FROM A LOGICAL POINT OF VIEW
9 Logico-Philosophical Essays

    

Willard Van Ormond Quine

    

Harvard University Press
Cambridge Massachusetts 1953

    

COPYRIGHT 1953
by the President and Fellows of Cambridge University

 

Printed in United States of America

    

    

CONTENTS
The word 'page', given at the top of each page, is a link to a picture of the actual page.
CHAPTER			PAGE
		Preface	i-viii
I.		On what there is	1
II.		Two dogmas of empiricism	20
III.		The problem of meaning in linguistics	47
Photos of original pages only from Ch.IV to end.
IV.		Identity, ostension, and hypostasis	65
V.		New foundations for mathematical logic	80
VI.		Logic and the reification of universals	102
VII.		Notes on the theory of reference	130
VIII.		Reference and modality	139
IX.		Meaning and existential inference	160
		Origin of the essays	169
		Bibliographical references	171
		Index	179

    

page v

PREFACE

    Several of these essays have been printed whole in journals; others are in varying degrees new. Two main themes run through them. One is the problem of meaning, particularly as involved in the notion of an analytic statement. The other is the notion of ontological commitment, particularly as involved in the problem of universals.

    Various previously published papers which seemed to call for inclusion presented twofold problems, For one thing, they overlapped as Ipapers will which are so written as to spare readers excessive use of libraries. For another, they contained parts which I had grown to recognize as badly formulated or worse. The upshot was that several essays seemed to warrant fairly integral reproduction under their original titles, while others had to be chopped, culled, mixed, eked out with new material, and redivided according to new principles of unification and individuation which brought new titles in their train. For the provenience of what is not new see Origins of the Essays, in the back pages.

    The pair of themes named at the top of this page is pursued through the book with the aid, increasingly, of the technical devices of logic. Hence there comes a point, midway, when those themes, must be interrupted for the purpose of some elementary technical preparation in logic. “New foundations” is reprinted both for this purpose and for its own sake; for it has figured in subsequent literature, and offprints continue to be sought. Its reproduction here creates an occasion also for supplementary remarks, touching on those subsequent findings and relating the

    

page vi

system of “New foundations” to other set theories. However, this intrusion of pure logic has been kept resolutely within bounds.

    As noted in some detail in the back pages, the content of this volume is in large part reprinted or adapted from the Review of Metaphysics, the Philosophical Review, the Journal of Philosophy, the American Mathematical Monthly, the Journul of Symbolsic Logic, the Proceedings of the American Academy of Arts and Sciences, and Philosophical Studies. I am grateful to the editors of these seven periodicals and to the University of Minnesota Press for their kind permission to make this further use of the material.

    I am obliged to Professors Rudolf Carnap and Donald Davidson for helpful criticisms of early drafts of “New foundations” and “Two dogmas” respectively, and to Professor Paul Bernays for noting an error in the first printing of “New foundations.” The critique of analyticity to which “Two dogmas” is in large part devoted is an outcome of informal discussions, oral and written, in which I have engaged from 1939 onward with Professors Carnap, Alonao Church, Nelson Goodman, Alfred Tarski, and Morton White; to them I am indebted certainly for stimulation of the essay, and probably for content. To Goodman I am indebted also for criticism of two of the papers from which “Logic and the reification of universals” was in part drawn; and to White for discussion which influenced the present form of that essay.

    I thank Mrs. Martin Juhn for her good typing, and the administrators of the Harvard Foundation for a grant in aid. I am grateful to Messrs. Donald P. Quimby and S. Marshall Cohen for able assistance with the index and proofs.

W. V. QUINE    

        Cambridge Massachusetts

    

    

page 1     ON WHAT THERE IS     I

    

    

I

ON WHAT THERE IS

 

    A curious thing about the ontological problem is its simplicity. It can be put in three Anglo-Saxon monosyllables: ‘What is there?’ It can be answered, moreover, in a word—‘Everything’—and everyone will accept this answer as true. However, this is merely to say that there is what there is. There remains room for disagreement over cases; and so the issue has stayed alive down the centuries.

Suppose now that two philosophers, McX and I, differ over ontology. Suppose McX maintains there is something which I maintain there is not. McX can, quite consistently with his own point of view, describe our difference of opinion by saying that I refuse to recognize certain entities. I should protest, of course, that he is wrong in his formulation of our disagreement, for I maintain that there are no entities, of the kind which he alleges, for me to recognize; but my finding him wrong in his formulation of our disagreelment is unimportant, for I am committed to considering him wrong in his ontology anyway.

When I try to formulate our difference of opinion, on the other hand, I seem to be in a predicament. I cannot admit that there are some things which McX countenances and I do not, for in admitting that there are such things I should be contradicting my own rejection of them.

It would appear, if this reasoning were sound, that in any ontological dispute the proponent of the negative side suffers the disadvantage of not beiing able to admit that his opponent disagrees with him. This is the old Platonic riddle of nonbeing. Nonbeing must

    

page 2     ON WHAT THERE IS     I

in some sense be, otherwise what is it that there is not? This tangled doctrine might be nicknamed Plato’s beard; historically it has proved tough, frequently dulling the edge of Occam’s razor.

    It is some such line of thought that leads philosophers like McX to impute being where they might otherwise be quite content to recognize that there is nothing. Thus, take Pegasus. If Pegasus were not, McX argues, we should not be talking about anything when we use the word; therefore it would be nonsense to say even that Pegasus is not. Thinking to show thus that the denial of Pegasus cannot be coherently maintained, he concludes that Pegasus is.

    McX cannot, indeed, quite persuade himself that any region of space-time, near or remote, contains a flying horse of flesh and blood. Pressed for further details on Pegasus, then, he says that Pegasus is an idea in men’s minds. Here, however, a confusion begins to be apparent. We may for the sake of argument concede that there is an entity, and even a unique entity (though this is rather implausible), which is the mental Pegasus-idea; but this mental entity is not what people are talking about when they deny Pegasus.

    McX never confuses the Parthenon with the Parthenon-idea. The Parthenon is physical; the Parthenon-idea is mental (according anyway to McX’s version of ideas, and I have no better to offer). The Parthenon is visible; the Parthenon-idea is invisible. We cannot easily imagine two things more unlike, and less liable to confusion, than the Parthenon and the Parthenon-idea. But when we shift from the Parthenon to Pegasus, the confusion sets in—for no other reason than that McX would sooner be deceived by the crudest and most flagrant counterfeit than grant the nonbeing of Pegasus.

    The notion that Pegasus must be, because it would otherwise be nonsense to say even that Pegasus is not, has been seen to lead McX into an elementary confusion. Subtler minds, taking the same precept as their starting point, come out with theories of Pegasus which are less patently misguided than McX’s, and correspondingly more difficult to eradicate. One of these subtler

    

page 3     ON WHAT THERE IS     I

minds is named, let us say, Wyman. Pegasus, Wyman maintains, has his being as an unactualized possible. When we say of Pegasus that there is no such thing, we are saying, more precisely, that Pegasus does not have the special attribute of actuality. Saying that Pegasus is not actual is on a par, logically, with saying that the Parthenon is not red; in either case we are saying something about an entity whose being is unquestioned.

    Wyman, by the way, is one of those philosophers who have united in ruining the good old word ‘exist’. Despite his espousal of unactualized possibles, he limits the word ‘existence’ to actuality—thus preserving an illusion of ontological agreement between himself and us who repudiate the rest of his bloated universe. We have all been prone to say, in our common-sense usage of ‘exist’, that Pegasus does not exist, meaning simply that there is no such entity at all. If Pegasus existed he would indeed be in space and time, but only because the word ‘Pegasus’ has spatio-temporal connotations, and not because ‘exists’ has spatio-temporal connotatians. If spatio-temporal reference is lacking when we affirm the existence of the cube root of 27, this is simply because a cube root is not a spatio-temporal kind of thing, and not because we are being ambiguous in our use of ‘exist’.¹ However, Wyman, in an ill-conceived effort to appear agreeable, genially grants us the nonexistence of Pegasus and then, contrary to what we meant by nonexistence of Pegasus, insists that Pegasus is. Existence is one thing, he says, and subsistence is another. The only way I know of coping with this obfuscation of issues is to give Wyman the word ‘exist’. I’ll try not to use it again; I still have ‘is’. So much for lexicography; let’s get back to Wyman’s ontology.

    ¹ The impulse to distinguish terminologically between existence as applied to objects actualized somewhere in space-time and existence (or subsistence or being) as applied to other entities arises in part, perhaps, from an idea that the observation of nature is relevant only to questions of existence of the first kind. But this idea is readily refuted by counterinstances such as ‘the ratio of the number of centaurs to the number of unicorns’. If there were such a ratio, it would be an abstract entity, viz. a number. Yet it is only by studying nature that we conclude that the number of centaurs and the number of unicorns are both 0 and hence that there is no such ratio.

    

page 4     ON WHAT THERE IS       I

    Wyman’s overpopulated universe is in many ways unlovely. It offends the aesthetic sense of us who have a taste for desert landscapes, but this is not the worst of it. Wyman’s slum of possibles is a breeding ground for disorderly elements. Take, for instance, the possible fat man in that doorway; and, again, the possible bald man in that doorway. Are they the same possible man, or two possible men? How do we decide? How many possible men are there in that doorway? Are there more possible thin ones than fat ones? How many of them are alike? Or would their being alike make them one? Are no two possible things alike? Is this the same as saying that it is impossible for two things to be alike? Or, finally, is the concept of identity simply inapplicable to unactualized possibles? But what sense can be found in talking of entities which cannot meaningfully be said to be identical with themselves and distinct from one another? These elements are well-nigh incorrigible. By a Fregean therapy of individual concepts,² some effort might be made at rehabilitation; but I feel we’d do better simply to clear Wyman's slum and be done with it.

    Possibility, along with the other modalities of necessity and impossibility and contingency, raises problems upon which I do not mean to imply that we should turn our backs. But we can at least limit modalities to whole statements. We may impose the adverb ‘possibly’ upon a statement as a whole, and we may well worry about the semantical analysis of such usage; but little real advance in such analysis is to be hoped for in expanding our universe to include so-called possible entities. I suspect that the main motive for this expansion is simply the old notion that Pegasus, for example, must be because otherwise it would be nonsense to say even that he is not.

    Still, all the rank luxuriance of Wyman’s universe of possibles would seem to come to naught when we make a slight change in the example and speak not of Pegasus but of the round square cupola on Berkeley College. If, unless Pegasus were, it would be nonsense to say that he is not, then by the same token, unless the round square cupola on Berkeley College were, it

    ² see below, p. 153.

    

page 5     ON WHAT THERE IS     I

would be nonsense to say that it is not. But, unlike Pegasus, the round square cupola on Berkeley College cannot be admitted even as an unactualized possible. Can we drive Wyman now to admitting also a realm of unactualizable impossibles? If so, a good many embarrassing questions could be asked about them. We might hope even to trap Wyman in contradictions, by getting him to admit that certain of these entities are at once round and sqnare. But the wily Wyman chooses the other horn of the dilemma and concedes that it is nonsense to say that the round square cupola on Berkeley College is not. He says that the phrase ‘round square cupola’ is meaningless.

    Wyman was not the first to embrace this alternative. The doctrine of the meaninglessness of contradictions runs away back. The tradition survives, moreover, in writers who seem to share none of Wyman’s motivations. Still, I wonder whether the first temptation to such a doctrine may not have been substantially the motivation which we have observed in Wyman. Certainly the doctrine has no intrinsic appeal; and it has led its devotees to such quixotic extremes as that of challenging the method of proof by reductio ad absurdum—a challenge in which I sense a reductio ad absurdum of the doctrine itself.

    Moreover, the doctrine of meaninglessness of contradictions has the severe methodological drawback that it makes it impossible, in principle, ever to devise an effective test of what is meaningful and what is not. It would be forever impossible for us to devise systematic ways of deciding whether a string of signs made sense—even to us individually, let alone other people—or not. For it follows from a discovery in mathematical logic, due to Church [2], that, there can be no generally applicable test of contradictoriness.

    I have spoken disparagingly of Plato’s beard, and hinted that it is tangled. I have dwelt at length on the inconveniences of putting up with it. It is time to think about taking steps.

    Russell, in his theory of so-called singular descriptions, showed clearly how we might meaningfully use seeming names without supposing that there be the entities allegedly named. The names to which Russell’s theory directly applies are complex

    

page 6     ON WHAT THERE IS     I

descriptive names such as ‘the author of Waverley’, ‘the present King of France’, ‘the round square cupola on Berkeley College’. Russell analyzes such phrases systematically as fragments of the whole sentences in which they occur. The sentence 'The author of Waverley was a poet’, for example, is explained as a whole as meaning ‘Someone (better: something) wrote Waverley and was a poet, and nothing else wrote Waverley’. (The point of this added clause is to affirm the uniqueness which is implicit in the word ‘the’, in ‘the author of Waverley’.) The sentence ‘The round square cupola on Berkeley College is pink’ is explained as ‘Something is: round and square and is a cupola on Berkeley College and is pink, and nothing else is round and square and a cupola on Berkeley College’.³

    The virtue of this analysis is that the seeming name, a descriptive phrase, is paraphrased in context as a so-called incomplete symbol. No unified expression is offered as an analysis of the descriptive phrase, but the statement as a whole which was the context of that phrase still gets its full quota of meaning—whether true or false.

    The unanalyzed statement ‘The author of Waverley was a poet’ contains a part, ‘the author of Waverley’, which is wrongly supposed by McX and Wyman to demand objective reference in order to be meaningful at all. But in Russell’s translation, ‘Something wrote Waverley and was a poet and nothing else wrote Waverley’, the burden of objective reference which had been put upon the descriptive phrase is now taken over by words of the kind that logicians call bound variables, variables of quantification, namely, words like ‘something’, ‘nothing’, ‘everything’. These words, far from purporting to be names specifically of the author of Waverley, do not purport to be names at all; they refer to entities generally, with a kind of studied ambiguity peculiar to themselves.⁴ These quantificational words or bound variables are, of course a basic part of language, and their meaningfulness, at least in context, is not

    ³ For more on .the theory of descriptions see below, pp. 85f, 166f.
    ⁴ For more explicit treatment of the bound variable see below, pp. 82, 102f.

