This is a book review

It’s striking nowadays to hear a philosopher say that ‘we want a unified account of our knowledge’; even more striking to hear him say ‘I think we can get it’; very striking indeed to hear this from a philosopher of language. That wouldn’t always have been so. A hundred years or so ago, there was great enthusiasm for looking closely at the structure of sentences and at the distinction Frege had drawn between their sense and reference (the difference between saying that Mark Twain wrote Huckleberry Finn and saying that Samuel Clemens did – Clemens was Twain’s real name – where the sense, the cognitive significance, is different but the reference is the same); a great will, too, to separate sentences that were true by definition from those that weren’t, and among those that weren’t, to admit only those that could be independently verified. This was at the heart of the logical positivism of the Vienna Circle for which Otto Neurath wrote the manifesto in 1929. And Neurath persuaded his colleagues (and the University of Chicago Press) that it was possible to bring all that was positively known into an International Encyclopedia of Unified Science. The unification would be provided by the purified language of meta-theoretical propositions. (Twenty monographs for the encyclopedia appeared between 1938 and 1969 but only two of the foundational volumes were published before the project eventually lapsed.)

The Vienna Circle’s radicalism, however, went only halfway round what came later to be called the linguistic turn in 20th-century philosophy. Within two years of writing his manifesto, Neurath himself declared that the idea of statements representing things in the world and being verified by experience was unacceptably metaphysical: ‘reality’ consists solely in the sum of true sentences and the relations between them. Language itself, indeed, with all its metaphysical baggage, was suspect, and Neurath, following the model of mathematical physics, proceeded to advocate a new way of doing propositions in what he called the Vienna Method of Pictorial Statistics, which his wife Marie refashioned as an International System of Typographic Picture Education. (She continued the work in Oxford after Otto’s death there in 1945, until her own, in London, in 1986; it is now archived at Reading.)

Neurath’s more enduring holistic conception of a language which we have no reliable way of connecting to the world was taken up by Quine in America and by the later Wittgenstein. Two men see a lolloping object, Quine imagined. One exclaims ‘rabbit’, the other ‘gavagai’, and there’s no way the first man can know whether ‘gavagai’ means ‘rabbit’ or something like ‘undetached rabbit part’ or ‘temporal slice of rabbithood’. (Quine also produced a powerful objection to the distinction between statements that were true by definition and those that were not.) In the 1920s, the early Wittgenstein’s Tractatus Logico-Philosophicus had been one of the Vienna Circle’s canonical texts. But it’s said that even then he’d insist on reciting poetry at their meetings, and by the 1950s, he too had taken the full turn to claiming, or so it could seem, that we are our language, though neither he nor Quine followed Neurath and his radical friends in the 1920s and 1930s in pressing the thought that a firm, clear, shared language would show the way to a true socialism. And although each took a holistic view of language, neither showed any interest in a unified account of the world.

All these philosophers, however, did assume that the point of language was to describe the world, however difficult or indeed impossible it might be to know that we’re doing so. So too did Husserl’s phenomenology, though from a different direction. Husserl concentrated on consciousness, which he saw, in his word, as intentional, directed towards objects in the world. We may exercise our consciousness in many ways, in hoping, declaring, commanding, promising and so on, as well as describing or referring, but Husserl regarded all such performances as one or another kind of objectifying act. For him, my declaring that you’ll get to the end of this piece would be to say that ‘your getting to the end of this piece is my declaration.’ That may prompt you into thinking that you might not, and not just because you don’t like being declared to. The reformulation, you will rightly think, misses the point. I’m not just saying something to myself. I’m saying something to you. The declaration is a social act. But what may be obvious to us now had not been so in phenomenological circles in Germany, and when Adolf Reinach suggested it in 1913, he caused a stir. Even Husserl, who was not known for listening to others, was impressed. But Reinach was killed in action in Flanders in 1917, aged 34, and although his thinking was revived in Munich after the first war, where it was described as the study of ‘speech acts’, it was not until after the second that it was developed by J.L. Austin in Oxford.

