Speech Acts

Primary:              Austin, J. L. (1961) ‘Performative Utterances’ (in Philosophical Papers, Oxford: OUP.)

                                                (1962) How to Do Things with Words, Oxford: Clarendon Press

                               Searle, J. (1965) ‘What is a Speech Act’ (in M. Black (ed.) Philosophy in America, Ithaca, NY: Cornell University Press.)

                                                (1969) Speech Acts, London: Cambridge University Press

Secondary:           Lycan, W. G. (2000) Philosophy of Language, London; Routledge.

                                Miller, A. (1998) Philosophy of Language,

                                Harrison, B. (1979) An Introduction to the Philosophy of Language,

Context

Last week we were looking at Grice’s two contributions to the philosophy of language; his ideas, first, of a distinction between speaker’s meaning and sentence meaning, and second, of a type of entailment in normal language use that was distinct and not reducible to so-called logical entailment. I said that I thought that this was an example of a type of approach to language that formed one of two broad categories of approach, and I contrasted it with the truth-conditional approach, which I also thought shared a philosophical kinship of some sort with the other naturalizing approaches that we’d seen – but I thought then that I couldn’t do much more than wave my hands in the general direction of a characterization of the distinction. Now, however, we’ve looked at the Strawson reaction to Russell (that is that people mean things when they use sentences rather than sentences meaning things); and we’ve looked at Wittgenstein’s reaction to, well, to Wittgenstein actually, but also to Russell and Frege (that meaning is just use and language is just like a game); and, as I say, we’ve looked at Grice’s claim that the ‘meaning’ of our utterances has to be defined in terms of the intentions that we have when we make the utterance, rather than being reducible to the meanings of the sentences that we utter. With all these examples before us perhaps it’s a little easier to identify a common thread here.

I think that what links all these approaches together is that language use is treated as a type of intentional action. Never mind what the actual language or system of signs looks like, any theory about meaning as expressed in the system is going to have to be a theory that talks about the beliefs and desires of the users of that system, as well as the rules that apply to its use (although some people also think that all talk of rules can also be eliminated as they can be derived from intentions of the rule followers – Grice is one of those people, but I won’t go into this here.) In this view language is just a really, really sophisticated version of the sort of miming and gesturing that we perform when we’re trying to communicate our intentions to children, foreigners, and other handicapped persons. So, if I stamp my foot, glare at someone who is annoying me, and point at the door, that person will interpret those actions as trying to convey my desire that they be elsewhere. If I say, ‘Michael, I find you annoying, Please leave.’ (This is only an example.) I am doing exactly the same thing. Considered purely pragmatically, as types of actions, there is no difference in principle in the natures of these acts.

This seems reasonable. We know that we can have meanings in utterances that are at wide variance with the meanings of the sentences that are uttered. I mentioned this before, when I used the example of my statement that “I’m really looking forward to reading all your long essays!” This could be said seriously, in which case the utterer’s meaning is identical to the conventional meaning; or it could be said sarcastically, in which case it means exactly the opposite; or it could be said in some idiosyncratic context, in which case it could mean just about anything. On the other hand, as I’ve also mentioned before, the sorts of things that I can mean by uttering that sentence seems to be largely dependent upon the meaning of the sentence itself. The only reason that the sarcastic way in which I uttered the sentence means that I’m not looking forward to reading your essays is because the sentence itself means that I am looking forward to it.

Nevertheless, this new understanding of communication – or speech – as a kind of action may be taken as a guide to further study. If we take this seriously as a way of thinking about language, we should, perhaps, try to set up classificatory schemes that will allow us to talk about the various types of utterance. Something of this nature can be found in Austin’s work, which is what we’re now going to have a look at.

Austin's Taxonomy

Utterances: Constative or Performative

To begin with Austin makes a distinction between constative utterances and performative utterances.

a.                   Identification

The distinction is supposed to be that a constative statement is just the sort of declarative statement that we are accustomed to talk about in philosophy of language courses. They are sentences like ‘The grass is green’ and ‘The sky is blue’ and ‘Scott is the author of ‘Waverley’’ and ‘The King of France is Bald.’ They are the sort of sentences for which it can actually be made plausible that their meanings are entirely reducible to their truth conditions.

On the other hand performative sentences are sentences like:

                I promise that I’ll come to your party.

                I bet you $10 that I can beat you at tiddlywinks.

                I pronounce you man and wife.

                I apologize for being late.

These sorts of sentences don’t seem in the first place to be stating facts about the way the world is - -although when one hears them one does know something new about the world. The statements do say that one has promised, laid a bet, married, or apologized; but what seems curious about the statements is that what they state to be the case is the case only in virtue of the utterance being uttered. If I hadn’t said ‘I promise …’ then I would not have promised; and similarly for all the other utterances. Actions like these Austin calls Speech Acts.