    

page 7     ON WHAT THERE IS     I

to be challenged. But their meaningfulness in no way praupposes there being either the author of Wave&y or the round square cupola on Berkeley College or any other specifically preassigned objects.

    Where descriptions are concerned, there is no longer any difficulty in affirming or denying being. ‘There is the author of Waverley’ is explained by Russell as meaning ‘Someone (or, more strictly, something) wrote Waverley and nothing else wrote Waverley’. ‘The author of Waverley is not’ is explained, correspondingly, as the alternation ‘Either each thing failed to write Waverley or two or more things wrote Waverley’. This alternation is false, but meaningful; and it contains no expression purporting to name the author of Waverley. The statement ‘The round square cupola on Berkeley College is not’ is analyzed in similar fashion. So the old notion that statements of nonbeing defeat themselves goes by the board. When a statement of being or nonbeing is analyzed by Russell’s theory of descriptions, it ceases to contain any expression which even purports to name the alleged entity whose being is in question, so that the meaningfulness of the statement no longer can be thought to presuppose that there be such an entity.

    Now what of ‘Pegasus’? This being a word rather than a descriptive phrase, Russell’s argument does not immediately apply to it. However, it can easily be made to apply. We have only to rephrase ‘Pegasus’ as a description, in any way that seems adequately to single out our idea; say, ‘the winged horse that was captured by Bellerophon’. Substituting such a phrase for ‘Pegasus’, we can then proceed to analyze the statement ‘Pegasus is’, or ‘Pegasus is not’, precisely on the analogy of Russell’s analysis of ‘The author of Waverley is’ and ‘The author of Waverley is not’.

    In order thus to subsume a one-word name or alleged name such as ‘Pegasus’ under Russell’s theory of description, we must, of course, be able first to translate the word into a description. But this is no real restriction. If the notion of Pegasus had been so obscure or so basic a one that no pat translation into a descriptive phrase had offered itself along familiar lines, we

    

page 8     ON WHAT THERE IS     I

could still have availed ourselves of the following artificial and trivial-seeming device: we could have appealed to the ex hypothesi unanalyzable, irreducible attribute of being Pegasus, adopting, for its expression, the verb ‘is-Pegasus’, or ‘pegasizes’. The noun ‘Pegasus’ itself could then be treated as derivative, and identified after all with a description: ‘the thing that is-Pegasus’, ‘the thing that pegasizes’.⁵

     If the importing of such a predicate as ‘pegasizes’ seems to commit us to recognizing that there is a corresponding attribute, pegasizing, in Plato’s heaven or in the minds of men, well and good. Neither we nor Wyman nor McX have been contending, thus far, about the being or nonbeing of universals, but rather about that of Pegasus. If in terms of pegs&zing we can interpret the noun ‘Pegasus’ as a description subject to Russell’s theory of descriptions, then we have disposed of the old notion that Pegasus cannot, be said not to be without presupposing that in some sense Pegasus is.

    Our argument is now quite general. McX and Wyman supposed that we could not meaningfully affirm a statement of the form ‘So-and-so is not’, with a simple or descriptive singular noun in place of ‘so-and-so’, unless so-and-so is. This supposition is now seen to be quite generally groundless, since the singular noun in question can always be expanded into a singular description, trivially or otherwise, and then analyzed out à la Russell.

    We commit ourselves to an ontology containing numbers when we say there are prime numbers larger than a million; we commit ourselves to an ontology containing centaurs when we say there are centaurs; and we commit ourselves to an ontology containing Pegasus when we say Pegasus is. But we do not commit ourselves to an ontology containing Pegasus or the author of Waverley or the round square cupola on Berkeley College when we say that Pegasus or the author of Waverley or the cupola in question is not. We need no longer labor under the delusion that the meaningfulness of a statement containing

     ⁵ For further remarks on such assimilation of all singular terms to descriptiona see below, p. 167; also Quine [2], pp. 218-224.

    

page 9     ON WHAT THERE IS     I

     a singular term presupposes an entity named by the term. A singular term need not name to be significant.

    An inkling of this might have dawned on Wyman and McX even without benefit of Russell if they had only noticed—a so few of us do—that there is a gulf between meaning and naming even in the case of a singular term which is genuinely a name of an object. The following example from Frege [3] will serve. The phrase ‘Evening Star’ names a certain large physical object of spherical form, which is hurtling through space some scores of millions of miles from here. The phrase ‘Morning Star’ names the same thing, as was probably first established by some observant Babylonian. But the two phrases cannot be regarded as having the same meaning; otherwise that Babylonian could have dispensed with his observations and contented himself with reflecting on the meanings of his words. The meanings, then, being different from one another, must be other than the named object, which is one and the same in both cases.

    Confusion of meaning with naming not only made McX think he could not meaningfully repudiate Pegasus; a continuing confusion of meaning with naming no doubt helped engender his absurd notion that Pegasus is an idea, a mental entity. The structure of his confusion is as follows. He confused the alleged named object Pegasus with the meaning of the word ‘Pegasus’, therefore concluding that Pegasus must be in order that the word have meaning. But what sorts of things are meanings? This is a moot point; however, one might quite plausibly explain meanings as ideas in the mind, supposing we can make clear sense in turn of the idea of ideas in the mind. Therefore Pegasus, initially confused with a meaning, ends up as an idea in the mind. It is the more remarkable that Wyman, subject to the same initial motivation as McX, should have avoided this particular blunder and wound up with unactualized possibles instead.

    Now let us turn to the ontological problem of universals: the question whether there are such entities as attributes, relations, classes, numbers, functions. McX, characteristically enough, thinks there are. Speaking of attributes, he says : “There

    

page 10     ON WHAT THERE IS     I

are red houses, red roses, red sunsets; this much is prephilosophical common sense in which we must all agree. These houses, roses, and sunsets, then, have something in common; and this which they have in common is all I mean by the attribute of redness.” For McX, thus, there being attributes is even more obvious and trivial than the obvious and trivial fact of there being red houses, roses, and sunsets. This, I think, is characteristic of metaphysics, or at least of that part of metaphysics called ontology: one who regards a statement on this subject as true at all must regard it as trivially true. One’s ontology is basic to the conceptual scheme by which he interprets all experiences, even the most commonplace ones. Judged within some particular conceptual scheme—and how else is judgement possible?—an ontological statement goes without saying, standing in need of no separate justification at all. Ontological statements follow immediately from all manner of casual statements of commonplace fact, just as—from the point of view, anyway, of McX’s conceptual scheme—‘There is an attribute’ follows from ‘There are red houses, red roses, red sunsets’.

Judged in another conceptual scheme, an ontological statement which is axiomatic to McX’s mind may, with equal immediacy and triviality, be adjudged false. One may admit that there are red houses, roses, and sunsets, but deny, except as a popular and misleading manner of speaking, that they have anything in common. The words ‘houses’, ‘roses’, and ‘sunsets’ are true of sundry individual entities which are houses and roses and sunsets, and the word ‘red’ or ‘red object’ is true of each of sundry individual entities which are red houses, red roses, red sunsets; but there is not, in addition, any entity whatever, individual or otherwise, which is named by the word ‘redness’, nor, for that matter, by the word ‘househood’, ‘rosehood’, ‘sunsethood’. That the houses and roses and sunsets are all of them red may be taken as ultimate and irreducible, and it may be held that McX is no better off, in point of real explanatory power, for all the occult entities which he posits under such names as ‘redness’.

    One means by which McX might naturally have tried to

    

page 11     ON WHAT THERE IS     I

impose his ontology of universals on us was already removed before we turned to the problem of universals. McX cannot argue that predicates such as ‘red’ or ‘is-red’, which we all concur in using, must be regarded as names each of a single universal entity in order that they be meaningful at all. For we have seen that being a name of something is a much more special feature than being meaningful. He cannot even charge us—at least not by that argument—with having posited an attribute of pegasizing by our adoption of the predicate ‘pegasizes’.

    However, McX hits upon a different strategem. “Let us grant,” he says, “this distinction between meaning and naming of which you make so much. Let us even grant that ‘is red’, ‘pegasizes’, etc., are not names of attributes. Still, you admit they have meanings. But these meanings, whether they are named or not, are still universals, and I venture to say that some of them might even be the very things that I call attributes, or something to much the same purpose in the end.”

    For McX, this is an unusually penetrating speech; and the only way I know to counter it is by refusing to admit meanings. However, I feel no reluctance toward refusing to admit meanings, for I do not thereby deny that words and statements are meaningful. McX and I may agree to the letter in our classification of linguistic forms into the meaningful and the meaningless, even though McX construes meaningfulness as the having (in some sense of ‘having’) of some abstract entity which he calls a meaning, whereas I do not. I remain free to maintain that the fact that a given linguistic utterance is meaningful (or signijcant, as I prefer to say so as not to invite hypostasis of meanings as entities) is an ultimate and irreducible matter of fact; or, I may undertake to analyze it in terms directly of what people do in the presence of the linguistic utterance in question and other utterances similar to it.

    The useful ways in which people ordinarily talk or seem to talk about meanings boil down to two: the having of meanings, which is significance, and sameness of meaning, or synonomy. What is called giving the meaning of an utterance is simply the uttering of a synonym, couched, ordinarily, in clearer language

    

page 12     ON WHAT THERE IS     I

than the original. If we are allergic to meanings as such, we can speak directly of utterances as significant or insignificant, and as synonymous or heteronymous one with another. The problem of explaining these adjectives ‘significant’ and ‘synonymous’ with some degree of clarity and rigor—preferably, as I see it, in terms of behavior—is as diflicult as it is important.⁶ But the explanatory value of special and irreducible intermediary entities called meanings is surely illusory.

    Up to now I have argued that we can use singular terms significantly in sentences without presupposing that there are the entities which those terms purport to name. I have argued further that we can use general terms, for example, predicates, without conceding them to be names of abstract entities. I have argued further that we can view utterances as significant, and as synonymous or heteronymous with one another, without countenancing a realm of entities called meanings. At this point McX begins to wonder whether there is any limit at all to our ontological immunity. Does nothing we may say commit us to the assumption of universals or other entities which we may find unwelcome?

    I have already suggested a negative answer to this question, in speaking of bound variables, or variables of quantification, in connection with Russell’s theory of descriptions. We can very easily involve ourselves in ontological commitments by saying, for example, that there is something (bound variable) which red houses and sunsets have in common; or that there is something which is a prime number larger than a million. But, this is, essentially, the only way we can involve ourselves in ontological commitments: by our use of bound variables. The use of alleged names is no criterion, for we can repudiate their namehood at the drop of a hat unless the assumption of a corresponding entity can be spotted in the things we affirm in terms of bound variables. Names are, in fact, altogether immaterial to the ontological issue, for I have shown, in connection with ‘Pegasus’ and ‘pegasize’, that names can be converted to descriptions, and Russell has shown that descriptions can be eliminated.

    ⁶See EsaysII and III.S

    

page 13     ON WHAT THERE IS     I

Whatever we say with the help of names can be said in a language which shuns names altogether. To be assumed as an entity is, purely and simply, to be reckoned as the value of a variable. In terms of the categories of traditional grammar, this amounts roughly to saying that to be is to be in the range of reference of a pronoun. Pronouns are the basic media of reference; nouns might better have been named propronouns. The variables of quantification, ‘something’, ‘nothing’, ‘everything’, range over our whole ontology, whatever it may be; and we are convicted of a particular ontological presupposition if, and only if, the alleged presuppositum has to be reckoned among the entities over which our variables range in order to render one of our affirmations true.

    We may say, for example, that some dogs are white and not thereby commit ourselves to recognizing either doghood or whiteness as entities. ‘Some dogs are white’ says that some things that are dogs are white; and, in order that this statement be true, the things over which the bound variable ‘something’ ranges must include some white dogs, but need not include doghood or whiteness. On the other hand, when we say that some zoological species are cross-fertile we are committing ourselves to recognizing as entities the several species themselves, abstract though they are. We remain so committed at least until we devise some way of so paraphrasing the statement as to show that the seeming referlence to species on the part of our bound variable was an avoidable manner of speaking.⁷

    Classical mathematics, as the example of primes larger than a million clearly illustrates, is up to its neck in commitments to an ontology of abstract entities. Thus it is that the great mediaeval controversy over universals has flared up anew in the modern philosophy of mathematics. The issue is clearer now than of old, because we now have a more explicit standard whereby to decide what ontology a given theory or form of discourse is committed to: a theory is committed to those and only those entities to which the bound variables of the theory

    ⁷For more on this topic see Essay VI.

        

page 14     ON WHAT THERE IS     I

must be capable of referring in order that the affirmations made in the theory be true.

    Because this standard of ontological presupposition did n.ot emerge clearly in the philosophical tradition, the modern phillosophical mathematicians have not on the whole recognized that they were debating the same old problem of universals in a newly clarified form. But the fundamental cleavages among modern points of view on foundations of mathematics do come down pretty explicitly to disagreements as to the range of entities to which the bound variables should be permitted to refer.

    The three main mediaeval points of view regarding universals are designated by historians as realism, conceptualism, and nominalism. Essentially these same three doctrines reappear in twentieth-century surveys of the philosophy of mathelmatics under the new names logicism, intuitionism, and formalism.

    Realism, as the word is used in connection with the mediaeval controversy over universals, is the Platonic doctrine that universals or abstract entities have being independently of the mind; the mind may discover them but cannot create them. Logicism, represented by Frege, Russell, Whitehead, Church, and Carnap, condones the use of bound variables to refer to abstract entities known and unknown, specifiable and unspecifiable, indiscriminately.