In what he described as the preliminary ‘cackle’ of his presidential address on excuses to the Aristotelian Society in 1956, Austin said that all he needed for his own ‘linguistic phenomenology’ of speech acts was a dictionary, distinctions in case law, especially tort, and where it might illuminate ‘the language of ordinary life’, some psychology also. ‘With these sources, and the aid of the imagination, it will go hard if we cannot arrive at the meanings of a large number of expressions and at the understanding and classification of large numbers of “actions”.’ But at the end of the address he doubted whether ‘true meanings … will fit into place in some single, interlocking, conceptual scheme. Not only is there no reason to assume this, but all historical probability is against it, especially in the case of a language derived from various civilisations as ours is.’ In any case, he said elsewhere, were a general philosophy to emerge from what he was doing, it would likely play ‘Old Harry’ with the long-standing fetishes of fact and value and with the very idea of the true and the false. But he died, at just 48, in 1960, and did not himself get to such a philosophy.

John Searle studied under Austin in Oxford in the 1950s, then went to Berkeley, where he has remained. He published his first book, Speech Acts, in 1969 and is Austin’s pre-eminent heir. He’s certainly not fussed about the distinction between statements of fact and of value; that’s plain in the purportedly political chapters at the end of this, his 13th book on a philosophical subject. He’s stayed with common sense. ‘If somebody tells you that we can never really know how things are in the real world,’ he said in a published conversation ten years or so ago, ‘or that consciousness doesn’t exist, or that we really can’t communicate with each other, or that you can’t mean “rabbit” when you say “rabbit”, I know that’s false.’ And he pays attention to the way we use language. But he’s always had his eye on a grand theory, insisting that any account of what we do with words should be consistent with the fact that particles and molecules are the stuff of reality. He has met this ‘basic requirement’, he claims, in what he’s written about intentions, minds and rationality, and he does so now in bringing his thoughts together in what he describes as his ‘philosophy of society’.

Aristotle, Durkheim, Weber, Simmel, Habermas, Bourdieu and Foucault are all dismissed in a sentence here, Locke in a footnote. They all took it for granted that we are language-speaking animals, and were then ‘off and running with an account of society, social facts, ideal types, political obligation, the social contract, communicative action, validity claims, discursive formations, the habitus, bio-power and all the rest of it’. None of them stopped to ask what language is. Searle does, and starts his inquiry at a point before it came about, in a state of nature populated by pre-linguistic hominids. They were not witless. They had beliefs, desires and intentions, and expressed themselves. They got through their day with what Searle mischievously describes as a whole bundle of Aristotelian and Kantian categories: space, time, individuation, object, causation, agency. Not with concepts as yet, or sentences. These only arrived when grunts and gestures became language. Nonetheless, four of the five kinds of intention (Searle believes there are only five) that we their descendants have in speaking – to represent things, to direct others, to commit ourselves and to express our feelings – have their analogues in pre-linguistic consciousness. Only the fifth kind – to declare something to be the case – does not. This is what distinguishes language users, and Searle’s argument turns on it. In social matters, language doesn’t merely reflect what exists: it can make it do so.

He accepts that this is not a new notion. Rather like Rousseau at the start of the Discourse on Inequality, he is suggesting that the first speech-acting hominid to declare ‘this is property,’ or ‘a husband’, ‘a king’ or whatever, and found others to believe him, started something. Searle’s argument is that Rousseau, like all the other philosophers, didn’t appreciate what he was saying. Names by themselves don’t create institutions. That requires what Searle calls a ‘status function declaration’: someone or something is declared to have a status in virtue of which he, she, other people or it is committed to perform one or more functions. What you are reading is a review on a sheet of paper. The second part of that statement presents no problems. A sheet of paper falls readily into the class of what Austin nicely described as medium-sized dry goods: it exists independently of anybody saying so, and were the pre-linguistics to have had paper, its objectness and individuation would have been evident to them in space and time, even if they couldn’t have given it a name. A review does not so exist, and could not have been evident in that way. But if you agree with me about what you’re reading now, you do so because we’ve come to accept that a certain kind of writing in a certain kind of place counts as a review rather than as a letter, say, or a short story. And if there was no previous sense of such a thing (though in this country at least the practice of reviewing books dates back at least to the 1640s, and the existence of publications containing reviews to 1700 or so), I could try to call it into existence by asserting that reviews do exist, and that what you have before you is an instance of one. In Searle’s way of putting it, I’d be declaring that X (what you’re reading now) counts as Y (a review) in C (some accepted set of circumstances).

