The Case Against Quine's Case for Psychologism

“Whoever remains caught up within the sphere of general reflections may allow himself to be deceived by the psychologistic arguments. A mere glance at any one logical Law, at its true meaning and at the insightfulness with which it is grasped in its own right as truth, necessarily puts an and to the delusion.”

E. Husserl¹

“I would almost say that psychologism lives on inconsequence. Whoever logically thinks it out to the end will have already abandoned it when he gets there.”
E. Husserl²

During the late 1950s and early 1960s, to make any significant association of logic with psychology was a dreadful faux pas, one almost utterly beyond redemption. It drew scorn and pity upon you. Although Dewey’s influence was not totally absent, extraordinary effort was required for dominant logicians such as Rudolph Carnap merely to be civil toward his views and treatments of (what Dewey took to be) logical matters. Defining psychologism in logic as essentially a confusion of the objective (in this case, logical relations) with the subjective (in this case, the psychological), Carnap exempted Dewey from it on the sole grounds that he consistently dealt with subjective or mentalistic processes such as ‘thinking’ or ‘inquiry’.³

The central connection here for an understanding of the fate of psychologism in contemporary philosophy is that between the subjective and the mental. It was Frege’s interpretation of the psychological or mental as the essentially subjective, as the ‘mentalistic’ in a certain opprobrious sense, that overshadowed philosophical logic in the middle quarters of the twentieth century and made Carnap’s view just referred to, for example, the obvious position to take. For Frege, ‘psychologism’ meant the misinterpretation of concepts and propositions proper to semantics and logic in such a way that they (supposedly) came to apply to ‘mental’ events and entities understood in his special way. Michael Dummett comments: “When Frege engages in polemic against psychologism, what he is concerned to repudiate is the invasion of the theory of meaning by notions concerned with mental processes, mental images, and the like, . . . . The psychological <287> was for him a realm of incommunicable inner experience . . . .“⁴ John Skorupski remarks in this same vein of interpretation: “ ‘Psychologism’ may be the view that laws of logic are, or hold in virtue of, the laws which govern our mental processes, or again it may be the view that ‘meanings’ are mental entities. Frege was clearly opposed to both, and thought they were bound up with idealism . . . .“⁵ I do not suggest that this is an adequate interpretation of Frege on psychologism, but it is a view of meaning and the mental that has come to be associated with him and handed on in current discussions.

The general attitude toward the relationship of logic and psychology has been radically transformed within the last decade or two. Now it is quite common to find logical laws described as laws of cognitive psychology or as aspects of the semantics of natural language or some ‘deep’ grammar. The return to psychologisms of these sorts all share the view that the mental, as they now interpret it, no longer falls under the reproach of subjectivism or ‘mentalism’ which was raised by Frege. Having supposedly come to terms with Frege through a repositioning of the ‘psychological’, it is thought to be safe to return logical laws to a central position in the actual processes of human life--where they in some sense obviously belong. This certainly is an inherently desirable shift. One of the most serious objections to anti-psychologistic treatments of the laws of logic has always been that such treatments cannot do justice to the practical use to which logical laws are actually put, if not in the direction, then at least in the critique of everyday thinking, talking and writing.⁶

It is unfortunate, in the light of the desirability of an appropriate reunion between logic and psychology, that a mere non-‘mentatistic’ interpretation of the mental (or, as would now more likely be said, of the ‘linguistic’, regarded as encompassing whatever of the mental should be taken seriously by philosophers) does not in fact deal with fundamental problems facing all attempted returns to psychologistic interpretations of logical laws. Frege and Husserl are often grouped together as critics of psychologism.⁷ Although there is considerable overlap between them, they turn out upon examination to make quite different points--in emphasis at least. The easy return to psychologism which has occurred in the recent past has disregarded the substance of Husserl’s critique of psychologism. That critique really had little to do with the ‘subjectification’ of logical law as Frege understood it. Instead, it focused upon the nature of the evidence for logical laws, and rested upon the distinction between the real (the individual) and the ideal (the universal). Insofar as the new psychologism interprets the laws of logic as laws of the real, and possibly as ‘naturalistic’, it remains subject to his critique, indifferently of whether the real or ‘natural’ events in question are opprobriously ‘mentalistic’ or not. If his arguments were good in the first decade of this century, they are still good today. I believe that they were and that they are.

