The ancient Greeks constantly harped on the contrast between words and actions, provoking Housman's parody in his Fragment of a Greek Tragedy
`Oh! I am smitten with a hatchet's jaw,
And that in deed and not in word alone.'
It seems a simple and basic distinction, but when one thinks about it it's not so simple as it looks. If I say `John promises to do that' I am simply reporting John's action of promising; but if I say `I promise to do that' I am actually doing the promising by saying so. Certain forms of words are actions as well, and not just in the trivial sense that to say something is to perform the act of saying something. Moreover, forms of words that seem very similar in meaning turn out not to behave in identical ways. `Apologise' behaves much like `promise', in the sense that when I say `I apologise for my behaviour' I am performing the act of apologising. However when I say `I am sorry for my behaviour' I may or may not be apologising - I may be reporting my feeling of sorrow, as if I had said `I am sad about my behaviour.'
The general idea is very easy to grasp, but the amount of variety in the ordinary expressions we use can seem mind-boggling. What's the story on `bequeath' for instance? If I say in my will `I bequeath you 1 million $' and if I have 1 million $ to bequeath you then I am performing the act of bequeathing by saying so. However if I don't have it I am bequeathing you nothing , whatever I say. Similarly, if I say `I anoint you Archbishop of Peoria' simply saying so doesn't make you that. In the first place I need the authority to perform this act, in the second place I need something to anoint you with, and in the third place you need to be willing to be so anointed. However even if all these conditions are present I will still not have anointed you Archbishop unless I also say so. It all goes on and on. If I am your commanding officer and I say `I reprimand you' I am thereby carrying out the act of reprimanding. However if I say `I insult you' and leave it at that I have done no insulting. Again, if one says `In saying that he made a mistake' this does not mean that the person's act consisted of something called `making a mistake'. And so on.
The series of twelve lectures in this book hauls us through any amount of fine and subtle detail about these so-called `performative utterances'. Normally the best way to read a book is to start at the beginning, but that's not what I'd recommend here. Once you have the general idea (even from this short review) I'd say start at the last lecture, go on to the second-last, and only then go back to the start. If you plough through it starting at page 1 it seems a bit of a catalogue of instances, almost as if linguistic philosophy is reduced to sweeping up after some majestic cavalcade of lexicography has passed by. Austin is always Austin of course, not just lucid and brilliant but witty too - there is one of his inimitable mixed metaphors somewhere, something about letting cards out of the bag or putting cats on the table. However after a while one yearns for a top-down perspective, for generalisation. That comes in the final two chapters. The most important statement in the book is in chapter XI, where he says that `...what we have to study is not the sentence but the issuing of an utterance in a speech situation.' That may not be Austin's most felicitous expression, at least not when quoted out of context, but it enshrines his basic argument, one that holes much ordinary linguistic philosophy below the water-line, that verbal expressions on their own do not enable us to understand what is said. Indeed I wish he had gone further in pointing out that non-verbal factors, such as tone of voice or facial expression, can cast doubt on what the verbal expression is ostensibly saying. I could, for example, say `Oh I do apologise' in such a manner as to make it very clear that I mean nothing of the kind.
In the last chapter Austin produces a short set of categories of expression in an attempt to classify the mass of detail in the foregoing chapters. He does not profess to think them anything but provisional, and the terms he coins are monstrosities - behabitives, expositives, verdictives, exercitives and commissives. Be not afraid. He explains them with all his characteristic clarity, and when you have seen the outlines of the wood you can go back to the beginning and inspect the trees individually. As always, Austin is a spoiler, and rightly so. He trains his guns on the illegitimate tyranny of `true and false' that has bedevilled so much philosophical thinking, saying that these terms constitute `a dimension of assessment' and do not stand in some supposedly unique relationship to `facts'. This is only a review, and if you want to know how he means that you have to read him for himself. For me, Austin's way of putting things is enjoyable and his thinking is liberating to the mind. Much philosophy is, in another of his great expressions, barking up the wrong gum tree, and I am only too grateful that Austin lived long enough to save us from the same fate.