    Conceptualism holds that there are universals but they are mind-made. Intuitionism, espoused in modern times in one form or another by Poincaré, Brouwer, Weyl, and others, countenances the use of bound variables to refer to abstract entities only when those entities are capable of being cooked up individually from ingredients specified in advance. As Fraenkel has put it, logicism holds that classes are discovered while intuitionism holds that they are invented—a fair statement indeed of the old opposition between realism and conceptualism. This opposition is no mere quibble; it makes an essential difference in the amount of classical mathematics to which one is willing to subscribe. Logicists, or realists, are able on their assumptions to get Cantor’s ascending orders of infinity;, intuitionists are compelled to stop with the lowest order of infinity,

    

page 15     ON WHAT THERE IS     I

and, as an indirect consequence, to abandon even some of the classical laws of real numbers.⁸ The modern controversy between logicism and intuitionism arose, in fact, from disagreements over infinity.

    Formalism, associated wilth the name of Hilbert, echoes intuitionism in deploring the logicist’s unbridled recourse to universals. But formalism also finds intuitionism unsatisfactory. This could happen for either of two opposite reasons. The formalist might, like the logicist, object to the crippling of classical mathematics; or he might, like the nominalists of old, object to admitting abstract entities at all, even in the restrained sense of mind-made entities. The upshot is the same: the formalist keeps classical mathematics as a play of insignificant notations. This play of notations can still be of utility—whatever utility it has already shown itself to have as a crutch for physicists and technologists. But utility nleed not imply significance, in any literal linguistic sense. Nor need the marked successof mathematicians in spinning out theorems, and in finding objective bases for agreement with one another’s results, imply significance. For an adequate basis for agreement among mathematicians can be found simply in the rules which govern the manipulation of the notations—these syntactical rules being, unlike the notations themselves, quite significant and intelligible.⁹

    I have argued that the sort of ontology we adopt can be consequential—notably in connection with mathematics, although this is only an example. Now how are we to adjudicate among rival ontologies? Certainly the answer is not provided by the semantical formula “To be is to be the value of a variable”; this formula serves rather, conversely, in testing the conformity of a given remark or doctrine to a prior ontological standard. We look to bound variables in connection with ontology not in order to know what there is, but in order to know what a given remark or doctrine, ours or someone else’s, says there is;

    ⁸See below, pp. 125ff.
    ⁹See Goodman and Quine. For further discussion of the general matters touched on in the past two pages, see Bernays [l], Fraenkel, Black.

    

page 16     ON WHAT THERE IS     I

and this much is quite properly a problem involving language. But what there is is another question.

    In debating over what there is, there are still re~ons for operating on a semantical plane. One reason is to escape from the predicament noted at the beginning of this essay: the predicament of my not being able to admit that there are things which McX countenances and I do not. So long aa I adhere to my ontology, aa opposed to McX’s, I cannot allow my bound variables to refer to entities which belong to McX’s ontology and not to mine. I can, however, consistently describe our disagreement by characterizing the statements which M:cX aEirm~. Provided merely that my ontology countenances :linguistic forms, or at least concrete inscriptions and utterances, I can talk about McX’s sentences.

    Another reason for withdrawing to a semantical plane is to find common ground on which to argue. Disagreement, in ontology involves basic disagreement in conceptual schemes; yet McX and I, despite these basic disagreements, find that our conceptual schemes converge sufficiently in their intermediate and upper ramifications to enable us to communicate successfully on such topics as politics, weather, and, in particular, language. In so.far as our basic controversy over ontology can be translated upward into a semantical controversy about words and what to do with them, the collapse of the controversy into question-begging may be delayed.

    It is no wonder, then, that ontological controversy should tend into controversy over language. But we must not ju:mp to the conclusion that what there is depends on words. Translatability of a question into semantical terms is no indication that the question is linguistic. To see Naples is to bear a :name which, when prefixed to the words ‘sees Naples’, yields a true sentence; still there is nothing linguistic about seeing Naples.

    Our acceptance of an ontology is, I think, similar in principle to our acceptance of a scientific theory, say a system of physics: we adopt, at least insofar as we are reasonable, the simplest conceptual scheme into which the disordered fragments of raw experience can be fitted and arranged. Our ontology is

    

page 17     ON WHAT THERE IS     I

determined once we have fixed upon the over-all conceptual scheme which is to accommodate science in the broadest sense; and the considerations which determine a reasonable construction of any part of that conceptual scheme, for example, the biological or the physical part, are not different in kind from the considerations which determine a reasonable construction of the whole. To whatever extent the adoption of any system of scientific theory may be said to be a matter of language, the same—but no more—may be said of the adoption of an ontology.

    But simplicity, as a guiding principle in constructing conceptual schemes, is not a clear and unambiguous idea; and it is quite capable of presenting a double or multiple standard. Imagine, for example, that we have devised the most economical set of concepts adequate to the play-by-play reporting of immediate experience. The entities under this scheme—the values of bound variables—are, let us suppose, individual subjective events of sensation or reflection. We should still find, no doubt, that a physicalistic conceptual scheme, purporting to talk about external objects, offers great advantages in simplifying our over-all reports. By bringing together scattered sense events and treating them as perceptions of one object, we reduce the complexity of our stream of experience to a manageable conceptual simplicity. The rule of simplicity is indeed our guiding maxim in assigning sense data to objects: we associate an earlier and a later round sensum with the same so-called penny, or with two different so-called pennies, in obedience to the demands of maximum simplicity in our total world-picture.

    Here we have two cormpeting conceptual schemes, a phenomenalistic one and a physicalistic one. Which should prevail? Each has its advantages; each has its special simplicity in its own way. Each, I suggest, deserves to be developed. Each may be said, indeed, to be the more fundamental, though in different senses: the one is epistemologically, the other physically, fundamental.

    The physical conceptual scheme simplifies our account of experience because of the way myriad scattered sense events come to be associated with single so-called objects; still there

    

page 18     ON WHAT THERE IS     I

is no likelihood that each sentence about physical objects can actually be translated, however deviously and complexly, into the phenomenalistic language. Physical objects are postulated entities which round out, and simplify our account of the flux of experience, just as the introduction of irrational numbers simplifies laws of arithmetic. From the point of view of the conceptual scheme of the elementary arithmetic of rational numbers alone, the broader arithmetic of rational and irrational numbers would have the status of a convenient myth, simpler than the literal truth (namely, the arithmetic of rationals) and yet, containing that literal truth as a scattered part. Similarly, from a phenamenalistic point of view, the conceptual scheme of physical objects is a convenient myth, simpler than the literal truth and yet containing that literal truth as a scattered part.¹⁰

    Now what of classes or attributes of physical objects, in turn? A platonistic ontology of this sort is, from the point of view of a strictly physicalistic conceptual scheme, as much a myth as that physicalistic conceptual scheme itself is for phenomenalism. This higher myth is a good and useful one, in turn, in so far as it simplifies our account of physics. Since mathematics is an integral part of this higher myth, the utility of this myth for physical science is evident enough. In speaking of it nevertheless as a myth, I echo that philosophy of mathematics to which I alluded earlier under the name of formalism. But an attitude of formalism may with equal justice be adopted toward the physical conceptual scheme, in turn, by the pure aesthete or phenomenalist.

    The analogy between the myth of mathematics and the myth of physics is, in some additional and perhaps fortuitous ways, strikingly close. Consider, for example, the crisis which was precipitated in the foundations of mathematics, at the turn of the century, by the discovery of Russell’s paradox and other antinomies of set theory. These contradictions had to be obviated by unintuitive, ad hoc devices;¹¹ our mathematical myth-making became deliberate and evident to all. But, what

    ¹⁰The arithmetical analogy is due to Frank, pp.108f.
    ¹¹See below, pp. 9Off, 96ff, 122ff.

    

page 19     ON WHAT THERE IS     I

    of physics? An antinomy arose between the undular and the corpuscular accounts of light; and if this was not as out-and-out a contradiction as Russell’s paradox, I suspect that the reason is that physics is not as out-and-out as mathematics. Again, the second great modern crisis in the foundations of mathematics—precipitated in 1931 by Gödel’s proof [2] that there are bound to be undecidable statements in arithmetic—has its companion piece in physics in Heisenberg’s indeterminacy principle.

    In earlier pages I undertook to show that some common arguments in favor of certain ontologies are fallacious. Further, I advanced an explicit standard whereby to decide what the ontological commitments of a theory are. But the question what ontology actually to adopt still stands open, and the obvious counsel is tolerance and an experimental spirit. Let us by all means see how much of the physicalistic conceptual scheme can be reduced to a phenomenalistic one; still, physics also naturally demands pursuing, irreducible in toto though it be. Let us see how, or to what degree, natural science may be rendered independent of platonistic mathematics; but let us also pursue mathematics and delve into its platonistic foundations.

    From among the various conceptual schemes best suited to these various pursuits, one—the phenomenalistic—claims epistemological priority. Viewed from within the phenomenalistic conceptual scheme, the ontologies of physical objects and mathematical objects are myths. The quality of myth, however, is relative; relative, in this case, to the epistemological point of view. This point of view is one among various, corresponding to one among our various interests and purposes.

    

page 20     TWO DOGMAS OF EMPIRICISM    

    

II

TWO DOGMAS OF EMPIRICISM

    Modern empiricism has been conditioned in large part by two dogmas. One is a belief in some fundamental cleavage between truths which are analytic, or grounded in meanings independently of matters of fact and truths which are synthetic, or grounded in fact. The other dogma is reductionism : the belief that each meaningful statement is equivalent to some logical construct upon terms which refer to immediate experience. Both dogmas, I shall argue, are ill founded. One effect of abandoning them is, as we shall see, a blurring of the supposed boundary between speculative metaphysics and natural science. Another effect is a shift toward pragmatism.

1. Background for Analyticity

    Kant's cleavage between analytic and synthetic truths was foreshadowed in Hume's distinction between relations of ideas and matters of fact, and in Leibniz's distinction between truths of reason and truths of fact. Leibniz spoke of the truths of reason as true in all possible worlds. Picturesqueness aside, this is to say that the truths of reason are those which could not possibly be false. In the same vein we hear analytic statements defined as statements whose denials are self-contradictory. But this definition has small explanatory value; for the notion of self-contradictoriness, in the quite broad sense needed for this definition of analyticity, stands in exactly the same need of clarification as does the notion of analyticity itself. The two notions are the two sides of a single dubious coin.

    Kant conceived of an analytic statement as one that attributes to its subject no more than is already conceptually contained

    

page 21     TWO DOGMAS OF EMPIRICISM     II, 1

in the subject. This formulation has two shortcomings: it limits itself to statements of subject-predicate form, and it appeals to a notion of containment which is left at a metaphorical level. But Kant's intent, evident more from the use he makes of the notion of analyticity than from his definition of it, can be restated thus: a statement is analytic when it is true by virtue of meanings and independently of fact. Pursuing this line, let us examine the concept of meaning which is presupposed.

    Meaning, let us remember, is not to be identified with naming¹. Frege's example of 'Evening Star' and 'Morning Star', and Russell’s of ‘Scott’ and 'the author of Waverly’, illustrate that terms can name the same thing but differ in meaning. The distinction between meaning and naming is no less important at the level of abstract terms. The terms '9' and 'the number of planets' name one and the same abstract entity but presumably must be regarded as unlike in meaning; for astronomical observation was needed, and not mere reflection on meanings, to determine the sameness of the entity in question.

    The above examples consist of singular terms, concrete and abstract. With general terms or predicates, the situation is somewhat different but parallel. Whereas a singular term purports to name an entity, abstract or concrete, a general term does not; but a general term is true of an entity, or of each of many, or of none. ² The class of all entities of which a general term is true is called the extension of the term. Now paralleling the contrast between the meaning of a singular term and the entity named, we must distinguish equally between the meaning of a general term and its extension. The general terms 'creature with a heart' and 'creature with kidneys', for example, are perhaps alike in extension but unlike in meaning.

    Confusion of meaning with extension, in the case of general terms, is less common than confusion of meaning with naming in the case of singular terms. It is indeed a common place in philosophy to oppose intention (or meaning) to extension, or, in a variant vocabulary, connotation to denotation.

¹ See above, p.9
² See abve, p.10. and below, pp. 107-115

    

page 22     TWO DOGMAS OF EMPIRICISM     II, 1

    The Aristotelian notion of essence was the forerunner, no doubt, of the modern notion of intension or meaning. For Aristotle it was essential in men to be rational, accidental to be two-legged. But there is an important difference between this attitude and the doctrine of meaning. From the latter point of view it may indeed be conceded (if only for the sake of argument) that rationality is involved in the meaning of the word 'man' while two-leggedness is not; but two-leggedness may at the same time be viewed as involved in the meaning of 'biped' while rationality is not. Thus from the point of view of the doctrine of meaning it makes no sense to say of the actual individual, who is at once a man and a biped, that his rationality is essential and his two-leggedness accidental or vice versa. Things had essences, for Aristotle, but only linguistic forms have meanings. Meaning is what essence becomes when it is divorced from the object of reference and wedded to the word.

    For the theory of meaning the most conspicuous question is as to the nature of its objects: what sort of things are meanings? A felt need for meant entities may derive from an earlier failure to appreciate that meaning and reference are distinct. Once the theory of meaning is sharply separated from the theory of reference, it is a short step to recognizing as the business of the theory of meaning simply the synonymy of linguistic forms and the analyticity of statements; meanings themselves, as obscure intermediary entities, may well be abandoned. ³

    The problem of analyticity then confronts us anew. Statements which are analytic by general philosophical acclaim are not, indeed, far to seek. They fall into two classes. Those of the first class, which may be called logically true, are typified by:

(1)          No unmarried man is married.

The relevant feature of this example is that it is not merely true as it stands, but remains true under any and all reinterpretations of 'man' and 'married.' If we suppose a prior inventory of logical particles, comprising 'no', 'un-', 'not', 'if', 'then', 'and', etc., then in general a logical truth is a statement which is true

³See above, pp. 11f, and below, pp. 48f.

    

page 23     TWO DOGMAS OF EMPIRICISM     II, 1

and remains true under all reinterpretations of its components other than the logical particles.

    But there is also a second class of analytic statements, typified by:

(2)         No bachelor is married.

