though for a different reason (semantic incompleteness), what the speaker means is more specific than what the sentence means. We might say that whereas what a user of (17) or (18) means is an expansion of the sentence meaning, what a user of (19) or (20) means is a completion of it. Several of Grice’s critics have pointed out that expansions and completions are not related closely enough to conventional meaning to fall under Grice’s notion of what is said but are too closely related to count as implicatures. Sperber and Wilson (1986, p. 182) coined the word explicature for this in-between category, since part of what is meant ‘explicates’ what is said. Bach (1994) proposed calling these cases of impliciture, since part of what is meant is communicated implicitly, by way of expansion or completion. Recanati (2004) suggests that the notion of what is said should be extended to cover these cases (in fact, he offers a series of progressively more liberal notions of saying), but clearly he is 20 going beyond Grice’s conception of what is said as corresponding to the meanings of the constituents of the sentence with its syntactic structure (Grice, 1969/1989, p. 87). 4. Going Further Grice's accounts of speaker meaning and conversational implicature have had lasting strategic impact on philosophy and linguistics. Despite disagreements about many particular cases, it is widely acknowledged that the Gricean strategy reduces the burden on semantics by seeking to diverse meaning phenomena with the help of general principles of rational language use rather than needlessly complex word meanings. It has been applied to diverse problems not only in philosophy of language, such as propositional attitude reports and referential uses of definite descriptions, but also in other areas of philosophy, notably epistemology and meta-ethics. It has been employed in linguistics and computer science to address such topics as the lexicon, anaphora, speech planning, and discourse analysis. The notion of speaker meaning, broadly construed, is relevant to literary theory and aesthetics, in connection with questions about author or artist intention. The distinction between saying and implicating bears on legal theory, in particular in helping to explain the difference between lying and misleading, a difference clearly central to the law of perjury, libel, and fraud.

We have focused on Grice’s most important and influential ideas, speaker meaning and conversational implicature. Stephen Neale (1992) presents a much fuller discussion of these and related ideas, Siobhan Chapman (2005) provides a full-length intellectual biography, which covers Grice’s views beyond philosophy of language, Larry Horn (2009) offers a 40- year retrospective on implicature in particular, and Bart Geurts (2010) gives an in-depth study of quantity implicatures. A fuller treatment here would go into more detail on Grice’s 21 attempts to improve upon his analysis of speaker meaning (Grice, 1969, 1982), to reduce linguistic meaning to speaker meaning (1968), to justify complicating his account with the controversial hybrid notion of conventional implicature (for the controversy see Bach, 1999), and to give a pragmatic treatment of the notion of presupposition (Grice, 1981). And, as most recently illustrated by the papers collected in Petrus 2010, it would document the extent of the influence of his ideas in philosophy and beyond.

References

Austin, J. L. (1961), ‘A plea for excuses’, in J. L. Austin, Philosophical Papers, J. O. Urmson and G. J. Warnock (eds). Oxford: Oxford University Press, pp. 123-52. _____ (1962), How to do Things with Words. Oxford: Oxford University Press. Bach, K. (1987), ‘On communicative intentions: A reply to Recanati’. Mind & Language, 2, 141- 154. _____ (1994), ‘Conversational impliciture’. Mind & Language, 9, 124-62. _____ (1999), ‘The myth of conventional implicature’. Linguistics and Philosophy 22: 327-66. _____ (2001), ‘You don’t say?’ Synthese, 128, 15-44. _____ (2005), ‘Context ex machina’. in Z. Szabó (ed), Semantics versus Pragmatics. Oxford: Oxford University Press, pp. 15-44. _____ (2006), ‘The top 10 misconceptions about implicature’, in B. Birner and G. Ward (eds), Drawing the Boundaries of Meaning. Amsterdam: John Benjamins, pp. 21-30. _____ and Harnish, R. M. (1979), Linguistic Communication and Speech Acts. Cambridge, MA: MIT Press. Carston, R. (2002), Thoughts and Utterances. Oxford: Blackwell. Chapman, S. (2005), Paul Grice: Philosopher and Linguist. Basingstoke, UK: Palgrave Macmillan. Geurts. B. (2010), Quantity Implicatures. Cambridge, UK: Cambridge University Press.

Grice, H. P. (1957), ‘Meaning’. The Philosophical Review, 66, 377-88 (reprinted in Grice, 1989). 22 _____ (1968), ‘Utterer’s meaning, sentence-meaning, and word meaning’. Foundations of Language, 4, 225-242 (reprinted in Grice, 1989).
 _____ (1969), ‘Utterer’s meaning and intentions’. The Philosophical Review, 78, 147-177 (reprinted in Grice, 1989). _____ (1975), ‘Logic and conversation’. in P. Cole and J. Morgan (eds), Syntax and Semantics, Vol. 3, Speech Acts, New York: Academic Press, pp. 41-58 (reprinted in Grice, 1989). _____ (1978), ‘Further notes on logic and conversation’, in P. Cole (ed), Syntax and Semantics, Vol. 9, Pragmatics, New York: Academic Press, pp. 113-127 (reprinted in Grice, 1989). '
_____ (1981), ‘Presupposition and conversational implicature’, in P. Cole (ed), Radical Pragmatics. New York: Academic Press, pp. 183-97 (reprinted abridged in Grice, 1989).
_____ (1982), ‘Meaning revisited’, in N. Smith (ed), Mutual Knowledge. London: Academic Press, pp. 223-43 (reprinted in Grice, 1989).
 _____ (1989), Studies in the Way of Words. Cambridge, MA, Harvard University Press.
_____ and Strawson, P. F. (1956), ‘In defense of a dogma’, Philosophical Review, 65, 141-58
(reprinted in Grice, 1989).
______(n.d.). The H. P. Grice Papers, BANC MSS 90/135c, The University of California, Berkeley.

Harman, G. (1986), Change in View. Cambridge, MA: MIT Press. Harnish, R. M., (1976), ‘Logical form and implicature’, in T. Bever, J. Katz, and T. Langendoen (eds), An Integrated Theory of Linguistic Ability. New York: Crowell, pp. 313-92. Horn, L. (2009), ‘WJ-40: Implicature, truth, and meaning’. International Review of Pragmatics, 1, 3-34. Levinson, S. (2000), Presumptive Meanings, The Theory of Generalized Conversational Implicature. Cambridge, MA: MIT Press. Lewis, D. (1979), ‘Scorekeeping in a language game’. Journal of Philosophical Logic, 8, 339-59. Neale, S. (1992), 'Paul Grice and the philosophy of language’. Linguistics and Philosophy, 15, 509- 59. Petrus, K. ed. (2010), Meaning and Analysis: New Essays on Grice. Basingstoke, UK: Palgrave Macmillan. 23 Recanati, F. (1986), ‘On defining communicative intentions’. Mind & Language, 1, 213-242. _____ (2004), Literal Meaning, Cambridge, UK, Cambridge University Press. Saul, J. (2002), ‘What is said and psychological reality, Grice’s project and relevance theorists’ criticisms’. Linguistics and Philosophy, 25, 347-72. Schiffer, S. (1972), Meaning. Oxford: Oxford University Press. Searle, J. (1969), Speech Acts, An Essay in the Philosophy of Language. Cambridge, UK: Cambridge University Press. Sperber, D. and Wilson, D. (1986), Relevance. Cambridge, MA: Harvard University Press. Strawson, P. F. (1964), ‘Intention and convention in speech acts’. Philosophical Review, 73, 439-60. Wittgenstein, L. (1953), Philosophical Investigations. G.E.M. Anscombe and R. Rhees (eds), G.E.M. Anscombe (trans). New York: Macmillan.