Of course, if you have a distinction like this you really also need to have a reliable way of deciding whether a particular utterance belongs to one or the other category. Austin thought that the way to do this was to apply a kind of paraphrase test: an utterance with a verb, like ‘promise’, ‘bet’, and so on, is a performative utterance if the verb can sensibly be prefaced by the word ‘hereby’ without changing the meaning. Thus for the examples:

                I hereby promise that I’ll come to your party.

                I hereby bet you $10 that I can beat you at tiddlywinks.

                I hereby pronounce you man and wife.

                I hereby apologize for being late.

The ‘hereby’ test is failed by constatives. ‘The grass hereby is green’, ‘Scott hereby is the author of ‘Waverley’’ are ridiculous statements. While there is a sense that can be given to them, it is a sense that seems obviously bizarre, since no one in their right mind thinks that merely saying that it is so is what makes the grass green, or Scott the author of ‘Waverley’.

b.                   Performative Rules: Regulative and Constitutive

Now, it’s clear that these speech acts are dependent upon a good deal more than the form of words in the utterance. If I said now that I promise to give you the moon, or if I said to any pair of you that I now pronounce you man and wife, these would not typically be considered to be successful speech acts. In fact you might even consider that they were not speech acts at all. In the first case, no one can be held to a promise to do the impossible, especially when the promiser and the promisee both know that it’s impossible; and if there’s no expectation or requirement that the conditions specified in the utterance will be met then it doesn’t seem that anything like a promising has occurred. The expectation of performance and idea of an obligation to perform seem to be constitutive of a promise having been made. This seems to indicate that there are rules which govern what sorts of utterance can count as promises which go beyond the simple rules of grammar. Similarly, when I pronounce any number of you married, this will fail to be a speech act because there are rules beyond merely grammatical rules that govern what sorts of utterance can count as a marriage-making utterance, and my utterance fails to satisfy those rules. For instance, I am not an ordained minister of any recognised religious order, and I am not currently in a social situation in which a marriage is occurring.

This is hardly surprising, for speech acts are like any other socially defined action: they have to obey the rules that apply to them. However, Searle has alerted us to the fact that there are at least two different ways in which rules can apply to such actions. As we have seen, some rules that apply to actions are such that if they are violated then the action simply does not occur. Such rules are said to be constitutive. My being an officially authorised figure is a constitutive rule for my utterance of ‘I pronounce you man and wife’ to be the speech act of marrying someone. On the other hand, there are rules that can be violated without entirely annihilating the action. They are rules that regulate the successful performance of the action and are therefore called regulative rules. For example, if a minister performs a marriage ceremony in which both parties recognise that they are merely doing it to get the tax advantages, then although the marriage does occur, it is not a good marriage. If I promise to play a game of squash with someone just to make them stop pestering me about it, and I have no intention at all of turning up to play, then I have made a promise, but it is a bad one. If I grant an extension to someone, but I mumble it so that they don’t hear me, then I have granted an extension, but I have done it badly. As you could guess from those examples, there are all sorts of ways in which the regulative rules can be violated and all sorts of ways in which the actions that are performed may be ‘bad’ ones of their kind. Austin calls these failures infelicities. 

Force: Locutionary, Illocutionary, and Perlocutionary

Now it turns out that this distinction was unsatisfactory, but before we look to see what is wrong with it, it’ll be necessary to have a look at a distinction that Austin drew regarding what he called the force of utterances. This ‘force’, as you’ll see, is related but not at all identical to what Frege called the force of a sentence. For Austin, there are three types of force, and any single utterance may be analyzed in terms of all these forces:

a.                Locutionary Force

This  is the simplest, and is to be identified with the propositional content of an utterance. It is what we think that various utterances like:

                The window is shut.

                Shut the window!

                Is the window shut?

                Would that the window were shut~

have in common.

Although this is a fairly simple concept, Austin did manage to complexicate it a little (which is the mark of a great philosopher.) He introduced a further subdivision of the acts that can be identified in any act which has locutionary force. These subsidiary acts are:

i.                     Phonic

At the most basic level, the Phonic act is just the act of making the noises, and the noises that are produced are phones. (Presumably, for someone who is using sign language the phonic act would be just making the movements, and the movements would be phones; for an octopus the phonic act is flashing patterns of colours over the skin, and the patterns would be phones.)

ii.                   Phatic

At a slightly more abstract level, the Phatic act is

the act of uttering certain vocables or words, i.e. noises of certain types belonging to and as belonging to a certain vocabulary, in a certain construction, i.e. conforming to and as conforming to a certain grammar, with a certain intonation, etc.