The characteristic of such a statement is that it can be turned into a logical truth by putting synonyms for synonyms; thus (2) can be turned into (1) by putting 'unmarried man' for its synonym 'bachelor.' We still lack a proper characterization of this second class of analytic statements, and therewith of analyticity generally, inasmuch as we have had in the above description to lean on a notion of “synonymy” which is no less in need of clarification than analyticity itself.

    In recent years Carnap has tended to explain analyticity by appeal to what he calls state-descriptions.⁴ A state-description is any exhaustive assignment of truth values to the atomic, or noncompound, statements of the language. All other statements of the language are, Carnap assumes, built up of their component clauses by means of the familiar logical devices, in such a way that the truth value of any complex statement is fixed for each state-description by specifiable logical laws. A statement is then explained as analytic when it comes out true under every state-description. This account is an adaptation of Leibniz's "true in all possible worlds." But note that this version of analyticity serves its purpose only if the atomic statements of the language are, unlike 'John is a bachelor' and 'John is married,' mutually independent. Otherwise there would be a state-description which assigned truth to 'John is a bachelor' and falsity to 'John is married,' and consequently 'All bachelors are married' would turn out synthetic rather than analytic under the proposed criterion. Thus the criterion of analyticity in terms of state-descriptions serves only for languages devoid of extralogical synonym-pairs, such as 'bachelor' and 'unmarried man'—synonym-pairs of the type which give rise to the "second class" of analytic statements. The criterion in terms of state-descrip-

⁴Carnap [3], pp. 9ff; [4], pp. 70ff.

    

page 24     TWO DOGMAS OF EMPIRICISM     II, 2

tions is a reconstruction at best of logical truth, not of analyticity.

    I do not mean to suggest that Carnap is under any illusions on this point. His simplified model language with its state descriptions is aimed primarily not at the general problem of analyticity but at another purpose, the clarification of probability and induction. Our problem, however, is analyticity; and here the major difficulty lies not in the first class of analytic statements, the logical truths, but rather in the second class, which depends on the notion of synonymy.

2. Definition

    There are those who find it soothing to say that the analytic statements of the second class reduce to those of the first class, the logical truths, by definition; 'bachelor,' for example, is defined as 'unmarried man.' But how do we find that 'bachelor' is defined as 'unmarried man'? Who defined it thus, and when? Are we to appeal to the nearest dictionary, and accept the lexicographer's formulation as law? Clearly this would be to put the cart before the horse. The lexicographer is an empirical scientist, whose business is the recording of antecedent facts; and if he glosses 'bachelor' as 'unmarried man' it is because of his belief that there is a relation of synonymy between these forms, implicit in general or preferred usage prior to his own work. The notion of synonymy presupposed here has still to be clarified, presumably in terms relating to linguistic behavior. Certainly the "definition" which is the lexicographer's report of an observed synonymy cannot be taken as the ground of the synonymy.

    Definition is not, indeed, an activity exclusively of philologists. Philosophers and scientists frequently have occasions to "define" a recondite term by paraphrasing it into terms of a more familiar vocabulary. But ordinarily such a definition, like the philologist's, is pure lexicography, affirming a relationship of synonymy antecedent to the exposition in hand.

    Just what it means to affirm synonymy, just what the inter-

    

page 25     TWO DOGMAS OF EMPIRICISM     II, 2

connections maybe which are necessary and sufficient in order that two linguistic forms be properly describable as synonymous, is far from clear; but, whatever these interconnections may be, ordinarily they are grounded in usage. Definitions reporting selected instances of synonymy come then as reports upon usage.

    There is also, however, a variant type of definitional activity which does not limit itself to the reporting of pre-existing synonymies. I have in mind what Carnap calls explication—an activity to which philosophers are given, and scientists also in their more philosophical moments. In explication the purpose is not merely to paraphrase the definiendum into an outright synonym, but actually to improve upon the definiendum by refining or supplementing its meaning. But even explication, though not merely reporting a pre-existing synonymy between definiendum and definiens, does rest nevertheless on other pre-existing synonymies. The matter may be viewed as follows. Any word worth explicating has some contexts which, as wholes, are clear and precise enough to be useful; and the purpose of explication is to preserve the usage of these favored contexts while sharpening the usage of other contexts. In order that a given definition be suitable for purposes of explication, therefore, what is required is not that the definiendum in its antecedent usage be synonymous with the definiens, but just that each of these favored contexts of the definiendum taken as a whole in its antecedent usage, be synonymous with the corresponding context of the definiens.

    Two alternative definientia may be equally appropriate for the purposes of a given task of explication and yet not be synonymous with each other; for they may serve interchangeably within the favored contexts but diverge elsewhere. By cleaving to one of these definientia rather than the other, a definition of explicative kind generates, by fiat, a relationship of synonymy between definiendum and definiens which did not hold before. But such a definition still owes its explicative function, as seen, to pre-existing synonymies.

    There does, however, remain still an extreme sort of defini-

    

page 26     TWO DOGMAS OF EMPIRICISM     II, 2

tion which does not hark back to prior synonymies at all; namely, the explicitly conventional introduction of novel notations for purposes of sheer abbreviation. Here the definiendum becomes synonymous with the definiens simply because it has been created expressly for the purpose of being synonymous with the definiens. Here we have a really transparent case of synonymy created by definition; would that all species of synonymy were as intelligible. For the rest, definition rests on synonymy rather than explaining it.

    The word "definition" has come to have a dangerously reassuring sound, due no doubt to its frequent occurrence in logical and mathematical writings. We shall do well to digress now into a brief appraisal of the role of definition in formal work.

    In logical and mathematical systems either of two mutually antagonistic types of economy may be striven for, and each has its peculiar practical utility. On the one hand we may seek economy of practical expression—ease and brevity in the statement of multifarious relationships. This sort of economy calls usually for distinctive concise notations for a wealth of concepts. Second, however, and oppositely, we may seek economy in grammar and vocabulary; we may try to find a minimum of basic concepts such that, once a distinctive notation has been appropriated to each of them, it becomes possible to express any desired further concept by mere combination and iteration of our basic notations. This second sort of economy is impractical in one way, since a poverty in basic idioms tends to a necessary lengthening of discourse. But it is practical in another way: it greatly simplifies theoretical discourse about the language, through minimizing the terms and the forms of construction wherein the language consists.

    Both sorts of economy, though prima facie incompatible, are valuable in their separate ways. The custom has consequently arisen of combining both sorts of economy by forging in effect two languages, the one a part of the other. The inclusive language, though redundant in grammar and vocabulary, is economical in message lengths, while the part, called primitive

    

page 27     TWO DOGMAS OF EMPIRICISM     II, 4

Notation is economical in grammar and vocabulary. Whole and part are correlated by rules of translation whereby each idiom not in primitive notation is equated to some complex built up of primitive notation. These rules of translation are the so-called definitions which appear in formalized systems. They are best viewed not as adjuncts to one language but as correlations between two languages, the one a part of the other.

    But these correlations are not arbitrary. They are supposed to show how the primitive notations can accomplish all purposes, save brevity and convenience, of the redundant language. Hence the definiendum and its definiens may be expected, in each case, to be related in one or another of the three ways lately noted. The definiens may be a faithful paraphrase of the definiendum into the narrower notation, preserving a direct synonymy⁵ as of antecedent usage; or the definiens may, in the spirit of explication, improve upon the antecedent usage of the definiendum; or finally, the Definiendum may be a newly created notation, newly endowed with meaning here and now.

    In formal and informal work alike, thus, we find that definition— except in the extreme case of the explicitly conventional introduction of new notation—hinges on prior relationships of synonymy. Recognizing then that the notation of definition does not hold the key to synonymy and analyticity, let us look further into synonymy and say no more of definition.

3. Interchangeability

    A natural suggestion, deserving close examination, is that the synonymy of two linguistic forms consists simply in their interchangeability in all contexts without change of truth value; interchangeability, in Leibniz's phrase, salva veritate.⁶ Note that synonyms so conceived need not even be free from vagueness, as long as the vaguenesses match.

    ⁵According to an important variant sense of ‘definition’, the relation preserved may be the weaker relation of mere agreement in reference; see below, p. 132. But definition in this sense is better ignored in the present connection, being irrelevant to the question of synonymy.
    ⁶CF. Lewis [1], p. 373.

    

page 28     TWO DOGMAS OF EMPIRICISM     II, 3

    But it is not quite true that the synonyms 'bachelor' and 'unmarried man' are everywhere interchangeable salva veritate. Truths which become false under substitution of 'unmarriedman' for 'bachelor' are easily constructed with help of 'bachelor of arts' or 'bachelor's buttons.' Also with help of quotation, thus:

        'Bachelor' has less than ten letters.

Such counterinstances can, however, perhaps be set aside by treating the phrases 'bachelor of arts' and 'bachelor's buttons' and the quotation 'bachelor' each as a single indivisible word and then stipulating that the interchangeability salva veritate which is to be the touchstone of synonymy is not supposed to apply to fragmentary occurrences inside of a word. This account of synonymy, supposing it acceptable on other counts, has indeed the drawback of appealing to a prior conception of "word" which can be counted on to present difficulties of formulation in its turn. Nevertheless some progress might be claimed in having reduced the problem of synonymy to a problem of wordhood. Let us pursue this line a bit, taking "word" for granted.

    The question remains whether interchangeability salva veritate (apart from occurrences within words) is a strong enough condition for synonymy, or whether, on the contrary, some non-synonymous expressions might be thus interchangeable. Now let us be clear that we are not concerned here with synonymy in the sense of complete identity in psychological associations or poetic quality; indeed no two expressions are synonymous in such a sense. We are concerned only with what may be called cognitive. synonymy. Just what this is cannot be said without successfully finishing the present study; but we know something about it from the need which arose for it in connection with analyticity in §1. The sort of synonymy needed there was merely such that any analytic statement could be turned into a logical truth by putting synonyms for synonyms. Turning the tables and assuming analyticity, indeed, we could explain cognitive synonymy of terms as follows (keeping to the familiar example): to say that 'bachelor' and 'unmarried man' are cognitively sy-

    

page 29     TWO DOGMAS OF EMPIRICISM     II, 3

nonymous is to say no more nor less than that the statement:

(3)    All and only bachelors are unmarried men

is analytic.⁷

    What we need is an account of cognitive synonymy not presupposing analyticity— if we are to explainan alyticity conversely with help of cognitive synonymy as undertaken in §1. And indeed such an independent account of cognitive synonymy is at present up for consideration, namely, interchangeability salva veritate everywhere except within words. The question before us, to resume the thread at last, is whether such interchangeability is a sufficient condition for cognitive synonymy. We can quickly assure ourselves that it is, by examples of the following sort. The statement:

(4)    Necessarily all and only bachelors are bachelors

is evidently true, even supposing 'necessarily' so narrowly construed as to be truly applicable only to analytic statements. Then, if 'bachelor' and 'unmarried man' are interchangeable salva veritate, the result:

(5)    Necessarily, all and only bachelors are unmarried men

of putting 'unmarried man' for an occurrence of 'bachelor' in (4) must, like (4), be true. But to say that (5) is true is to say that (3) is analytic, and hence that 'bachelor' and 'unmarried man' are cognitively synonymous.

    Let us see what there is about the above argument that gives it its air of hocus-pocus. The condition of interchangeability salva veritate varies in its force with variations in the richness of the language at hand. The above argument supposes we are working with a language rich enough to contain the adverb 'necessarily,' this adverb being so construed as to yield truth

    ⁷This is cognitive synonymy in a primary, broad sense. Carnap ([3], pp. 56ff) and Lewis ([2], pp. 83ff) have suggested how, once this notion is at hand, a narrower sense of cognitive synonymy which preferable for some purposes can in turn be derived. But this special ramification of concept-building lies aside from the present purposes and must not be confused with the broad sort of cognitive synonymy here concerned.

    

page 30     TWO DOGMAS OF EMPIRICISM     II, 3

when and only when applied to an analytic statement. But can we condone a language which contains such an adverb? Does the adverb really make sense? To suppose that it does is to suppose that we have already made satisfactory sense of 'analytic.' Then what are we so hard at work on right now?

    Our argument is not flatly circular, but something like it. It has the form, figuratively speaking, of a closed curve in space.

    Interchangeability salva veritate is meaningless until relativized to a language whose extent is specified in relevant respects. Suppose now we consider a language containing just the following materials. There is an indefinitely large stock of oneplace predicates (for example 'F' where 'Fx' means that x is a man) and many-place predicates,(for example, 'G' were 'Gxy' means that x loves y), mostly having to do with extralogical subject matter. The rest of the language is logical. The atomic sentences consist each of a predicate followed by one or more variables 'x', 'y'. etc.; and the complex sentences are built up of atomic ones by truth functions ('not', 'and', 'or', etc.) and quantification.⁸ In effect such a language enjoys the benefits also of descriptions and indeed singular terms generally, these being contextually definable in known ways.⁹ Even abstract singular terms naming classes, classes of classes, etc., are contextually definable in case the assumed stock of predicates includes the two-place predicate of class membership.¹⁰ Such a language can be adequate to classical mathematics and indeed to scientific discourse generally, except in so far as the latter involves debatable devices such as contrary-to-fact conditionals or modal adverbs like 'necessarily'.¹¹ Now a language of this type is extensional, in this sense: any two predicates which agree extensionally (that is, are true of the same objects) are interchangeable salva veritate.¹²

    ⁸Pp. 81ff, below, contain a description of just such a language, except that there happens there to be just one predicate, the two-place predicate 'ε'
    ⁹See above, pp. 5-8; also below, pp. 85f, 166f.
    ¹⁰See below, p. 87.
    ¹¹On such devices see also Essay VIII.
    ¹²This is the substance of Quine [1], *121.

    

page 31     TWO DOGMAS OF EMPIRICISM     II, 3

    In an extensional language, therefore, interchangeability salva veritate is no assurance of cognitive synonymy of the desired type. That 'bachelor' and 'unmarried man' are interchangeable salva veritate in an extensional language assures us of no more than that (3) is true. There is no assurance here that the extensional agreement of 'bachelor' and 'unmarried man' rests on meaning rather than merely on accidental matters of fact, as does the extensional agreement of 'creature with a heart' and 'creature with kidneys'.