This doesn’t, of course, say what a review is. For that, if the idea was quite new, I’d have to say what its function is and spell out the conventions it observes. In the case of reviews, we might believe that we know these things. As Searle puts it, we may have a ‘collective intention’ in using the term. So far, perhaps, so good, though the notion of a ‘collective intention’ can make one nervous. But can we be said to know about more casual practices: about letters, say, or love affairs, parties, dancing or walks together in the park? Searle thinks not. We do have performative verbs for some of these practices, and when we use them we may think we know what we’re doing. ‘I’ll write to you’ suggests a letter, ‘let’s dance’ dancing, ‘let’s take a walk in the park’ a polite kind of stroll. For others, we have no such verb. ‘Let’s party’ might just do but ‘let’s love’ would be odd. If there are conventions for doing such things they’re no more than discretionary. They’re not part of what Searle thinks of as ‘the structure of human civilisation’. That requires firm rules. I can certainly declare that X (what you’re up to) counts as Y (a love affair) in circumstance C (what exactly?), but why should you take any notice? You know there are no rules for such goings on. But for less casual practices, call them ‘public’ rather than ‘private’, we believe that there are. Take states. If someone declares that X (Flanders) counts as Y (a state) in circumstance C (Belgium doesn’t count as one, the Netherlands doesn’t want Flanders, its inhabitants do, and others give in), where X is specific and C is clear, the Y term commits Flanders to govern as a state. In this respect, declarations are for Searle like commissives. To declare that something is a state is in this respect analogous to one of those performative acts, like promising, which binds. Reviewing is one of the many practices that stands somewhere in between. It has its conventions, but no firm rules. Yet to say ‘I review’ commits me to the practice of reviewing, however indeterminate that may be, just as agreeing that Flanders is a state would commit it to the rather more determinate business of governing. The commitments are internal to the declarative act, and commitments, for Searle, are obligations. Status function declarations entail deontic powers, these constitute institutions, and it’s in institutions that the wonder of civilisation consists.

The special wonder, for Searle, is philosophical. Institutions are created by speech acts, yet become ‘non-linguistic facts’. Someone somewhere at some time ascribed a status and a function to ‘review’ and thereby conferred powers and obligations that go beyond the semantics of declaring this or any other piece of writing to be a review. And in going beyond semantics, these powers and obligations are by definition not linguistic. A review is subjective in its creation and its substance, in what a philosopher might call its ontology, but epistemologically, in the claims that can be made about its existence, it’s as objective as can be. A word has become a part of the world about which we can make true and false statements. And if we could trace this kind of thing back to the people who more or less overtly declared it to be so and back even further, Searle would expect us to find, as he likes to put it, that it ‘bottoms out’ in those brute facts of our biology that give us the capacity linguistically to create.

This is not to say that Searle is engaged in an evolutionary anthropology. His ‘state of nature’ is just a construct to think with. But he says that he does like to treat problems in philosophy as problems in engineering, and at this point in his argument, one might wonder whether he hasn’t been too successful. Beyond the twilight world of private life, he can seem to make us victims of our own inventions: his civilisation of obligation-inducing institutions suggests one big mouse-wheel. Yet we’re conscious of being free. He is pleased to allow this, although he’d prefer us not to go on talking about ‘freedom of the will’, a notion which, he says, has too ‘sordid’ a history in philosophy to be acceptable any longer. He suggests that we think instead of ‘the gap’ between what causes us to act and our actually doing so. This is nothing so simple as the capacity to jump off the wheel, to say no to an institutional commitment. It’s the space in which we exercise our reasonableness, and it’s the commitments and obligations that come with institutions which give us the ‘desire-independent’ reasons we need if we’re to be ‘reasonable’. (‘Desire-independent’ against those like Hume and Bernard Williams who have argued that external reasons to act are effective only if they fit a pre-existing desire, an internal ‘motivational set’.) Institutions need reasonableness and reasonableness needs institutions. Without institutions and the reasons they give us to act, without things like states and banks and schools and journals of literary discussion and all their various requirements, we’d be at the mercy of mere inclination, racketing about in the twilight, innocent of the advantages and pleasures civilised life provides.

