The Harvard University Press asked ‘the most distinguished and influential of living philosophers’ (Strawson’s description of Quine, on the dust-jacket) to produce a collection of loosely-connected essays on topics of his choice in a format inspired by Voltaire’s Philosophical Dictionary – and the result is a remarkable addition to English literature. Quine is not only a great philosopher, but also a master of the English language and a genuine polymath; and the ‘dictionary’ format – more than eighty articles ranging from A (Alphabet) to Z (Zero), and including entries on Belief, Communication, Free Will, Idiotisms, Longitude and Latitude, Marks, Prizes, Latin Pronunciation, Tolerance and Trinity – gives him ample opportunity to write (and, he tells us in the Preface, to have ‘more than half the fun’) about ‘lowlier themes’ than philosophy (which occupies less than half the book). Apart from philosophy, the subjects most fully represented in the book are mathematics, logic and language (including English etymologies, stylistics, and the philology of the Romance languages), but there are also many short essays in which Quine pokes fun or grumbles good-naturedly about various pet peeves. (The essays on Artificial Languages, Extravagance, Mathematosis, Usage and Abuses are wonderful examples.) Perhaps the most charmingly lighthearted essay in the book is the one titled ‘Misling’. Many people have been misled to pronounce ‘misled’ as ‘mizzled’. ‘But the verb misle that is born of that misconception is too pat to pass up, descriptive as it is of the very circumstance that engendered it,’ Quine tells us. ‘Perhaps we can press it into service as a mild word for the restrained sort of deception, not quite actionable as fraud even in Ralph Nader’s day, that has a respected place in enlightened modern merchandising.’ Although Quine steers clear of political themes for the most part, there is one beautifully formulated statement of his conservative creed – the essay on Freedom. An added charm is the not quite self-deprecating humour exhibited by some of the best remarks in the book, as when Quine writes (in the essay on Communication):
Examples taper off to where communication is less firmly assured, as when Hegel writes, ‘Truth is in league with reality against consciousness,’ or I write: ‘Logic chases truth up the tree of grammar.’ I am confident that I grasp and appreciate this message of Hegel’s, and that there are philosophers of logic who grasp mine. But mere acknowledgment, however sincere – ‘I dig you,’ or ‘I read you. Roger and over’ – is not conclusive evidence of successful communication. The Latin pupil gets low marks who says: ‘Oh, I know what it means but I can’t quite put it into words.’
In spite of the immense range of subjects that Quine covers, there are very few errors that I could detect. (The most surprising is the totally wrong statment about quantum physics which comprises the last paragraph of the essay no Discreteness. Time in quantum mechanics is not discrete, and Planck’s constant, in any case, is not in units of time.) Not only is there very little error in this book, but there is much that was discovered by Quine himself (although Quine modestly refrains from saying so), including the beautiful combinatorial equivalent to Fermat’s last Theorem.
If Quine had ‘more than half the fun’ writing the non-philosophical essays in the Dictionary, still the philosophical essays – perhaps because of their very informality – give a remarkably good picture of how Quine pictures the universe, and I must devote much more than half of this review to them. Quine is often considered to have destroyed Logical Positivism, with his rejection of the analytic-synthetic distinction and his likening of philosophy to natural science rather than to pure logic; and, indeed, a generation of young ‘scientific realist’ philosophers has been inspired by him to denounce Logical Positivism root and branch. But reading these essays, I must say that I am inclined to class Quine as the last and greatest of the Logical Positivists, in spite of his criticisms of the movement. Not only is the reverential appraisal of the philosophical achievements of modern logic still there – ‘Gottlob Frege, however, seems to have been the first to offer a coherent account of what [the numbers] are,’ Quine writes in the essay on Natural Numbers – but so, I seem to detect, is something of the Positivist picture of the world as a system of ‘posits’.
The External World as a Construction
The greatest work produced during the existence of the Vienna Circle was Carnap’s The Logical Construction of the World, and Quine describes Carnap’s views very sympathetically. (Carnap, Quine tells us in the essay on Things, ‘gave us a masterful scheme or caricature of how the maxim [‘posit the simplest and laziest of all worlds compatible with our observations’] governs our conceptual construction of the world’ (emphasis added). Physical objects are, fundamentally, constructions. Our purpose in introducing them is to store up ‘observation categoricals’ in a logically compact form, where ‘observation categorical’ is Quine’s technical term for a general if-then assertion whose antecedent and consequent describe observable situations, as in ‘Whenever there is a fire at a place and one’s wristwatch shows ten o’clock, then there is smoke at that place when one’s wristwatch shows ten minutes after ten.’