We now turn to consider how those arguments apply to the position of W.V. Quine, who has strongly influenced the contemporary return to psychologism. On September 9, 1968, at the Fourteenth International Congress of Philosophy in Vienna, Quine read a lecture entitled, “Epistemology Naturalized: or, The Case for Psychologism.” Later, stripped of its sub-title, it was published as chapter 3 of his Ontological Relativity and Other Essays,⁸ and became a focal point of subsequent discussions concerning the nature of epistemology and logic. It is fair to say that it substantially contributed to the monumental change of attitude that has since taken place with regard to the relationship between logic and psychology. In what follows I will attempt to show that Quine’s case for psychologism depends entirely upon neglecting the details of what the working logician, including Quine himself, actually does in establishing the truths--‘the Laws of Logic’--for which logic as an intellectual discipline is primarily responsible. It is such neglect and its implications that originally opened the way for Husserl’s critique and that sustains its relevance today.

What exactly was the main point of Husserl’s critique as it concerns the new psychologism? (Here I will be very concise, referring for a more comprehensive account to pp. 143-166 of my Logic and the Objectivity of Knowledge). Addressing, in particular, thinkers such as John Stuart Mill and Theodore Lipps, who adopted the position that logic is a branch of psychology, Husserl raises the issue of the evidential base of the laws of logic. If logic is a subsection of psychology, then the evidence for those laws must be drawn from the psychological realm. What then are the psychological events, facts and regularities to be invoked in showing that the laws are true? Surveying the actual literature of logic, where is there so much as a single demonstration of a single logical law in a single logic text or treatise that, in the familiar manner of demonstrations in logic, rests the truth of the law upon psychological facts and regularities? It was Husserl’s view that the universal absence of such demonstrations of logical laws points to a rigorous separation of logic from psychology. The evidence supporting the familiar logical laws taught in standard courses in formal logic is complete.⁹ We are not waiting for further evidence on the Distribution Laws, for example. But this quite conclusive evidence turns out, in addition, to be completely free of statements of empirical laws or facts of any kind, including the psychological. While numerous thinkers have held the laws of logic to be psychological, their contention has no actual bearing upon the proofs of logical laws carried out by working logicians. It is therefore a mere general prejudice with no foundation in the actual conduct of logic as a field of research and teaching. Anyone is, of course, invited to overturn this claim of general prejudice by simply producing one professionally acceptable proof of an unquestionable law of logic from psychological facts interpreted in any way one wishes. Certainly no such proof is to be found in Quine’s writings.

But what, then, is Quine’s case for psychologism? It chiefly consists in despair of anything other than psychology to guide us in understanding how evidence works in providing us with our view of reality. It is “the bankruptcy of epistemology” conceived along the lines of logical reconstruction.¹⁰ It is “The impossibility of strictly deriving the science of the external world from sensory evidence.“¹¹ Quine points out (what is generally conceded) that Carnap’s attempts at rational reconstruction of physicalistic discourse in terms of sense experience, logic, and set theory, even if successful as an account of meaning, would not enable us to prove sentences about physical objects from observation sentences (plus set theory) by acceptable logic. The move to psychologism in epistemology and logic is a response to this failure:

But why all this creative reconstruction, all this make believe? The stimulation of his sensory receptors is all the evidence anybody has had to go on, ultimately, in arriving at his picture of the world. Why not just see how this construction really proceeds? Why not settle for psychology?¹²

The project of rational reconstruction was in fact doomed, Quine holds, by the indeterminacy of translation: “The typical statement about bodies has no fund of experiental implications it can call its own.“¹³ All sentences incorporate a substantial mass of theory, allowing for manifold adaptations to their ‘subject matter’, so that “...there can be no ground for saying which of two glaringly unlike translations of individual sentences is right.“¹⁴“The indeterminacy of translation of theoretical sentences is the natural conclusion.... This conclusion..., seals the fate of any general notion of propositional meaning or, for that matter, state of affairs.“¹⁵ Therewith it seals the fate of logic as a theory of the structure and interrelationships of such propositional meanings, requiring that a different subject matter be settled upon for logic itself.