Searle’s account of how language constructs institutions is certainly ingenious. We declare that something has the responsibility for this or that ‘function’ and discharges it by conferring rights, obligations, requirements, authorisations and so forth on those involved with it. Often, in forming a new company, say, or a society for the preservation of endangered spiders, the ‘something’ is itself something that we create. An institutional purpose or function, an initially ‘free-standing’ Y term, as he would put it, will lead someone to call an X into being. What matters is the difference between the ‘conditions of satisfaction’ for references made to the spiders and references made to the preservation society. The meaning of the sentence ‘we preserve spiders’ is satisfied by the statement itself. The meaning of a sentence that includes ‘the preservation society’ (but not what it preserves) is not. The society would have had already to be represented in the declarations that created it. And having thus been created, it carries rights and duties and so on, some of which will be its own, some of which will derive from others (in this case from the law on charities). This is the way Searle solves the riddle of how subjective creation can licence claims of non-linguistic, objective fact about what it creates. And although he’s not particularly concerned to do so, he can also account for the way new institutions can derive from the old, most obviously by fission. But whose declarations become authoritative, and why? And who or what polices what happens in the gap between the desire-independent reasons to act and actually acting? What about power?

Searle argues such questions away. Someone suggests an institution through which we can get something done. We may accept the suggestion or we may not. If we do, we collectively accept the powers of authorisation, rights and obligations that are entailed in doing so. Powers also pervade what he rather limply calls the ‘background’, itself the product of declarations few of us have heard, which tells us where and when we are permitted to take our clothes off (or keep them on), or to laugh, or shout at the moon. Politics, he suggests, governs the creation of institutions and relies on the ‘background’. The state and its apparatuses are the outcome of a status function declaration whereby certain public powers have authority over others. In the United States and Western Europe at least, we can be confident that there is a ‘background presupposition of tolerance of disagreement’, so that those who compete for offices from which to exercise these powers can’t afford to stray too far from the middle way; after all, he observes, the outcomes of elections don’t seem to make much difference. All public functions and institutional obligations, he appears to believe, are compatible, or where they’re not, immaterial. (One’s tempted to retort, as Sidney Morgenbesser did when Austin told his New York audience that no double positive ever meant a negative: ‘yeah, yeah.’)

Searle risks making ‘collective intentionality’ do the work that ‘general values’ used to be asked to do in an old, apolitical, non-linguistic sociology. In his one essay here on value – specifically, on human rights, declarations that X counts as a right in the circumstance of being human – he is similarly brisk. Rights are either given by his central argument that reasonable speech actors must have freedom of speech, or are a matter of practicality (not everyone can have a positive right to all the goods of life) or a frank expression of taste (he’d like to be able to claim a right to silence against those who make life so noisy).

Yet the concept of ‘the political’, he concedes, is clearly what Wittgenstein called a ‘family resemblance concept’. There are overlapping similarities in all the goings-on we think of as political, but there’s no set of necessary and sufficient conditions that defines them all. And one may ask whether this is not the case for most of the rest of social life, including its institutions. It’s not difficult to accept that Searle’s linguistically created non-linguistic facts come to exist by virtue of what we may think of as a more or less hypothetical declaration. Nor is it difficult to accept that where the existence of such a fact is marked by a constitution, it will be irreversible and obligatory unless the constitution itself says otherwise. But as Reinach remarked in 1913, few of our practices are constituted by constitutions. Even where they are, what a constitution enjoins will be as much permissive as obligatory. And if a constitution should attempt completeness, its provisions will have to be general. For all these reasons, our rules, conventions and tacit ways of going on will be open to continual interpretation, argument and revision. Lawyers will be asked to tie things down. The more ambitious and imaginative will want to reshape what we do. And people will all the time make more or less knowing changes of less noticeable kinds. There will be continual fights and fuzziness, and surprises too. Yet this side of a thoroughgoing revolution, most of the institutions and less formal linguistically created non-linguistic facts of social life will continue, and most manifestations of these will bear a family resemblance to the ones that have gone before. We may certainly accept, with Searle, that if we speak carefully, none of these facts is obviously inconsistent with the further fact that at bottom, we are particles and molecules. The one non-natural entity we need, if it is one, is the faith that we’ll one day get to know how they’re not.