    For most purposes extensional agreement is the nearest approximation to synonymy we need care about. But the fact remains that extensional agreement falls far short of cognitive synonymy of the type required for explaining analyticity in the manner of §I. The type of cognitive synonymy required there is such as to equate the synonymy of 'bachelor' and 'unmarried man' with the analyticity of (3), not merely with the truth of (3).

    So we must recognize that interchangeability salva veritate, if construed in relation to an extensional language, is not a sufficient condition of cognitive synonymy in the sense needed for deriving analyticity in the manner of Section I. If a language contains an intensional adverb 'necessarily' in the sense lately noted, or other particles to the same effect, then interchangeability salva veritate in such a language does afford a sufficient condition of cognitive synonymy; but such a language is intelligible only if the notion of analyticity is already clearly understood in advance.

    The effort to explain cognitive synonymy first, for the sake of deriving analyticity from it afterward as in Section I, is perhaps the wrong approach. Instead we might try explaining analyticity somehow without appeal to cognitive synonymy. Afterward we could doubtless derive cognitive synonymy from analyticity satisfactorily enough if desired. We have seen that cognitive synonymy of 'bachelor' and 'unmarried man' can be explained as analyticity of (3). The same explanation works for any pair of one-place predicates, of course, and it can be extended in obvious fashion to many-place predicates. Other syntactical categories can also be accommodated in fairly parallel fashion.

    

page 32     TWO DOGMAS OF EMPIRICISM     II, 4

    Singular terms may be said to be cognitively synonymous when the statement of identity formed by putting '=' between them is analytic. Statements maybe said simply to be cognitively synonymous when their biconditional (the result of joining them by 'if and only if') is analytic.¹³ If we care to lump all categories into a single formulation, at the expense of assuming again the notion of "word" which was appealed to early in this section, we can describe any two linguistic forms as cognitively synonymous when the two forms are interchangeable (apart from occurrences within "words") salva (no longer veritate but) analyticitate. Certain technical questions arise, indeed, over cases of ambiguity or homonymy; let us not pause for them, however, for we are already digressing. Let us rather turn our backs on the problem of synonymy and address ourselves anew to that of analyticity.

4. Semantical Rules

Analyticity at first seemed most naturally definable by appeal to a realm of meanings. On refinement, the appeal to meanings gave way to an appeal to synonymy or definition. But definition turned out to be a will-o'-the-wisp, and synonymy turned out to be best understood only by dint of a prior appeal to analyticity itself. So we are back at the problem of analyticity.

    I do not know whether the styatement 'Everything green is extended' is analytic. Now does my indecision over thisexample really betray an incomp[llete understanding, an incomplete grasp of the "meanings", of 'green' and 'extended'j? I think not. Thew trdfouble is not with 'green' or 'extended', but with 'analytic'.

    It is often hinted that the difficulty in separating analytic statements from synthetic ones in ordinary language is due to the vagueness of ordinary language and that the distinction is clear when we have a precise artificial language with explict "semantic rules." This, however, as I shall now attempt to show, is a confusion.

    ¹³The 'of and only if' itself is intended in the truth functional sense. See Carnap [3], p.14.

    

page 33     TWO DOGMAS OF EMPIRICISM     II, 4

    The notion of analyticity about which we are worrying is a purported relation bewtween statements and languages: a statement S is said to be analytic for a language L, and the problem is to make sense of this relation generally, that is, for variable 'S' and 'L'. The gravity of this problem is not perceptibly less for artificial languages than for natural ones. The problem of making sense of the idiom 'S' is analytic for L', with variable 'S' and 'L', retains its stubbornness even if we limit the range of the variable 'L' to artificial languages. Let me now try to make this point evident.

    For artificial languages and semantical rules we look naturally to the writings of Carnap. His semantical rules take various forms, and to make my point I shall have to distinguish certain of the forms. Let us suppose, to begin with, an artificial language L₀ whose semantical rules have the form explicitly of a specification, by recursion or otherwise, of all the analytic statements of L₀. The rules tell us that such and such statements, and only those, are the analytic statements of L₀. Now here the difficulty is simply that the rules contain the word 'analytic', which we do not understand! We understand what expressions the rules attribute analyticity to, but we do not understand what the rules attribute to those expressions. In short, before we can understand a rule which begins 'A statement S is analytic for language L₀ if and only if ...', we must understand the general relative term 'analytic for'; we must understand 'S is analytic for L' where S and L are variables.

    Alternatively we may, indeed, view the so-called rule as a conventional definition of a new simple symbol 'analytic-for-L₀', which might better be written untendentiously as K so as not to seem to throw light on the interesting word 'analytic'. Obviously any number of classes of K,M,N, etc., of statements of L₀ can be specified for various purposes or for no purpose; what does it mean to say that K, as against M, N, erc., is the class of the "analytic" statements of L₀.

    By saying what statements are analytic for L₀ we explain 'analytic-for-L₀' but not 'analytic', not 'analytic for'. We do not begin to explain the idiom 'S is analytic for L' with variable

    

page 34     TWO DOGMAS OF EMPIRICISM     II, 4

'S' and 'L', even if we are content to limit the range of 'L' to the realm of artificial languages.

    Actually we do know enough about the intended significance of 'analytic' to know that analytic statements are suposed to be true. Let us then turn to a second form of semantical rule, which says not that such and such statents are analaytic but simply that such and such statements are included among the truths. Such a rule is not subject to the criticism of contining the un-understood word 'analytic'; and we may grant for the sake of argument that there is no difficulty over the broader term 'true'. A semantical rule of this second ypte, a rule of truth, is not supposed to specify all the truths of the language; it merely stipulates, recursively or otherwise, a certain multitude of statements which, along with others unspecified, are to count as true. Such a rule may be conceded to be quite clear. Derivatively, afterward, analyticity can be demarcated thus: a statement is analytic if it is (not merely true but) true according to the semantic rule.

    Still there is really no progress. Instead of appealing to an unexplained word 'analytic', we are now appealing to an unexplained phrase 'semantical rule'. Not every true statement which says that the statements of some class are true can count as a semantical rule—otherwise all truths would be "analytic" in the sense of being true according to semantical rules. Semantical rules are distinguishable, apparently, only by the fact of appearing on a page under the heading 'Semantical Rules'; and this heading is itself then meaningless.

    We can say indeed that a statement is analytic-for-L₀ if and only if it is true according to such and such specifically appended "semantical rules," but then we find ourselves back at essentially the same case which was originally discusses: 'S is analytic-for-L₀ if and only if. . . .' Once we seek to explain 'S is analytic-for-L' generally for variable L (even allowing limitation of L to artificial languages), the explanation 'true according to the semantical rules of L' is unavailaing; for the relative term 'semantical rule of' is as much in need of clarification, at least, as 'analytic for'.

    

page 35     TWO DOGMAS OF EMPIRICISM     II, 4

    It may be instructive to compare the notion of semantical rule with that of postulate. Relative to a given set of postulates, it is easy to say what a postulate is; it is a member of the set. Relative to a given set of semantical rules, it is equally easy to say what a semantical rule is. But given simply a notation, mathematical or otherwise, and indeed as thoroughtly understood a notation as you please in point of the translations or truth conditions of its statements, who can say which of its true statements rank as postulates? Obviously the question is meaningless—as meaningless as asking which points in Ohio are starting points. Any finite (or effectively specifiable infinite) selection of statements (preferably true ones, perhaps) is as much a set of postulates as any other. The word 'postulate' is significant only relative to an act of inquiry; we apply the word to a set of statements just in so far as we happen, for the year or the moment, to be thinking of those statements in relation to the statements which can be reached from them by some set of transformations to which we have seen fit to direct our attention. Now the notion of semantical rule is as sensible and meaningful as that of postulate, if conceived in a similarly relative spirit—relative, this time, to one or another particular enterprise of schooling unconversant persons in sufficient conditions for truth of statements of some natural or artificial language L. But from this point of view no one signalization of a subclass of the truths of L is intrinsically more a semantical rule than another; and , if 'analytic' means 'true by semantical rules', no one truth of L is analytic to the exclusion of another.¹⁴

    It might conceivably be protested that an artificial language L (unlike a natural one) is a language in the ordinary sense plus a set of explicit semantical rules — the whole constituting, let us say, an ordered pair; and that the semantical rules of L then are specifiable simply as the second component of the pair L. But, by the same token and more simply, we might construe an artificial language L outright as an ordered pair whose second

    ¹⁴The foregoing paragraph ws not part of the present essay as originally published. It was prompted by Martin (see Bibliography), as was the end of Essay VII.

    

page 36     TWO DOGMAS OF EMPIRICISM     II, 4

component is the class of its analytic statements; and then the analytic statements of L become specifiable simply as the statements in the second component of L. Or better still, we might just stop tugging at our bootstraps altogether.

    Not all the explanations of analyticity known to Carnap and his readers have been covered explicitly in the above considerations, but the extension to other forms is not hard to see. Just one additional factor should be mentioned which sometimes enters: sometimes the semantical rules are in effect rules of translation into ordinary language, in which case the analytic statements of the artificial language are in effect recognized as such from the analyticity of their specified translations in ordinary language. Here certainly there can be no thought of an illumination of the problem of analyticity from the side of the artificial language.

    From the point of view of the problem of analyticity the notion of an artificial language with semantical rules is a feu follet par ercellence [will o’ the wisp]. Semantical rules determining the analytic statements of an artificial language are of interest only in so far as we already understand the notion of analyticity; they are of no help in gaining this understanding.

    Appeal to hypothetical languages of an artificially simple kind could conceivably be useful in clarifying analyticity, if the mental or behavioral or cultural factors relevant to analyticity—whatever they may be — were somehow sketched into the simplified model. But a model which takes analyticity merely as an irreducible character is unlikely to throw light on the problem of explicating analyticity.

    It is obvious that truth in general depends on both language and extra-linguistic fact. The statement 'Brutus killed Caesar' would be false if the world had been different in certain ways, but it would also be false if the word 'killed' happened rather to have the sense of 'begat.' Hence the temptation to suppose in general that the truth of a statement is somehow analyzable into a linguistic component and a factual component. Given this supposition, it next seems reasonable that in some statements the factual component should be null; and these are the analytic

    

page 37     TWO DOGMAS OF EMPIRICISM     II, 5

statements. But, for all its a priori reasonableness, a boundary between analytic and synthetic statement simply has not been drawn. That there is such a distinction to be drawn at all is an unempirical dogma of empiricists, a metaphysical article of faith.

5. THE VERIFICATION THEORY AND REDUCTIONISM

    In the course of these somber reflections we have taken a dim view first of the notion of meaning, then of the notion of cognitive synonymy, and finally of the notion of analyticity. But what, it may be asked, of the verification theory of meaning? This phrase has established itself so firmly as a catchword of empiricism that we should b’e very unscientilic. indeed not to look beneath it for a possible key to the problem of meaning and the associated problems.

    The verification theory of meaning, which has been conspicuous in the literature f.rom Peirce onward, is that the meaning of a statement is the method of empirically confirming or infirming it. An analytic statement is that limiting case which is confirmed no matter what.

    As urged in §1, we can as well pass over the question of meanings as entities and move straight to sameness of meaning, or synonymy. Then what the verification theory says is that statements are synonymous if and only if they are alike in point of method of empirical confirmation or infirmation.

    This is an account of cognitive synonymy not of linguistic forms generally, but of statements.¹⁵ However, from the concept of synonymy of statements we could derive the concept of synonymy for other linguistic forms, by considerations somewhat similar to those at the end of §3. Assuming the notion of “word,” indeed, we could explain any two forms as synonymous when the

    ¹⁵The doctrine can indeed be formulated with terms rather than statements as the units. Thus ILewis describes the meaning of a term as “a criterion in mind, by reference to which one is able to apply or refuse to apply the expression in question in the case of presented, or imagined, things or situations” ([2], p. 133).—For an instructive account of the vicissitudes of the verification theory of meaning, centered however on the question of meaningfulness rather than synonymy and analyticity, see Hempel.

    

page 38     TWO DOGMAS OF EMPIRICISM     II, 5

putting of the one form for an occurrence of the other in any statement (apart from occurrences within “words”) yields a synonymous statement. Finally, given the concept of synonymy thus for linguistic forms generally, we could define analyticity in terms of synonymy and logical truth as in §1. For that matter, we could define analyticity more simply in terms of just synonymy of statements together with logical truth; it is not necessary to appeal to synonymy of linguistic forms other than statements. For a statement may be described as analytic simply when it is synonymous with a logically true statement.

    So, if the verification theory can be accepted ,asan adequate account of statement synonymy, the notion of analyticity is saved after all. However, let us reflect. Statement synonymy is said to be likeness of method of empirical confirmation or infirmation. Just what are these methods which are to be compared for likeness? What, in other words, is the nature of the relation between a statement and the experiences which contribute to or detract from its confirmation?

    The most naive view of the relation is that it is one of direct report. This is radical reductionism. Every meaningful statemon t is held to be translatable into a statement (true or false) abou,t immediate experience. Radical reductionism, in one form or another, well antedates the verification theory of meaning explicitly so called. Thus Locke and Hume held that every idea must either originate directly in sense experience or else be compounded of ideas thus originating; and taking a hint from Tooke we might rephrase this doctrine in semantical jargon by saying that a term, to be significant at all, must be either a name of a sense datum or a compound of such names or an abbrevia,tion of such a compound. So stated, the doctrine remains ambiguous as between sense data as sensory ev’ents and sense data as sensory qualities; and it remains vague as to the ad,missible ways of compounding. Moreover, the doctrine is unnecessarily and intolerably restrictive in the term-by-term critique which it imposes. More reasonably, an.d without yet exceeding the limits of what I have called radical reductionism, we may take full statements as our significant units—thus

    

page 39     TWO DOGMAS OF EMPIRICISM     II, 5

demanding that our statements as wholes be translatable into sense-datum language, but not that they be translatable term by term.