What is left with such meanings gone? ‘Empirical’ meaning: the linguistic interactions of human organisms with factors present in the public environment of which they are themselves a part. The child:

... learns his first words and sentences by hearing and using them in the presence of appropriate stimuli. These must be external stimuli, for they must act both on the child and on the speaker from whom he is learning. Language is socially inclucated and controlled; the inculcation and control turn strictly on the keying of sentences to shared stimulation. Internal factors may vary ad libitum without prejudice to communication as long as the keying of language to external stimuli is undisturbed.¹⁶

Laws of evidence must show up in the same public arena. Epistemology is ‘naturalized’, and thus (supposedly) saved from total bankruptcy, when psychological laws of stimulus and response are taken to be the only possible account of “...how evidence relates to theory, and in what ways one’s theory of nature transcends any available evidence.“¹⁷“Epistemology, or something like it, simply falls into place as a chapter of psychology and hence of natural science. It studies a natural phenomena, viz. a physical human subject...“¹⁸--specifically, the relation between the human being’s linguistic (including scientific) output and its environmental stimuli. “We are after an understanding of science as an institution or process in the world, and we do not intend that that understanding be any better than the science which is its object.“¹⁹

This completes Quine’s abstract line of reasoning in support of psychologism. Although he makes a few comments about analyticity in “Epistemology Naturalized,“details of a psychological interpretation of evidence (and the laws of logic) are lacking there. Other works offer some help on the specifics of his view. The sub-section of Word and Object²⁰ on “The Interanimaton of Sentences,” for example, considers a situation where someone mixes the contents of two test tubes, notes a green tint in the compound, and emits the sentence: “There was copper in it.“²¹ In this case chemical theory has intervened between the observation of color and the sentence emitted. The intervening theory is a “verbal network .... composed of sentences associated with one another in multifarious ways,” causally and logically. But whether causally or logically:

... any such interconnections of sentences must finally be due to the conditioning of sentences as responses to sentences as stimuli. If some of the connections count more particularly as logical or as causal, they do so only by reference to so-called logical or causal laws which in turn are sentences within the theory. The theory as a whole--a chapter of chemistry, in this case, plus relevant adjuncts from logic and elsewhere--is a fabric of sentences variously associated to one another and to nonverbal stimuli by the mechanism of conditioned response.²²

This helps a bit with the details. It leaves no doubt, in any case, that Quine intends to interpret the evidentiating or logical relations between sentences in terms of the psychological process of stimulus and response, and that logical laws will simply be one element (indeed, as themselves sentences) in that process. A few pages later he comments that “...any realistic theory of evidence must be inseparable from the psychology of stimulus and response, applied to sentences.“²³ We take this to mean that the fundamental logical relations, such as implication and contradiction, are matters of social conditioning, and that to say A implies B. for example, is to make a statement about the conditioned responses of the relevant language community with respect to the sentences named by ‘A’ and ‘B.’

The later book, Roots of Reference,²⁴ makes this even clearer, and provides a discussion of a specific logical law: the one usually called ‘Addition’. This is, so far as I can tell, the only logical law of which Quine even attempts a specifically psychologistic interpretation, which surely indicates the great distance from ‘the things themselves’ he maintains in his philosophical speculations about logic. He remarks: “The law that an alternation is implied by its components is thus learned, we might say, with the word ‘or’ itself, and similarly with the other laws.“²⁵ This clearly applies what is said here about alternation to the laws of logic generally: laws such as the familiar Modus Tollens, the Distributive Laws, and Universal Instantiation. He takes care to restate his opposition to that linguistic interpretation of the laws of logic according to which they rest solely on the meanings of the logical words, ‘or’, ‘not’, ‘all’, etc. “But now,” he asks, “can we perhaps find some sense for the doctrine [just opposed]...in terms of the learning process?” Not surprizingly, the answer is affirmative: “We learn the truth functions,...by finding connections of dispositions: e.g. that people are disposed to assent to an alternation when disposed to assent to a component.“²⁶ From this, then, he proceeds to the general statement about the laws of logic quoted above. The laws of logic are laws of tendency to assert or deny, which (with a linguistic twist, of course) fits the well-known psychologistic pattern of interpretation.

