Can Meaning Be Described In Terms Of Speakers' Intentions?
Many accounts of meaning begin by asking what it is that sentences represent, and most frequently the response is that the meaning of a sentence is defined by what would have to be the case for that sentence to be true. There is, however, another way to approach the question of meaning, and that is the stance taken by Locke and Grice amongst others - to approach meaning from the opposite direction, viewing language as a means of communication, and asking how sentences are formed in view of this. Such accounts tend to focus on the intentions of the speaker (or writer, signaller, etc.) in saying (or writing, signalling, etc.) a particular sentence, suggesting that the sentence gains its meaning because the person that uses it intends it to produce some belief or cause some action in the audience. In this essay I shall ask whether this is how language is used and whether an analysis of the meaning of sentences in terms of what the speaker's intentions are is useful.
The first task Grice sets himself is to distinguish things with that which he calls 'natural meaning' from those with 'non-natural meaning' (or meaningNN). He seems to be suggesting that some things have a meaning in virtue of what they directly imply: so 'Spots mean measles' because if someone has spots (in a rather simplistic medical fantasy) this can only mean that they have measles; 'Being out in the rain without an umbrella means you'll get wet' because it seems unavoidable - and this usage of 'means' he terms 'natural'. What Grice is discussing when talking about 'natural meaning' seems simply to be the sort of sentences where we express a causal idea (or at least a constant conjunction) using the verb 'to mean', so that when observing the signifier someone can't help but realise the signified. With ' meaningNN', on the other hand, Grice is talking about sentences where no direct implication is involved, but where someone uses particular words to represent some idea, and it is only by some form of convention or understanding of language that the listener comes to understand what the sentence 'means'. It is the concept of meaningNN that Grice was most interested in, and it is this usage of the word 'means' that he tried to clarify.
Grice suggested that if speaker A means p by uttering sentence s then A wants some audience to believe that p (or perhaps to act in act in a way in line with a belief that p); meaning is imbued upon sentences in order for those sentences to communicate something from one person to another. Ignoring the possibility that the person is talking to himself, we should therefore construe a sentence as an attempt by the speaker to create such a belief in the person he is talking to - but if the audience is to gain such a belief, Grice argues, he must realise that the speaker has just this intention. This is how it comes to be the case that if I say 'As I write this essay I am banging my head against the wall violently' a reader may well not gain the belief that I am in fact banging my head against the wall, thinking rather that I have just used the same words as would express such a suggestion merely as an example sentence. In order for a reader to gain a particular belief due to my writing a particular sentence he must recognise that I am trying to create this belief in his mind; he must believe that I am not trying to mislead, be sarcastic, or use the sentence in a way (such as as an example) which is not actually intended to be read as expressing something I am actually thinking.
Furthermore, Grice argues, A must want the recognition of his intentions in using sentence s to be the reason why his audience comes to believe that p - so A must believe that his saying sentence s will have the effect of producing the belief that p in his audience because they see what A is trying to do, and that the audience will not simply come to believe that p by some other means, since then his attempt to affect their beliefs using s would be entirely unnecessary. Grice seems to see the process of communication not just as one person broadcasting his beliefs to everyone else, but as other people also agreeing to receive the broadcast. This seems obviously true in a fairly vacuous form, that people can always ignore someone speaking at them (and can sing 'Mary had a little lamb' to block out the noise), and must allow someone to speak to them if any information is to be passed on, but is also very plausible in terms of the more complicated idea that communication does not occur unless the listener engages in some way with the speaker.
Grice's account seems very reasonable in terms of how language might come about. Few people would argue with the claim that language emerged or evolved in order to enable people to share ideas amongst one another, if only ideas such as where to find food, or how to protect oneself. Consider the example of the caveman Bob, who is experienced at hunting attempting to help newly come-of-age caveman Tim learn to hunt. No amount of gesticulating on the part of Bob will help Tim to find food if he has failed to realise that Bob is attempting to communicate with him, so in some sense he must acknowledge Bob's intention that information be exchanged. If, however, Tim realises that Bob is not waving his arms about randomly, but attempting to signify something, then he will have achieved the first step towards communication, and will attempt to understand what Bob means by his gesticulations, and together they will attempt to form a rudimentary language.
Grice argues that anyone is able to introduce a particular set of sounds (or gestures, symbols etc.) into conversation, and that through their intention that a belief be communicated these sounds gain a meaning. Furthermore he claims that continued usage of a given example is all that is needed in order to make it fully a part of a language - sentences are used repeatedly and then are adopted into language once they are understood and people are able to use them without confusion. How is it, then, that some sentences are accepted into common usage and understood whereas others are rejected? Whilst it might initially seem likely that we will be unable to account for sounds or motions becoming associated with objects, it is certainly at least plausible that we will be able to invoke some sense of 'salience' in order to choose what might be used to represent simple concepts. A fist, therefore, may come to represent fighting, a 'chomping' sound might represent eating, and hugging oneself could represent being cold. We also seem to have an intuitive grasp of what certain ways of behaving might mean, and it might be a reasonable claim to suggest that some things, such as whether facial expressions are positive or negative, are in some sense 'hard-wired' into our understanding. Whilst the extension of this theory to more complex ideas might seem a great deal more complicated, there does not seem to be any serious contender to this theory, since to deny that people are able to imbue meaning upon words would seem to require a dictionary that is not formulated by the people that use it, but comes from somewhere external - and as Blackburn points out, the idea that our dictionary is given to us (perhaps by God?) in the form we now use it seems very implausible indeed.