    This emendation would unquestionably have been welcome to Locke and Hume and Tooke, but historically it had to await an important reorientation in semantics—the reorientation whereby the primary vehicle of meaning came to be seen no longer in the term but in the statement. This reorientation, explicit Frege ([1], §60), underlies Russell’s concept of incomplete symbols defined in use;¹⁶ also it is implicit in the verification theory of meaning, since the objects of verification are statements.

    Radical reductionism, conceived now with statements as uniti, set itself the task of specifying a sensedatum language and showing how to translate the rest of significant discourse, statement by statement, into it. Carnap embarked on this project in the Aufbau.

    The language which Carnap adopted as his starting point was not a sense-datum language in the narrowest conceivable sense, for it included also the notations of logic, up through higher set theory. In effect it included the whole language of pure mathematics. The ontology implicit in it (that is, the range of values of its variables) embraced not only sensory events but classes, classes of classes, and so on Empiricists there are who would boggle at such procligality. Carnap’s starting point is very parsimonious, however, in its extralogicel or sensory part. In a series of constructions :in which he exploits the resources of modern logic with much ingenuity, Carnap succeeds in defining a wide array of import.ant additional sensory concepts which, but for his constructions, lone would not have dreamed were definable on so slender a basis. He was the first empiricist who, not content with asserting the reducibility of science to terms of immediate experience, took serious steps toward carrying out the reduction.

    If Carnap’s starting point is satisfactory, still his construc-

    ¹⁶See above, p. 6.

    

page 40     TWO DOGMAS OF EMPIRICISM     II, 5

tions were, as he himself stressed, only a fragment of the full program. The construction of even the simplest statements about the physical world was left in a sketchy state. Carnap’s suggestions on this subject were, despite their sketchiness, very suggestive. He explained spatio-temporal point-instants as quadruples of real numbers and envisaged assignment of sense qualities to point-instants according to certain canons. Roughly summarized, the plan was that qualities should be assigned to point-instants in such a way as to achieve the laziest worl’d compatible with our experience. The principle of least action was to be our guide in constructing a world from experience.

    Carnap did not seem to recognize, however, that his treatment of physical objects fell short of reduction not merely through sketchiness, but in principle. Statements of the form ‘Quality q is at point-instant x; y; z; t’ were, according to his canons, to be apportioned truth values in such a way as to maximize and minimize certain over-all features, and with growth of experience the truth values were to be progressively revised in the same spirit. I think this is a good schematieation (deliberately oversimplified, to be sure) of what science really does; but it provides no indication, not even the sketchiest, of how a statement of the form ‘Quality q is at x; y; z; t’ could ever be translated into Carnap’s initial language of sense data and logic. The connective ‘is at’ remains an added undefined connective; the canons counsel us in its use but not in its elimination.

    Carnap seems to have appreciated this point afterward; for in his later writings he abandoned all notion of the translatability of statements about the physical world into statements about immediate experience. Reductionism in its radical form has long since ceased to figure in Carnap’s philosophy.

    But the dogma of reductionism has, in a subtler and more tenuous form, continued to influence the thought of empiricist.. The notion lingers that to each statement, or each synthetic statement, there is associated a unique range of possible sensory events such that the occurrence of any of them would add to the likelihood of truth of the statement, and that there is associated

    

page 41     TWO DOGMAS OF EMPIRICISM     II, 5

also another unique range of possible sensory events whose occurrence would detract from that likelihood. This notion is of course implicit in the verification theory of meaning.

    The dogma of reductionism survives in the supposition that each statement, taken in isolation from its fellows, can admit of confirmation or infirmation at all. My countersuggestion, issuing essentially from Carnap’s doctrine of the physical world in the Aufbau, is that our statements about the external world face the tribunal of sense experience not individually but only as a corporate body.¹⁷

     The dogma of reductionism, even in its attenuated form, is intimately connected with the other dogma—that there is a cleavage between the analytic and the synthetic. We have found ourselves led, indeed, from the latter problem to the former through the verification theory of meaning. More directly, the one dogma clearly supports the other in this way: as long as it is taken to be significant in general to speak of the confirmation and infirmation of a statement, it seems significant to speak also of a limiting kind of statement which is vacuously confirmed, ipso facto, come what may; and such a statement is analytic.

    The two dogmas are, indeed, at root identical. We lately reflected that in general the truth of statements does obviously depend both upon language and upon extralinguistic fact; and we noted that this obvious circumstance carries in its train, not logically but all too naturally, a feeling that the truth of a statement is somehow analyzable into a linguistic component and a factual component. The factual component must, if we are empiricists, boil down to a range of confirmatory experiences. In the extreme case where the linguistic component is all that matters, a true statement is analytic. But I hope we are now impressed with how stubbornly the distinction between analytic and synthetic has resisted any straightforward drawing. I am impressed also, apart from prefabricated examples of black and white balls in an urn, with how baffling the problem has always

    ¹⁷This doctrine was well argued by Duhem, pp. 303-328. Or see Lowinger, pp. 132-140.

    

page 42     TWO DOGMAS OF EMPIRICISM     II, 6

been of arriving at any explicit theory of the empirical confirmation of a synthetic statement. My present suggestion is that it is nonsense, and the root of much nonsense, to speak of a linguistic component and a factual component in the truth of any individual statement. Taken collectively, science has its double dependence upon language and experience; but this duality is not significantly traceable into the statements of science taken one by one.

    The idea of defining a symbol in use was, as remarked, an advance over the impossible term-by-term empiricism of Locke and Hume. The statement, rather than the term, came with. Bentham to be recognized as the unit accountable to an empiricist critique. But what I am now urging is that even in taking the statement as unit we have drawn our grid too finely. The unit of empirical significance is the whole of science.

6. Empiricism without the Dogmas

    The totality of our so-called knowledge or beliefs, from the most casual matters of geography and history to the profoundest laws of atomic physics or even of pure mathematics and logic, is a man-made fabric which impinges on experience only along the edges. Or, to change the figure, total science is like a field of force whose boundary conditions are experience. A conflict with experience at the periphery occasions readjustments in the interior of the field. Truth values have to be redistributed over some of our statements. Reevaluation of some statements entails reevaluation of others, because of their logical interconnections—the logical laws being in turn simply certain further statements of the system, certain further elements of the field. Having reevaluated one statement we must reevaluate some others, which may be statements logically connected with the first or may be the statements of logical connections themselves. But the total field is so underdetermined by its boundary conditions, experience, that there is much latitude of choice as to what statements to reevaluate in the light of any single contrary

    

page 43     TWO DOGMAS OF EMPIRICISM     II, 6

experience. No particular experiences are linked with any particular statements in the interior of the field, except indirectly through considerations of equilibrium affecting the field as a whole.

    If this view is right, it is misleading to speak of the empirical content of an individual statement—especially if it is a statement at all remote from the experiential periphery of the field. Furthermore it becomes folly to seek a boundary between synthetic statements, which hold contingently on experience, and analytic statements, which hold come what may. Any statement can be held true come what may, if we make drastic enough adjustments elsewhere in the system. Even a statement very close to the periphery can be held true in the face of recalcitrant experience by pleading hallucination or by amending certain statements of the kind called logical laws. Conversely, by the same token, no statement is immune to revision. Revision even of the logical law of the excluded middle has been proposed as a means of simplifying quantum mechanics; and what difference is there in principle between such a shift and the shift whereby Kepler superseded Ptolemy, or Einstein Newton, or Darwin Aristotle?

    For vividness I have bleen speaking in terms of varying distances from a sensory periphery. Let me try now to clarify this notion without metaphor. Certain statements, though about physical objects and not sense experience, seem peculiarly germane to sense experience—and in a selective way: some statements to some experiences, others to others. Such statements, especially germane to particular experiences, I picture as near the periphery. But in this relation of “germaneness” I envisage nothing more than a loose association reflecting the relative likelihood, in practice, of our choosing one statement rather than another for revision in the event of recalcitrant experience. For example, we can imagine recalcitrant experiences to which we would surely be inclined to accommodate our system by reevaluating just the statement that there are brick houses on Elm Street, together with related statements on the same

    

page 44     TWO DOGMAS OF EMPIRICISM     II, 6

topic. We can imagine other recalcitrant experiences to which we would be inclined to accommodate our system by reëvaluating just the statement that there are no centaurs, along with kindred statements. A recalcitrant experience can, I have urged, be accommodated by any of various alternative reevaluations in various alternative quarters of the total system; but, in the cases which we are now imagining, our natural tendency to disturb the total system as little as possible would lead us to focus our revisions upon these specific statements concerning brick houses or centaurs. These statements are felt, therefore, to have a sharper empirical reference than highly theoretical statements of physics or logic or ontology. The latter statements may be thought of as relatively centrally located within the total network, meaning merely that little preferential connection with any particular sense data obtrudes itself.

    As an empiricist I continue to think of the conceptual scheme of science as a tool, ultimately, for predicting future experience in the light of past experience. Physical objects are conceptually imported into the situation as convenient intermediaries—not by definition in terms of experience, but simply as irreducible posits¹⁸ comparable, epistemologically, to the gods of Homer. For my part I do, qua lay physicist, believe in physical olbjects and not in Homer’s gods; and I consider it a scientific error to believe otherwise. But in point of epistemological footing the physical objects and the gods differ only in degree and not in kind. Both sorts of entities enter our conception only as cultural posits. The myth of physical objects is epistemologically sulperior to most in that it has proved more efficacious than other myths as a device for working a manageable structure into the flux of experience.

    Positing does not stop with macroscopic physical objects. Objects at the atomic level are posited to make the laws of macroscopic objects, and ultimately the laws of experience, simpler and more manageable; and we need not expect or demand full definition of atomic and subatomic entities in terms of macroscopic ones, any more than definition of macroscopic

    ¹⁸Cf. pp. 17f above.

    

page 45     TWO DOGMAS OF EMPIRICISM     II, 6

things in terms of sense data. Science is a continuation of common sense, and it continues the common-sense expedient of swelling ontology to simplify theory.

    Physical objects, small and large, are not the only posita. Forces are another example; and indeed we are told nowadays that the boundary between energy and matter is obsolete. Moreover, the abstract entities which are the substance of mathematics—ultimately classes and classes of classes and so on up—are another posit in the same spirit. Epistemologically these are myths on the Bame footing with physical objects and gods, neither better nor worse except for differences in the degree to which they expedite our dealings with sense experiences.

    The over-all algebra of rational and irrational numbers is underdetermined by the algebra of rational numbers, but is smoother and more convenient; and it includes the algebra of rational numbers as a jagged or gerrymandered part.¹⁹ Total science, mathematical and natural and human, is similarly but more extremely underdetermined by experience. The edge of the system must be kept squared with experience; the rest, with all its elaborate myths or fictions, has a+~its objective the simplicity of laws.

    Ontological questions, under this view, are on a par with questions of natural science.²⁰ Consider the question whether to countenance classes as entities. This, as I have argued elsewith respect to where,²¹ is the question ,whlether to quantify variables which take classes so values. Now Carnap [6] haa maintained that this is a question not of matters of fact but of choosing a convenient language form, a convenient conceptual scheme or framework for science. With this I agree, but only on the proviso that the same be conceded regarding scientific hypotheses generally. Carnap ([6], p. 32n) has recognized that he is able to preserve a double standard for ontological questions and scientific hypotheses only by assuming an absolute distinc-

    ¹⁹Cf. p. 18 above.
    ²⁰“L’ontologie fait corps avec la science ell-même et ne peut en être separée.” Meyerson, p. 439.
    ²¹Above, pp. 12f; below, pp. 102ff.

    

page 46     TWO DOGMAS OF EMPIRICISM     II, 6

tion between the analytic and the synthetic; and I need not say again that this is a distinction which I reject.²²

    The issue over there being classes seems more a questilon of convenient conceptual scheme; the issue over there being centaurs, or brick houses on Elm Street, seems more a questi’on of fact. But I have been urging that this difference is only one of degree, and that it turns upon our vaguely pragmatic inclination to adjust one strand of the fabric of science rather than anNother in accommodating some particular recalcitrant experience. Conservatism figures in such choices, and so does the quest for simplicity.

    Carnap, Lewis, and others take a pragmatic stand on the question of chloosing between language forms, scientific frameworks; but their pragmatism leaves off at the imagined boundary between the analytic and the synthetic. In repudiating such a boundary I espouse a more thorough pragmatism. Each man is given a scientific heritage plus a continuing barrage of sensory stimulation; and the considerations which guide him i:n warping his scientific heritage to fit his continuing sensory promptings are, where rational, pragmatic.

    ²²For an effective see White [2].

    

page 47     MEANING IN LINGUISTICS    

    

    

III

THE PROBLEM OF MEANING
IN LINGUISTICS

1

Lexicography is concerned, or seems to be concerned, with identification of meanings, and the investigation of semantic change is concerned with change of meaning. Pending a satisfactory explanation of the notion of meaning, linguists in semantic fields are in the situation of not knowing what they are talking about. This is not an untenable situation. Ancient astronomers knew the movements of the planets remarkably well without knowing what sort of things the planets were. But it is a theoretically unsatisfactory situation, as the more theoretically minded among the linguists are painfully aware.

    Confusion of meaning with reference¹ has encouraged a tendency to take the notion of meaning for granted. It is felt that the meaning of the word ‘man’ is as tangible as our neighbor and that the meaning of the phrase ‘Evening Star’ is as clear as the star in the sky. And it is felt that to question or repudiate the notion of meaning is to suppose a world in which there is just language and nothing for language to refer to. Actually we can acknowledge a worldful of objects, and let our singular and general terms refer to those objects in their several ways to our hearts’ content, without ever taking up the topic of meaning.

    An object referred to, named by a singular term or denoted by a general term, can be anything under the sun. Meanings, however, purport to be entities of a special sort: the meaning of

    ¹See above, pp. 9, 21f.

    

page 48     MEANING IN LINGUISTICS     III, 1

an expression is the idea expressed. Now there is considerable agreement among modern linguists that the idea of an idea, the idea of the mental counterpart of a linguistic form, is worse than worthless for linguistic science. I think the behaviorists are right in holding that talk of ideas is bad business even for psychology. The evil of the idea idea is that its use, like the appeal in Molière to a virtus dormitiva, engenders an illusion of having explained something. And the illusion is increased by the fact that things wind up in a vague enough state to insure a certain stability, or freedom from further progress.