Now the crucial point for our discussion is the link postulated by Quine between meaning (as communalized behavioral dispositions with regard to sentences) and truth. For the laws of logic are in some indispensable sense laws of truth. The linkage is, on his view, embedded in the learning process. “We learn to understand and use and create sentences only by learning conditions for the truth of such sentences.“²⁷ The earliest language learning, that of observation sentences, “is simply a matter of learning the circumstances in which those sentences count as true.“²⁸’Eternal’ sentences are, of course, learned differently, since their truth values do not alter with circumstances.

Truth in these discussions becomes a social concept. Learning a sentence, whether an observation sentence, a logical law, or an eternal sentence, is strictly a matter of developing a disposition to affirm or deny it in harmony with the linguistic community in its physical and social environment. The following comments about a particular eternal sentence makes this ‘social’ concept of truth clear. At least it makes clear that assent or ‘counting as trite’ is what we learn in learning a language, and is what the logical laws mediate:

Still, the learning of ‘A dog is an animal’ as I represented it consisted in learning to assent to it, and this hinged upon the truth of the sentence. It hinged anyway on our having learned to assent to ‘dog’ only in circumstances in which we learned also to assent to ‘animal’. If we learned to use and understand ‘A dog is an animal’ in the way I described then we learned at the same time to assent to it, or account it true.²⁹

Thus the foundation is laid for a definitive statement about analyticity, incorporating Quine’s reinterpretation of ‘the linguistic theory of logic’. Language is social, and since analyticity is truth grounded in language it will be social as well: “A sentence is analytic if everybody learns that it is true by learning its words. Analyticity, like observationality, hinges on social uniformity.“³⁰ Quine thus makes “...analyticity hinge... on a community-wide uniformity in the learning of certain words.“³¹

We must then be clear that, for the purposes of this discussion, Quine is not distinguishing between being true and being communally counted or accepted as true. In particular, learning a logical law is not a conditioned response to truth as something distinct from acceptance within the linguistic community. There are for him no meanings, no propositions, which might be bearers of truth regardless of acceptance or rejection. The ‘necessity’ which the laws of mathematics and logic seem to enjoy merely amounts to the psycho-social tendency toward conservation with regard to our belief system as a whole: a preference to modify our acceptances and rejections only along lines that will least disturb it as a whole.³²

There is very little more than this to Quine’s discussions of how the laws of logic are psychological laws. The thinness of the account of how those laws may be so understood is due to the fact that in the works cited above he is coming to them von oben, from his speculative theory of language and mind, not von unten, from an examination of those laws themselves. When dealing with the laws themselves, on the other hand, as is done in writings which engage with logic as a field of scientific research, a quite different picture of logical law emerges. When one examines any of Quine’s expositions or proofs of logical laws, one finds that they are presented as theoretically complete without a single reference to, or invocation of support from, any psychological matter of fact. We will make just a few comments about his actual procedure in the process of logical research and exposition, since his writings in this field are so well known.

Those who have used Quine’s Methods of Logic will recall that the primary subjects of discussion and theoretical analysis in that work are what he calls ‘schemata’. “Schemata are logical diagrams of statements.“³³ That is, they are symbols used to pick out various kinds of statements in terms of features determinative of the logical properties and relationships of those statements. “Schemata,” he tells us, “are the medium of our technical work”--although our results are to be applied to sentences.³⁴ The ‘results’ of the technical work will be primarily that given schemata are or are not ‘valid’. “A truth-functional schema is called consistent if it comes out true under some interpretations of its letters; otherwise inconsistent. A truth-functional schema is called valid if it comes out true under every interpretation of its letters.“³⁵“One truth-functional schema is said to imply another if there is no way of so interpreting the letters as to make the first schema true and the second false.“³⁶ Quantificational schemata submit to a similar treatment. “We can define an open [quantificational] schema as valid when, under every nonempty choice of universe U, and all interpretations of ‘Fx’, ‘Gx’, etc. within U, the schema comes out true of every object in U.“³⁷ The argumentation for the validity or non-validity of truth-functional or quantificational schemata accordingly is solely a matter of examining certain possibilities of interpretation of the schemata in question.