Platts, however, argues against Grice's conception on the basis that, since it seems likely that the only indication of someone's intentions is going to be what they actually say, we can't judge their intentions before we know the meaning of the sentences they utter. This seems a little bit strange as an objection, since it seems to be simply a denial of what Grice is saying without any attempt to explain what is wrong with it - Grice is suggesting that from intentions comes language, and Platts seems to respond by saying almost just 'Well, I think language comes before we can judge intentions'. Platts is right to see that Grice's position rests upon the idea that there are other ways to judge (roughly) someone's intentions than simply to interpret what they are saying, and yet if he is to deny this he is going to have a very difficult time explaining how it is that the whole communication process can begin. If Tim has to know the meaning of Bob's gesticulations before he is able to gauge Bob's motives in attempting to communicate with him, then it seems that Bob's attempt to advise this stranger are doomed. According to Grice, Tim is able to recognise Bob's intention to begin the communication process, and developing from this recognition he can (imagining for a moment that this process would be far more simplistic than it almost certainly would be) establish Bob's altruism, and understand that Bob is intending to help him, and, by consideration of the sort of ways that Bob is moving he is able to establish that the communication is one about food, and putting everything together, about how to get food. Whilst Platts may be rightly sceptical about such a reductionist explanation, I cannot see how he would account for the (very real, in my opinion) possibility that Bob would be able to begin a communicative process with Tim.
Furthermore, Grice can offer in his defence a large number of sentences that can only be understood in the context of the speaker's intentions. If Platts is to claim that what a sentence 'means' can be entirely understood without reference to the way in which it is used and the intention behind its usage then it would seem to be the case (in my opinion) that language would have to be very strange indeed, with some terms which most would claim to be used in one way having to mean a very large number of seemingly unrelated or opposing things, because of what Grice called 'Conversational implicature'. Seemingly the most popular example of Gricean conversational implicature comes in the form of 'damning with faint praise'. Consider the case of someone responding to the question 'Is your cleaner very efficient?' with the reply 'She is indeed very efficient at reading magazines'; it seems to be the case that the reply comes in the form of a compliment, and we seem to view efficiency as a positive attribute, and yet what is meant by this appears to be far from positive, and actually suggests that he isn't a very good cleaner. According to Grice such sentences gain their additional content due to the fact that the speaker intends the listener to gain from what he says the belief that, in fact, the cleaner is no good, and this seems to be an accurate analysis.
It is not clear how Platts can respond to this sort of case. He can simply deny that it does mean that the cleaner is bad at cleaning, and can argue it means just what it says, that he is good at reading magazines (whatever would qualify someone as that), and that the response is completely agnostic as to his cleaning ability, yet this seems to be somewhat implausible, especially considering that the only people that would take it this way would be people who failed to recognise the speaker's intention to cast aspersions on the cleaner's cleaning ability - failing just the criterion Grice suggests is essential. Another (popular) response is to argue that whilst the context is important it is not important because of the speaker's intentions. The argument is that since the question was about cleaning ability but the response was about his magazine-reading ability is perfectly compatible with the suggestion that he is a bad cleaner, we should take it to mean that he is. I am not convinced by this argument since the response was equally compatible with the suggestion that he is a good cleaner - we can only decide between them by knowing about how people conventionally reply to questions when they are trying to be positive, and the sorts of responses they may give if they intend to be negative, but want to appear positive. This seems to lead into the third response, which is that the reply did include a comment about his cleaning ability, and that was that it is bad. This seems the least attractive response of all, since then if a child receives a school report which says 'Billy has become very good at reading magazines' this would presumably also mean that Billy was bad at cleaning.
Clearly there are going to be a very large number (in fact, of course, infinite) of sentences where the meaning is not expressed entirely by the most usual usage of the words, but where an understanding of how the speaker is trying to use them is necessary. Grice, however, only needs a single example to show that an understanding of intentions may be vital to the meaning of some sentences, and thus cannot always simply be ignored. Sarcasm, again, is a good example of this, unless we are to construe 'It really is very cold in here' as actually meaning 'It is very hot' on random occasions (rather than on the occasions when people use it to mean this!).
Overall, however, I may have to step back from the most extreme way that one might take Grice's theory of meaning. I would not agree that every time we hear a sentence we have to consciously take a number of steps to decide whether we believe the intentions of the speaker to be pure (rather than deceitful or simply flippant), nor would I agree that people never say things which would evoke in their audience a belief that they are perfectly aware the audience already have, or at least will definitely come to believe - just consider small-talk. I would, however, agree with Grice's theory in a slightly more relaxed form, since it does seem to be the case that meaning arises in sentences due to a decision on the part of the speaker to express an idea and convey it to someone else, and that the listener must recognise that such a transaction is being made in order for the vibrations hitting their eardrums to have any effect.