    Let us then look back to the lexicographer, supposed as he is to be concerned with meanings, and see what he is really trafficking in if not in mental entities. The answer is not far to seek: the lexicographer, like any linguist, studies linguistic forms. He differs from the so-called formal linguist only in that he is concerned to correlate linguistic forms with one another in his own special way, namely, synonyms with synonyms. The characteristic feature of semantical parts of linguistics, notably lexicography, comes to be not that there is an appeal to meanings but that there is a concern with synonymy.

    What happens in this maneuver is that we fix on one important context of the baffling word ‘meaning’, namely the context ‘alike in meaning’, and resolve to treat this whole context in the spirit of a single word ‘synonymous', thus not being tempted to seek meanings as intermediary entities. But, even supposing that the notion of synonymy can eventually be provided with a satisfactory criterion, still this maneuver only takes care of the one context of the word ‘meaning’—the context ‘alike in meaning’. Does the word also have other contexts that should concern linguists? Yes, there is certainly one more—the context ‘having meaning’. Here a parallel maneuver is in order: treat the context ‘having meaning’ in the spirit of a single word, ‘significant’, and continue to turn our backs on the supposititious entities called meanings.

    Significance is the trait with respect to which the subject matter of linguistics is studied by the grammarian. The grammarian catalogues short forms and works out the laws of their

    

page 49     MEANING IN LINGUISTICS     III,2

concatenation, and the end product of this is no more nor less than a specification of the class of all possible linguistic forms, simple and composite, of the language under investigation—the class of all significant sequences, if we accept a liberal standard of significance. The lexicographer, on the other hand, is concerned not with specifying the class of significant sequences for the given language, but rather with specifying the class of pairs of mutually synonymous sequences for the given language or, perhaps, pair of languages. The grammarian and the lexicographer are concerned with meaning to an equal degree, be it zero or otherwise; the grammarian wants to know what forms are significant, or have meaning, while the lexicographer wants to know what forms are synonymous, or alike in meaning. If it is urged that the grammarian’s notion of significant sequences should not be viewed as resting on a prior notion of meaning, I applaud; and I say the lexicographer’s notion of synonymy is entitled to the same compliment. What had been the problem of meaning boils down now to a pair of problems in which meaning is best not mentioned; one is the problem of making sense of the notion of significant sequence, and the other is the problem of making sense of the notion of synonymy. What I want to emphasize is that the lexicographer had no monopoly on the problem of meaning. The problem of significant sequence and the problem of synonymy are twin offspring of the problem of meaning.

2

    Let us suppose that our grammarian is at work on a hitherto unstudied language, and that his own contact with the language has been limited to his field work. As grammarian he is concerned to discover the bounds of the class K of significant sequences of the language. Synonymy correlations of members of K with English sequences and with one another are not his business; they are the business of the lexicographer.

    There is presumably no upper limit to the lengths of members of K. Moreover, parts of significant sequences count aa significant, down to the smallest adopted units of analysis; so such

    

page 50     MEANING IN LINGUISTICS     III, 2

units, whatever they are, are the shortest members ofK . Besides the length dimension, however, there is a dimension of thickness to consider. For, given two utterances of equal and arbitrary length and fairly similar acoustical make-up, we must know whether to count them as occurrences of two slightly different occurrences of one members of K or as two slightly different and the same member of K . The question of thickness is the question what acoustical differences to count as relevant and what ones to count merely as inconsequential idiosyncrasies of voice and accent.

    The question of thickness is settled by cataloguing the phonemes—the single sounds, distinguished as coarsely as possible for purposes of the language. Two subtly differing sounds count as the same phoneme unless it is possible, by putting one for the other in some utterance, to change the meaning of the utterance.² Now the notion of phoneme, thus formulated, depends obviously and notoriously on the notion of sameness of meaning, or synonymy. Our grammarian, if he is to remain pure grammarian and eschew lexicography, must carry out his program of delimiting K without the help of a notion of phoneme so defined.

    There seems indeed, at first glance, to be an easy way out: he can simply enumerate the phonemes needed for the particular language at hand, and dispense with the general notion of phoneme defined in terms of synonymy. This expedient would be quite admissible as a mere technical aid to solving the grammarian’s problem of specifying the membership of K , if the problem of specifying the membership of K could itself be posed without prior appeal to the general notion of phoneme. But the fact is otherwise. The class K which it is the grammarian’s empirical business to describe is a class of sequences of phonemes, and each phoneme is a class of brief events. (It will be convenient to swallow this much platonism for present purposes, though some logical maneuvers might serve to reduce it.) The grammarian’s problem is in part objectively set for him thus: every speech event which he encounters in his field work

    ²Cf. Bloch and Trager, pp. 38-52, or Bloomfield, pp. 74-92.

    

page 51     MEANING IN LINGUISTICS     III,2

counts as a sample of a member of K. But the delimiting of the several members of K, that is, the grouping of mutually resemblant acoustical histories into bundles of proper thickness to qualify as linguistic forms, needs also to have some objective significance if the task of the field grammarian is to be made sense of as an empirical and objective task at all. This need is fulfilled if the general notion of phoneme is at hand, as a general relative term: ‘x is a phoneme for language L’, with variable ‘x’and ‘L’, or ‘x is a phoneme for speaker s’, with variable ‘x' and ‘s’. Thereupon the grammarian’s business, with respect to a language L, can be stated as the business of finding what sequences of phonemes of L are significant for L. Statement of the grammarian’s purpose thus depends not only on ‘significant’, as we had been prepared to expect, but also on ‘phoneme’.

    But we might still seek to free grammar of dependence on the notion of synonymy, by somehow freeing the notion of phoneme itself of such dependence. It has been conjectured, for example, by Bühler, that this might in principle be accomplished. Let the continuum of sounds be arranged in acoustical or physiological order in one or more dimensions, say two, and plotted against frequency of occurrence, so that we come out with a three-dimensional relief map in which altitude represents frequency of occurrence. Then it is suggested that the major humps correspond to the phonemes. There are abundant reasons to suspect that neither this oversimplified account nor anything remotely resembling it can possibly provide an adequate definition of the phoneme; and phonologists have not neglected to adduce such reasons. As a means of isolating other points of comparison between grammar and lexicography, however, let us make the unrealistic assumption that our grammarian has some such nonsemantical definition of phoneme. Then his remaining task is to devise a recursive description of a class K of forms which will comprise all and only those sequences of phonemes which are in fact significant.

    The basic point of view is that the class K is objectively determinate before the grammatical research is begun; it is the class of the significant sequences, the sequences capable of

    

page 52     MEANING IN LINGUISTICS     III, 2

occurring in the normal stream of speech (supposing for the moment that this terminology is itself significant). But the grammarian wants to reproduce this same class in other terms, formal terms; he wants to devise, in terms of elaborate conditions of phoneme succession alone, a necessary and sufficient condition for membership in K. He is an empirical scientist, and his result will be right or wrong according as he reproduces that objectively predetermined class K or some other.

    Our grammarian’s attempted recursive specification of K will follow the orthodox line, we may suppose of listing “morphemes” and describing constructions. Morphemes, according to the books,³ are the significant forms which are not resoluble into shorter significant forms. They comprise affixes, word stems, and whole words in so far as these are not analyzable into subsidiary morphemes. But we can spare our grammarian any general problem of defining morpheme by allowing him simply to list his so-called morphemes exhaustively. They become simply a convenient segmentation of heard phoneme sequences, chopped out as convenient building blocks for his purpose. He frames his constructions in the simplest way that will enable him to generate all members of K from his morphemes, and he cuts his morphemes to allow for the simplest constructions. Morphemes, like higher units such as might be called words or free forms, may thus be viewed simply as intermediate stages in a process which, over all, is still describable as reproduction of K in terms of conditions of phoneme succession.

    There is no denying that the grammarian’s reproduction of K, as I have schematized it, is purely formal, that is, free of semantics. But the setting of the grammarian’s problem is quite another matter, for it turns on a prior notion of significant sequence, or possible normal utterance. Without this notion, or something to somewhat the same effect, we cannot say what the grammarian is trying to do—what he is trying to match in his formal reproduction of K—nor wherein the rightness or wrongness of his results might consist. We are thus squarely

    ³Bloch and Trager, p. 54; Bloomfield, pp. 161-168.

    

    

page 53     MEANING IN LINGUISTICS     III, 3

confronted with one of the twin offspring of the problem of meaning, namely, the problem of defining the general notion of significant sequence.

3

    It is not satisfactory to say that a significant sequence is simply any sequence of phonemes uttered by any of the Naturkinder of our grammarian's chosen valley. What are wanted as significant sequences include not just those uttered but also those which could be uttered without reactions suggesting bizarreness of idiom. The joker here is ‘could’; we cannot substitute ‘will’. The significant sequences, being subject to no length limit, are infinite in variety; whereas, from the dawn of the language under investigation to the time when it will have evolved to the point where our grammarian would disown it, only a finite sample of this infinite manifold will have been uttered.

    The desired class K of significant sequences is the culmination of a series of four classes of increasing magnitude, H, I, J, and K, as follows. H is the class of observed sequences, excluding any which are ruled inappropriate in the sense of being nonlinguistic or belonging to alien dialects. I is the class of all such observed sequences and all that ever will happen to be professionally observed, excluding again those which are ruled inappropriate. J is the class of all sequences ever occurring, now or in the past or future, within or without professional observation—excluding, again, only those which are ruled inappropriate. K, finally, is the infinite class of all those sequences, with exclusion of the inappropriate ones as usual, which could be uttered without bizarreness reactions. K is the class which the grammarian wants to approximate in his formal reconstruction, and K is more inclusive even than J, let alone H and I. Now the class H is a matter of finished record; the class I is, or could be, a matter of growing record; the class J goes beyond any record, but still has a certain common-sense reality; but not even this can very confidently be said of K, because of the ‘could’.

    I expect we must leave the ‘could’ unreduced. It has some

    

page 54     MEANING IN LINGUISTICS     III, 3

operational import, indeed, but only in a partial way. It does require our grammarian to bring into his formal reconstruction of K all of the actually observed cases, that is, all of H. Further, it commits him to the prediction that all cases to be observed in the future will conform, that is, all of I belongs in K . Further still, it commits him to the scientific hypothesis that all unobserved cases fall in this K , that is, all of J. Now what more does the ‘could’ cover? What is the rationale behind that infinite additional membership of K , over and above the finite part J? This vast supplementary force of ‘could’, in the present instance and elsewhere, is perhaps a vestige of Indo-European myth, fossilized in the subjunctive mood.

    What our grammarian does is evident enough. He frames his formal reconstruction of K along the grammatically simplest lines he can, compatibly with inclusion of H, plausibility of the predicted inclusion of I, plausibility of the hypothesis of inclusion of J, and plausibility, further, of the exclusion of all sequences which ever actually do bring bizarreness reactions. Our basis for saying what ‘could’ be generally consists, I suggest, in what is plus simplicity of the laws whereby we describe and extrapolate what is. I see no more objective way of construing the conditio irrealis.

     Concerning the notion of significant sequence, one of the two survivals of the notion of meaning, we have now observed the following. It is needed in setting the grammarian’s task. But it is describable, without appeal to meanings as such, as denoting any sequence which could be uttered in the society under consideration without reactions suggesting bizarreness of idiom. This notion of a reaction suggesting bizarreness of idiom would want some refinement eventually. A considerable problem of refinement is involved also in the preliminary putting aside of so-called nonlinguistic noises, as well as utterances in alien dialects. Also there is the general methodological problem, of a pretty philosophical kind, which is raised by the word ‘could’. This is a problem common to concept-building in most subjects (apart from logic and mathematics, where it happens

    

page 55     MEANING IN LINGUISTICS     III, 3

to be well cleared up); I have outlined one attitude toward it.

    We should also remind ourselves of the oversimplification which I made with regard to morphemes, when I treated them merely as convenient phoneme sequences which our grammarian specifies by enumeration in the course of his formal reconstruction of the class of significant sequences from the phonemes. This is unrealistic because it requires our grammarian to exhaust the vocabulary, instead of allowing him to leave certain open categories, comparable to our nouns and verbs, subject to enrichment ad libitum. Now if on the other hand we allow him some open morpheme categories, his reconstruction of the class K of significant sequences ceases to be a formal construction from phonemes; the most we can say for it is that it is a formal reconstruction from phonemes and his open morpheme categories. So the problem remains how he is going to characterize his open morpheme categories—since enumeration no longer serves. This gap must be watched for possible intrusion of an unanalyzed semantical element.

    I do not want to take leave of the topic of significant sequence withodt mentioning one curious further problem which the notion raises. I shall speak now of English rather than a hypothetical heathen tongue. Any nonsensical and thoroughly un-English string of sounds can occur within a perfectly intelligible English sentence, even a true one, if in effect we quote the nonsense and say in the rest of our sentence that the quoted matter is nonsense, or is not English, or consists of four syllables, or rimes with ‘Kalamazoo’, etc. If the whole inclusive sentence is to be called normal English speech, then the rubbish inside it has occurred in normal English speech, and we have thus lost the means of excluding any pronounceable sequence from the category of significant sequence. Thus we must either narrow our concept of normality to exclude, for present purposes, sentences which use quotation, or else we must narrow our concept of occurrence to exclude occurrence within quotation. In either event we have the problem of identifying the spoken analogue of quotation marks, and of doing so in general enough

    

page 56     MEANING IN LINGUISTICS     III, 4

terms so that our concept of significant sequence will not be limited in advance to some one preconceived language such as English.

    In any case we Ihave seen that the problem of significant sequence admits of considerable fragmentation; and this is one of the two aspects into which the problem of meaning seemed to resolve, namely, the aspect of the having of meaning. The fact that this aspect of the problem of meaning is in such halfway tolerable shape accounts, no doubt, for the tendency to think of grammar as a formal, nonsemantical part of linguistics. Let us turn now to the other and more forbidding aspect of the problem of meaning, that of likeness in meaning, or synonymy.