This pattern of argumentation is still followed in arguing for the validity of the schemata underlying rules for manipulating quantifiers. For example, with the schema underlying the rule of Universal Instantiation:

(1). (y)[(x)Fx g Fy],

Quine comments:

To see that (1) is valid, consider any universe and any interpretation, within that universe. of ‘F’. Case 1: ‘F’ is interpreted as true of everything in the universe. Then, for each object y in the universe, ‘(x)Fx g Fy’ becomes true because of true consequent; so (1) comes out true. Case 2: ‘F’ is interpreted otherwise. Then for each object y in the universe, ‘(x)Fx g Fy’ becomes true because of false antecedent; so (1) comes out true.³⁸

The typical language in such analyses, ‘consider any universe and any interpretation of...’, is of course the heart of the matter. It is precisely such considerations of universes and interpretations that here yields insight into the validity of logical laws. Our point in relation to the attempted psychologization of those laws can be put very simply now. What is considered in the proof of a logical law has absolutely nothing to do with psychological events, facts or regularities, however the psychological is interpreted. The precise character of those things--universes, interpretations, etc.--considered in exposition and proof of logical laws need not be cleared up here. It is enough for our present purposes that logical laws are proven, and that the premises of those proofs do not state or presuppose psychological matters of any kind. The evidence for them is complete, as cited by Quine himself, but it is not psychological. What has been said here with reference to Methods of Logic, probably his most widely used book on logic proper, is equally true of his other logical treatises. In no case do we find theorems proved from psychological premises. We would, for completeness sake, need to add some comments about principles of substitution and of derivations generally. But such additions would change nothing for the point here at issue.

Thus while, in his speculative writings , Quine suggests, for example, that the validity of the schema ‘p g (p v r)’ might rest upon certain facts pertaining to how ‘or’ is learned and used, he never offers such considerations when he is doing research and exposition in the discipline of logic, as distinct from speculating about it. Rather, in logical work his practice completely conforms to what Husserl said about the science of pure logic. Namely, that it is a theoretical discipline which stands entirely on its own feet,³⁹ and that “logical laws and forms in the pregnant sense of these words, belong to a theoretically closed round of abstract truth, that cannot in any way be fitted into previously delimited theoretical disciplines.“⁴⁰

Thus the really substantial point here does not concern psychology only, or the reduction of logical laws to psychological regularities. Given any domain of objects whatever, the claim that the laws of logic are laws of that domain imposes the task of proving the laws of logic from statements about the objects of that domain. (Alternative strategies, such as that of treating those laws as ‘rules’, will, I believe, invariably generate parallel problems. The justification of ‘rules’, for example, or the explanation of how they accomplish what they do, can only be carried out by reference to some domain of entities, including the rules themselves, what they ‘govern’, and whatever ‘governing’ itself amounts to.) Moreover, any ontology places rigorous restraints upon the possible interpretation of logical law. If, for example, one is a ‘Naturalist’, the problem of how to fit the laws of logic into that framework becomes urgent. A Physicalist, for example, owes us an account of how Modus Tollens and the Barbara syllogism are to be derived from the facts and laws proper to Physics.

With specific reference to Quine’s “Case for Psychologism,” we of course do not wish to deny that there are psychological laws of linguistic assent and denial. Those laws would find their evidence in behavioral or other psychological facts. But they and their accompanying facts do in fact never serve as the basis for laws of logic or count as evidence to be expounded within the science of logic. Nor is it even clear how they could. From within the discipline of logic, they present themselves as simply irrelevant. This is acknowledged by Quine’s practice, as well as by the practice of the logic profession generally. Husserl’s case against psychologism as an interpretation of the laws of logic is as strong today as ever. It will remain so until someone demonstrates, in a professionally acceptable way, at least one law of logic from psychological premises.