The point of saying something is uttered ‘as belonging to’ a certain vocabulary/grammar is presumably to indicate that the intention in uttering the vocables was to follow the rules of the vocabulary/grammar. It is not enough just that they do belong; the fact that they belong must be part of the reason that they were uttered. It is difficult to say just what Austin was thinking about when he talks about Phatic acts and phemes, but it seems to me that a plausible version of what he should have said (and might even have meant to say) is that the phatic act is the act of using a language, the act of fitting the vocables produced as phones by the phonic act to the rules of a language. The product of a phatic act is a Pheme. On my reading a pheme could be the utterance of a grammatical sentence or phrase.

iii.                 Rhetic

Finally, the Rhetic act is

The act of using that pheme or its constituents with a certain more or less definite “sense” and a more or less definite “reference” (which together are equivalent to “meaning”)

Again, there is some controversy about what is meant here, but I think that it’s obvious that what he was really getting at was that the rhetic act was what finally produced the propositional meaning that is supposed to be expressed by an utterance of a sentence – i.e. the production of a pheme. The product of a rhetic act is a Rheme. On my reading of things a rheme could be the expression of a meaning. Certainly something like this has to occur at the final stage of Austin’s deconstruction of the locutionary act, because,  if you’ll remember, the entire point of the locutionary force of an utterance is to express a proposition.

b.                Illocutionary Force

So much for locutionary force. More interest attaches to Austin’s proposal that utterances also have what he called Illocutionary force, by which he meant: that which is done when the utterance is made. Sometimes the grammatical form of the utterance – in Austin’s terminology, the evidence of the pheme – is sufficient to indicate what the illocutionary force of an utterance is supposed to be. For example, when we see the sentences that introduced this section:

                The window is shut.

                Shut the window!

                Is the window shut?

                Would that the window were shut~

we can see immediately that in normal conditions they will have the illocutionary forces of, respectively, asserting, commanding, enquiring, and wishing. Of course, the interesting thing is that most utterances can be associated with a variety of illocutionary forces. (Maybe all utterances can be associated with a variety of illocutionary forces; maybe all utterances can be associated with any illocutionary force – but let’s not get overexcited.) For example depending on the circumstances in which the simple declarative sentence ‘Help is on the way’ is on the way is uttered the illocutionary force may be simply assertive (e.g. I have called the fire department), or promissory (e.g. hang on there I intend to do my best), or judgmental (e.g. I see that you are incompetent, well, I’ll look after you) and so on. Utterances that look like questions may have the illocutionary force of accusations (e.g. Are you mad?)

c.                Perlocutionary Force

Finally, Austin describes the Perlocutionary force of an utterance, which is what is actually achieved by an utterance. For example, when I say ‘I wonder: should I give a surprise exam at the end of this lecture?’ the locutionary force is provided by the rheme expressed by the sentence ‘I’ll give a surprise exam at the end of this lecture’; the illocutionary force is interrogative, I’m enquiring about the possibility; the perlocutionary effect is alarming, I have frightened my dear student chums.

Utterances: Force and Content

Now let’s return to the division of utterances between constative and performative. What would Austin have to say about utterances like

                I state that the time is 12:45

This seems to be a performative on Austin’s rough definition because when I state that I am stating something then I really am stating it, given that the constitutive rules for making a statement are followed. And yous’ll notice that

                I hereby state that the time is 12:46

passes Austin’s proposed ‘hereby’ test, because it makes sense and doesn’t seem to change the locutionary or illocutionary or perlocutionary force of the utterance. On the other hand, there’s more to the utterance than just that performative, because it is actually making a statement and has a certain constative aspect.

Because of examples like this, as well as examples involving reviewing, judging, critiquing, demanding, etc. Austin came to believe that many, perhaps all, utterances had two aspects – a performative and a constative aspect. In fact, we might conclude that the appropriate way to analyze an utterance was in terms of the ‘content’, which covers both the locutionary effect and the constative aspect of the utterance, and the ‘force’ – meaning the illocutionary force – which covers the performative aspect. Thus each utterance could be analyzed as

                U = <C, F>

For example:

                The window is shut.                                   < S(w), dec. >

                Shut the window!                                        < S(w), imp. >

                Is the window shut?                                   < S(w), int. >

                Would that the window were shut~         < S(w), opt. >

and in this form we start to see how we can get back to the one true semantic analysis in terms of logical forms and interpretations and so on (Just Kidding!)

But this brings us to a problem pointed out by L. Jonathon Cohen. On that way of looking at utterances one is strongly tempted to provide the following utterances with the following analyses

                I am pleased to announce that …               < S(w), announce. >

                I regret to inform that ….                              < S(w), inform. >

                I [name], being of sound mind and body, do hereby declare that …

                                                                                          < S(w), declare. >

in which the content is entirely restricted to the subordinate clause, and the prefatory material associated with the ‘performative’ verb is treated as a mere marker for the type of performative. It simply tells us what the illocutionary force associated with the ‘real’ content of the utterance is going to be. But how plausible is this? There seems to be an increasing amount of actual content, actual relevant information, in the prefatory material that is ignored. And yet, if we take the other option of including that prefatory material as content of the utterance then almost every such utterance is going to take the form of a simple declaration. And in almost every case it will be a true declaration. For example suppose that I say

                I declare that today is my birthday

Now, you happen to know that it is not my birthday, so you say to me: ‘You lie like a dog!’ But you are wrong. I am declaring something to be the case. And it is true that I am declaring it to be the case.

This is a curious puzzle, and you should think about it a great deal over the next few months.