Quine is not saying that our sole aim is to make predictions. We want and need correct observation categoricals, of course. But we also want and need a tidy system. What we want is a system of generalisations from which as many correct observation categoricals (and as few incorrect ones) as possible can be derived. In short, we also want both simplicity and generality.
This desire is, I think, what Quine means when he says that we want not just successful prediction but satisfaction of pure intellectual curiosity. (‘This is not to say that prediction is the purpose of science ... an overwhelming [purpose] is satisfaction of pure intellectual curiosity.’) This quotation from the essay on Prediction might seem to say that we want to make true statements about the way the world really is: but a look at some of the other articles (Reification, Truth, Reference) quickly dispels the illusion that Quine allows any notion at all of ‘how the world really is’ that transcends the (traditional positivist) idea of a simple system of general laws that leads to successful prediction. There is, to be sure, one other desideratum besides generality, simplicity and predictive power: we inherit a body of past doctrine and, like William James, Quine attaches value to ‘minimum mutilation’ of this past doctrine: but this is an internal constraint on the acceptability of a construction of the world. The world, as Quine views it, seems to be a human construction.
Quine’s view of truth is (fundamentally) Alfred Tarski’s (as well as Carnap’s). There is no general philosophical problem about truth, according to these thinkers. If we list all the sentences of the form ‘S’ is true if and only if S, e.g. ‘Snow is white’ is true if and only if snow is white, ‘There are infinitely many prime numbers’ is true if and only if there are infinitely many prime numbers, ‘Australia is a continent’ is true if and only if Australia is a continent, we will have succeeded in stating for each sentence of the English language a condition for its membership in the set of true sentences which we understand as well as we understand the sentence itself. Moreover – this is the technical part of Tarski’s theory – Tarksi’s work shows how, after we have formalised English, this infinite set of, so to speak, axioms for the property of truth in English can be captured by a finite definition in an appropriate ‘metalanguage’.
To this there is an objection – or at least a question – which I would put thus: when you say that I understand ‘ “Snow is white” is true’ just as well as I understand ‘Snow is white,’ what do you mean by understand? If Quine were to answer, ‘To understand a sentence is to know the conditions under which it is true,’ obviously no progress would have been made. But Quine’s account of understanding does not use the notion of truth (which is how circularity is avoided). To understand an observation sentence is to be conditioned so that appropriate sensory stimulations will prompt one’s assent to the sentence. And to understand a non-observation sentence is to master its role in the system. On this picture, cognitively meaningful speaking or thinking is simply an attempt to get someone – myself or someone else – to change their dispositions-to-predict. Calling a sentence someone (myself or someone else) utters (or thinks) ‘true’ is just an indication that I would currently include that sentence in the system I use to predict.
Indeterminacy of Translation
Quine’s most famous doctrine is, of course, the doctrine of the indeterminacy of translation. The doctrine is widely misunderstood; the articles in this Dictionary hint at the doctrine, but they do not explicitly state or defend it. In Word and Object Quine explained the doctrine with the aid of a thought experiment which has been endlessly quoted and discussed. Imagining a ‘jungle language’ in which the natives say (or at least ‘assent to’) gavagai whenever they see a rabbit, Quine argued that one could translate the native utterance either as ‘Rabbit!’ or as ‘Rabbithood!’
A point which many critics missed is that Quine was arguing at least as much for the determinacy of translation in the case of observation sentences as for indeterminacy. True, Quine wants us to see that the use of a sentence need not fix its exact translation into another language, or even determine what objects the sentence is about. But he also wants us to see that there is an important sense in which (considered just as observation sentences, apart from their contribution to inferences) the English sentences ‘That is a rabbit’ and ‘That is rabbithood exemplifying itself’ have the same meaning. They correspond to exactly the same perceptual stimuli. ‘Taken holophrastically’ (i.e. as unanalysed wholes), why should ‘That is a rabbit’ and ‘That is rabbithood exemplifying itself’ be thought of as differing in meaning at all? We do know what gavagai means, according to Quine: it means ‘Rabbit!’ and it means ‘Rabbithood!’ Quine does not discuss these technical issues in the present work, but such a belief in the determinacy of meaning of the observation sentence is what is behind his insistence (in the essay on Communication) that statements about ‘tangible, visible and audible reality’ are ‘unfailing vehicles of communication’.