4

    A lexicographer may be concerned with synonymy between forms in one language and forms in another or, as in compiling a domestic dictionary, he may be concerned with synonymy between forms in the same language. It is an open question how satisfactorily the two cases can be subsumed under a single general formulation of the synonymy concept, for it is an open question whether the synonymy concept can be satisfactorily clarified for either case. Let us first limit our attention to synonymy within a language.

     So-called substitution criteria, or conditions of interchangeability, have in one form or another played central roles in modern grammar. For the synonymy problem of semantics such an approach seems more obvious still. However, the notion of the interchangeability of two linguistic forms makes sense only in so far as answers are provided to these two questions: (a) In just what sorts of contextual position, if not in all, are the two forms to be interchangeable? (b) The forms are to be interchangeable salvo quo? Supplanting one form by another in any context changes something, namely, form at least; and (b) asks what feature the interchange is to leave invariant. Alternative answers to (a) and (b) give alternative notions of interchangeability, some suited to defining grammatical correspondences and others, conceivably, to defining synonymy.

    

page 57     MEANING IN LINGUISTICS     III, 4

    In §3 of Essay II we tried answering (b), for purposes of synonymy, with veritate. We found that something had still to be done about (a), in view, for example, of the difficulty presented by quotation. So we answered (a), lamely appealing to a prior conception of “word.” Then we found that interchangeability salva veritate was too weak a condition for synonymy if the language as a whole was “extensional,” and that in other languages it was an unilluminating condition, involving something like a vicious circle.

    It is not clear that the problem of synonymy discussed in those pages is the same as the lexicographer’s problem. For in those pages we were concerned with “cognitive” synonymy, which abstracts from much that the lexicographer would want to preserve in his translations and paraphrases. Even the lexicographer is indeed ready to equate, as synonymous, many forms which differ perceptibly in imaginative associations and poetic value;₄ but the optimum sense of synonymy for his purpose is probably narrower than synonymy in the supposed cognitive sense. However this may be, certainly the negative findings which were summed up in the preceding paragraph carry over; the lexicographer cannot answer (b) with veritate. The interchangeability which he seeks in synonymy must not merely be such as to assure that true statements remain true, and false ones false, when synonyms are substituted within them; it must assure further that statements go over into statements with which they as wholes are somehow synonymous.

    This last observation does not recommend itself as a definition, because of its circularity: forms are synonymous when their interchange leaves their contexts synonymous. But it has the virtue of hinting that substitution is not the main point, and that what we need in the first place is some notion of synonymy for long segments of discourse. The hint is opportune; for, independently of the foregoing considerations, three reasons can be adduced for approaching the problem of synonymy from the point of view of long segments of discourse.

    First, any interchangeability criterion for synonymy of short

    ⁴See above, p. 28.

    

page 58     MEANING IN LINGUISTICS     III, 4

forms would obviously be limited to synonymy within a language; otherwise interchange would produce polyglot jumbles. Interlinguistic synonymy must be a relation, primarily, between segments of discourse which are long enough to bear consideration in abstraction from a containing context peculiar to one or the other particular language. I say “primarily” because interlinguistic synonymy might indeed be defined for the component forms afterward in some derivative way.

    Second, a retreat to longer segments tends to overcome the difficulty of ambiguity or homonymy. Homonymy gets in the way of the law that if a is synonymous with b and b with c, then a is synonymous with c. For, if b has two meanings (to revert to the ordinary parlance of meanings), a may be synonymous with b in one sense of b and b with c in the other sense of b. This difficulty is sometimes dealt with by treating an ambiguous form as, two forms, but this expedient has the drawback of making the concept of form depend on that of synonymy.

    Third, there is the circumstance that in glossing a world we have so frequently to content ourselves with a lame partial synonym plus stage directions. Thus in glossing ‘addled’ we say ‘spoiled’ and add ‘said of an egg’. This widespread circumstance reflects the fact that synonymy in the small is no primary concern of the lexicographer; lame synonyms plus stage directions are quite satisfactory in so far as they expedite his primary business of explaining how to translate or paraphrase long speeches. We may continue to characterize the lexicograplher’s domain squarely as synonymy, but only by recognizing synonymy as primarily a relation of sufficiently long segments of discourse.

    So we may view the lexicographer as interested, ultimately, only in cataloguing synonym pairs which are sequences of sufficient length to admit of synonymy in some primary sense. Naturally he cannot catalogue these true synonym pairs directly, in any exhaustive way, because they are altogether limitless in number and variety. His case is parallel to that of the grammarian, who for the same reason was unable to catalogue the significant sequences directly. The grammarian accomplished

    

page 59    MEANING IN LINGUISTICS   III, 4

his end indirectly, by fixing on a class of atomic units capable of enumeration and then propounding rules for compounding them to get all significant sequences. Similarly the lexicographer accomplishes his end indirectly, the end of specifying the infinitely numerous genuine pairs of long synonyms; and this he does by fixing on a class of short forms capable of enumeration and then explaining as systematically as he can how to construct genuine synonyms for all sufficiently long forms compounded of those short ones. These short forms are in effect the word entries in his glossary, and the explanations of how to construct genuine synonyms of all sufficiently long compounds are what appear as the glosses in his glossary, typically a mixture of quasi synonyms and stage directions.

    Thus the lexicographer’s actual activity, his glossing of short forms by appeal to quasi synonyms and stage directions, is not antithetical to his being concerned purely and simply with genuine synonymy on the part of forms sufficiently long to admit of genuine synonymy. Something like his actual activity is indeed the only possible way of cataloguing, in effect, the limitless class of pairs of genuinely synonymous longer forms.

    I exploited just now a parallelism between the grammarian’s indirect reconstruction of the limitless class of significant sequences and the lexicographer’s indirect reconstruction of the limitless class of genuine synonym pairs. This parallelism bears further exploiting. It brings out that the lexicographer’s reconstruction of the class of synonym pairs is just as formal in spirit as the grammarian’s reconstruction of the class of significant sequences. The invidious use of the word ‘formal’, to favor grammar as against lexicography, is thus misleading. Both the lexicographer and the grammarian would simply list the membership of the respective classes in which they are interested, were it not for the vastness, the infinitude even, of the numbers involved. On the other hand, just as the grammarian needs over and above his formal constructions a prior notion of significant sequence for the setting of his problem, so the lexicographer needs a prior notion of synonymy for the setting of his. In the setting of their problems, the grammarian and the

    

page 60   MEANING IN LINGUISTICS     III, 5

lexicographer draw equally on our heritage from the old notion of meaning.

    It is clear from the foregoing reflections that the notion of synonymy needed in the statement of the lexicographer’s problem is synonymy only as between sequences which are long enough to be pretty clean-cut about their synonymy connections. But in conclusion I want to stress what a baffling problem this remaining problem of synonymy, even relatively clean-cut and well-behaved synonymy, is.

    

5

    Synonymy of two forms is supposed vaguely to consist in an approximate likeness in the situations which evoke the two forms, and an approximate likeness in the effect of either form on the hearer. For simplicity let us forget this second requirement and concentrate on the first—the likeness of situations. What I have to say from here on will be so vague, at best, that this further inaccuracy will not much matter.

    As everyone is quick to point out, no two situations are quite alike; situations in which even the same form is uttered are unlike in myriad ways. What matters rather is likeness in relevant respects. Now the problem of finding the relevant respects is, if we think of the matter in a sufficiently oversimplified way, a problem typical of empirical science. We observe a speaker of Kalaba, say—to adopt Pike’s myth—and we look for correlations or so-called causal connections between the noises he makes and the other things that are observed to be happening. As in any empirical search for correlations or so-called causal connections, we guess at the relevance of one or another feature and then try by further observation, or even experiment, to confimrm or refute our hypothesis. Actually, in lexicography this guessing at possible relevances is expedited by our natural familiarity with the basic lines of human interest. Finally, having found fair evidence for correlating a given Kalaba sound sequence with a given combination of circumstances, we conjecture synnonymy of that sound sequence with another, in English, say, which is correlated with the same circumstances.

    

page 61    MEANING IN LINGUISTICS     III, 5

    As I unnecessarily remarked, this account is oversimplified. Now I want to stress one serious respect in which it is oversimplified: the relevant features of the situation issuing in a given Kalaba utterance are in large part concealed in the person of the speaker, where they were implanted by his earlier environment. This concealment is partly good, for our purposes, and partly bad. It is good in so far as it isolates the subject’s narrowly linguistic training. If we could assume that our Kalaba speaker and our English speaker, when observed in like external situations, differed only in how they say things and not in what they say, so to speak, then the methodology of synonymy determinations would be pretty smooth; the narrowly linguistic part of the causal complex, different for the two speakers, would be conveniently out of sight, while all the parts of the causal complex decisive of synonymy or heteronymy were open to observation. But of course the trouble is that not only the narrowly linguistic habits of vocabulary and syntax are imported by each speaker from his unknown past.

    The difficulty here is not just that those subjective components of the situation are hard to ferret out. This difficulty, if it were all, would make for practical uncertainty and frequent error in lexicographical pronouncements, but it would be irrelevant to the problem of a theoretical definition of synonymy irrelevant, that is, to the problem of coherently stating the lexicographer’s purpose. Theoretically the more important difficulty is that, as Cassirer and Whorf have stressed, there is in principle no separating language from the rest of the world, at least as conceived by the speaker. Basic differences in language are bound up, as likely as not, with differences in the way in which the speakers articulate the world itself into things and properties, time and space, elements, forces, spirits, and so on. It is not clear even in principle that it makes sense to think of words and syntax as varying from language to language while the content stays fixed; yet precisely this fiction is involved in speaking of synonymy, at least as between expressions of radically different languages.

    What provides the lexicographer with an entering wedge is

    

page 62    MEANING IN LINGUISTICS     III, 5

the fact that there are many basic features of men’s ways of conceptualizing their environment, of breaking the world down into things, which are common to all cultures. Every man is likely to see an apple or breadfruit or rabbit first and foremost as a unitary whole rather than as a congeries of smaller units or as a fragment of a larger environment, though from a sophisticated point of view all these attitudes are tenable. Every man will tend to segregate a mass of moving matter as a unit, separate from the static background, and to pay it particular attention. Again there are conspicuous phenomena of weather which one man may be expected to endow with much the same conceptual boundaries as another; and similarly perhaps for some basic internal states such as hunger. As long as we adhere to this presumably common fund of conceptualization, we can successfully proceed on the working assumption that our Kalaba speaker and our English speaker, observed in like external. situations, differ only in how they say things and not in what they say.

    The nature of this entering wedge into a strange lexicon encourages the misconception of meaning as reference, since words at this stage are construed, typically, by pointing to the object referred to. So it may not be amiss to remind ourselves that meaning is not reference even here. The referen.ce might be the Evening Star, to return to Frege’s example, and hence also the Morning Star, which is the same thing; but ‘Evening Star’ might nevertheless be a good translation and ‘Morning Star’ a bad one.

    I have suggested that our lexicographer’s obvious first moves in picking up some initial Kalaba vocabulary are at bottom a matter of exploiting the overlap of our cultures. From this nucleus he works outward, ever more fallibly and conjecturally, by a series of clues and hunches. Thus he begins with a fund of correlations of Kalaba sentences with English sentences at the level where our cultures meet. Most of these sentences classify conspicuously segregated objects. Then he breaks these Kakaba sentences down into short component elements, and makes

    

page 63    MEANING IN LINGUISTICS     III, 5

tentative English translations of these elements, compatible with his initial sentence translations. On this basis, he frames hypotheses as to the English translations of new combinations of those elements—combinations which as wholes have not been translated in the direct way. He tests his hypotheses as best he can by making further observations and keeping an eye out for conflicts. But, as the sentences undergoing translation get further and further from mere reports of common observations, the clarity of any possible conflict decreases; the lexicographer comes to depend increasingly on a projection of himself, with his Indo-European Weltanschauung, into the sandals of his Kalaba informant. He comes also to turn increasingly to that last refuge of all scientists, the appeal to internal simplicity of his growing system.

    The finished lexicon is a case, evidently, of ex pede Herculem. But there is a difference. In projecting Hercules from the foot we risk error, but we may derive comfort from the fact that there is something to be wrong about. In the case of the lexicon, pending some definition of synonymy, we have no statement of the problem; we have nothing for the lexicographer to be right or wrong about.

    Quite possibly the ultimately fruitful notion of synonymy will be one of degree: not the dyadic relation of a as synonymous with b, but the tetradic relation of a as more synonymous with b than c with d. But to classify the notion as a matter of degree is not to explain it; we shall still want a criterion or at least a definition for our tetradic relation. The big difliculty to be surmounted in devising a definition, whether of a dyadic relation of absolute synonymy or a tetradic relation of comparative synonymy, is the difficulty of making up our minds as to just what we are trying to do when we translate a Kalaba statement which is not a mere report on fairly directly observable features of the surrounding situation.

    The other branch of the problem of meaning, namely the problem of defining significant sequence, led us into a contrary-to-fact conditional: a significant sequence is one that could be

    

page 64    MEANING IN LINGUISTICS     III, 5

uttered without such and such adverse reactions. I urged that the operational content of this ‘could’ is incomplete, leaving scope for free supplementary determinations of a grammatical theory in the light of simplicity considerations. But we are well schooled in acquiescing in contrary-to-fact conditionals. In the case of synonymy the tyranny of the developing system, the paucity of explicit objective controls, is more conspicuous.

    

    

Chapters IV to IX, Bibliography, Index etc. are only available as photos of the pages. OCR etc. still underway.

CONTENTS
CHAPTER		PAGE
	Preface	i-viii
I.	On what there is	1
II.	Two dogmas of empiricism	20
III.	The problem of meaning in linguistics	65
Photos of original pages only from Ch.IV to end.
IV.	Identity, ostension, and hypostasis	65
V.	New foundations for mathematical logic	80
VI.	Logic and the reification of universals	102
VII.	Notes on the theory of reference	130
VIII.	Reference and modality	139
IX.	Meaning and existential inference	160
	Origin of the essays	170
	Bibliographical references	171
	Index	179