When it comes to exact word-for-word translation – the sort of translation we need to determine whether a sentence commits the speaker to to belief in the existence of universals (‘rabbithood’) or not – and when it comes to non-observational sentences, things do become much more indeterminate on Quine’s view, though still not completely so. Here Quine is drawing a conclusion deeply disturbing to positivists from positivist premises. If we construct the world by making up a system of posits that helps us to predict our sensations, then why should we suppose that the sentences which comprise that system of posits are associated with such suspiciously non-empirical entities as ‘meanings’ or ‘ideas’? (Quine ends the essay on Ideas by writing: ‘The idea idiom is an entrenched and consequently useful element of our vernacular. In daily discourse we cannot easily do without it, nor need we try. But it is a snare to the philosopher or scientist who admits it to his theory. There is no place in science for ideas.’) To say that a sentence in one language has the same ‘meaning’ as a sentence in another is to say their roles are similar: but we should not expect similarity of role to be a precise or well-defined notion. Sameness of meaning is a useful notion in everyday life, but it has no role to play in our ‘first-class’ scientific theory of the world. Or as Quine writes, ending his essay on Communication: ‘We get an exaggerated idea of how well we have been understood, simply for want of checkpoints to the contrary. The miracle of communication, in its outer reaches, is a little like the miracle of transubstantiation: what transubstantiation?’
Here is a little bit of evidence in support of Quine’s claim. I recall that when I visited China in 1984 I lectured on Quine’s views at Fudan University in Shanghai, and sophisticated Chinese told me that they did not think that the Chinese word mao (‘cat’) could be determinately translated into English as ‘cat’/‘cathood’. What they claimed was that ‘Are you saying there is a cat or that there is cathood exemplifying itself?’ is the wrong question to put to a Chinese speaker. There is no special suffix in Chinese to distinguish mao from ‘mao-hood’ (mao is used both in contexts in which we would translate it as ‘cat’ and in contexts in which we would translate it as ‘cathood’), nor are there articles in Chinese. ‘Cat there’ and ‘Cathood there’ would go into the same sentence in Chinese.
Of course, one could say that the Chinese word mao is ambiguous. But is the fact that one language has two words where another language has one really a proof of the existence of an ambiguity in the one word? Quine is arguing that such questions are bad questions. The Chinese conceptual scheme works as well as ours at producing those ‘observation categoricals’, but the sentences are not exactly isomorphic to our sentences: the consequence of the failure of isomorphism is that our ‘parochial’ ontology of particulars and universals can be projected onto the Chinese utterances in more than one way. There is no ‘fact of the matter’ as to whether the Chinese speaker is really talking about the cats or the cathood.
This conclusion was unwelcome to Quine’s fellow positivists (and to analytic philosophers generally) because analytic philosophy was conceived of as ‘analysis of meaning’, and Quine is telling us that the notions of ‘meaning’ and ‘analysis of meaning’ are hopelessly vague. If analytic philosophy has become more of a style than a school since Quine published Word and Object, that is in large part the consequence of the corrosive effect of Quine’s views on the whole notion of ‘analysis’.
The role of the positivist premises in Quine’s argument is most easily seen by contrasting Quine’s view with a more standard ‘realist’ view of language. A realist who sees objects as just ‘there’ independently of language naturally expects that different languages will have ‘names’ for at least the more salient features of the external world, and expects there to be a ‘fact of the matter’ as to which words in a language ‘correspond’ to which mind-independent objects. Someone like Quine, who sees objects as constructions within language, does not expect the ontologies of different languages to line up exactly. A philosopher of this second kind – call him a ‘linguistic idealist’ – may either regard the ontologies of different discourses as incommensurable (the line of thinking of a Kuhn or a Saussure or a Derrida) or, like Quine, regard the attribution of an ontology to a discourse as a useful but sloppy bit of projection. An excellent new study of these issues, and of Quine’s philosophy in general, is Christopher Hootway’s Quine.
Quine and Realism
That Quine’s perspectivalism about meaning flows from his positivist picture of knowledge and language does not mean that his arguments can safely be ignored by philosophers of a more realist persuasion. Quine’s writing has sensitised philosophers to problems which had been long ignored, and lost innocence is impossible to recapture. Realists tend to assume that it is possible to draw a sharp line between ‘the facts’ and ‘what we project onto the facts’ (or between ‘the facts’ and ‘our conventions’). What Quine’s work suggests is that no line can be drawn. We cannot translate the Chinese word mao just any which way. (To translate mao as ‘rabbit’ would be simply wrong.) But it does not follow that there is a fact of the matter as to whether mao ‘corresponds’ to rabbits or to rabbithood. (From within Chinese it is easy to say what mao corresponds to, of course: ‘mao’ corresponds to mao.) The notion of ‘correspondence’ on which traditional realism leans has proved to be a weak reed.
Some facts are, if not totally independent of mind and language, at least independent of any particular conceptual scheme on Quine’s view. These are the facts expressed by observation sentences (taken holophrastically). They are not so much ‘pre-analytic’ facts as ‘dis-analysed’ facts; when we take an observation sentence holophrastically what we are doing is subtracting whatever in it is peculiar to the particular language in which it is expressed to arrive at a fact which is to the maximum degree independent of conceptual scheme. That criterion of what is ‘absolute’ – namely, what is to the maximum degree independent of conceptual scheme – has been thought by Bernard Williams, in his recent Ethics and the Limits of Philosophy, to delimit science, and to distinguish science from ethics. In Quine’s view, Williams is wrong. It is not total science that is ‘to the maximum degree independent of conceptual scheme’, but only the observation sentences and observation categoricals.
Quine’s views have provoked an enormous variety of responses. My own work has been deeply influenced by Quine’s rejection of a sharp convention/fact distinction and by his insistence that the notion of ‘sameness of meaning’ that we actually possess is constituted by our actual practice of translation and interpretation; meanings are not to be seen as platonic objects which somehow explain translation and interpretation. But that does not mean that I can accept the whole of Quine’s view, although I enormously admire Quine’s willingness to push his view to the limits, and to bring out and emphasise controversial or paradoxical consequences of his view rather than try to sweep them under the rug.
Here I should like to mention three points at which I am in disagreement with Quine.
1. The Privileged Status of Observation Sentences. It seems to me that Quine’s increasing insistence on the determinacy of the meaning of observation sentences (taken holophrastically) represents a backing-away from his own insight that it is impossible to draw a sharp fact/convention distinction. Quine put this insight beautifully some years ago when he wrote: ‘The lore of our fathers is a fabric of sentences ... It is a pale grey lore, black with fact and white with convention. But I have found no substantial reasons for concluding that there are any quite black threads in it, nor any white ones.’Although Quine would not agree that this is what he is doing, it is as if he were now telling us what the real facts are (telling us that they are expressed by the true observation sentences, taken holophrastically, and everything else is the product of convention).
In fact, Quine’s claim that observation sentences are ‘unfailing vehicles of communication’ is not right. Imagine that we come across a jungle language (call it Natool) and succeed in working out a successful ‘translation scheme’. Suppose the sentence Hu bosarka turns out to have the translation ‘She is a witch.’ Suppose further that only women with moles on their noses are called witches. Then Hu bosarka might well turn out to be conditioned to the same stimuli as our English sentence ‘She has a wart on her nose’: but it would not have the same meaning, not even ‘taken holophrastically’. Quine is right to suppose that sometimes two sentences which consist of totally non-equivalent parts can make the same claim taken as wholes: but wrong to suppose that whenever two sentences are conditioned to the same stimuli they can be regarded as in some (‘holophrastic’) sense equivalent in meaning. Quine’s retrograde motion here testifies to the enomous strength of the urge to find some level of foundational ‘facts’.
2. Quine’s View of Truth and Understanding. The combination of a Tarskian theory of truth and a positivist theory of understanding leads, as I said, to the picture of cognitively meaningful speaking and thinking as an attempt to get someone – myself or someone else – to change their dispositions to predict, or better, to change the complex system which they employ to generte those dispositions to predict (or their linguistic counterparts, the ‘observation categoricals’). The distance of such a view from our common-sense picture is enormous. Consider a sentence about the past: on the common-sense view, when I say, ‘Caesar crossed the Rubicon’, I am making a true assertion and calling the sentence true is ascribing a genuine property to the sentence, not merely making a noise which is equivalent to the original noise according to the rules of some system. On Quine’s view, I gets the feeling, uttering a sentence about the past is making a move in a game whose real point is to predict the future. I share Quine’s dissatisfaction with traditional ‘realist’ philosophy: but the right alternative cannot be such a radical positivism as this.
3. Quine’s Emphasis on Prediction. His devotion to empiricism comes out most strongly in his emphasis on successful prediction as the sole ultimate ‘evidence’ for anything. On Quine’s view, mathematical statements, for example, are only justified in so far as they help to make successful predictions in physics, engineering etc. I find this claim almost totally unsupported by actual mathematical practice. As I have argued elsewhere, we have many more cognitive interests than prediction and, correspondingly, many more kinds of justification than are included in this narrow notion of ‘evidence’.
Discussion of these issues will certainly continue: in the meantime, anyone who wants to encounter a great philosophical mind in a less technical mood, and to get some feeling for Quine as the peerless companion, raconteur, and amused commentator on the passing show, that his friends know him to be cannot do better than read this book. I began by quoting what Strawson says of the author on the jacket of the book; I cannot end with a more fitting tribute than the one Colin McGinn paid to Quine in the Journal of Philosophy (also quoted on the jacket of the book): ‘Quine pursues philosophical vision with an uncompromising consistency of purpose that makes his doctrines impossible to ignore. You either go with him or define your position in reaction to his. And this is one mark of a great philosopher.’