SOURCE: "Dewey's Epistemology and Metaphysics," in The Philosophy of John Dewey, Northwestern University, 1939, pp. 195-225.

[In the following essay, Murphy declares that Dewey's philosophical methods are unsound because they do not adopt scientific or practical methodology.]

In harmony with Spinoza's observation that Peter's idea of Paul is likely to give us a better notion of Peter than of Paul, it will doubtless be observed that the essays in this volume reflect the preconceptions and interpretative limitations of their authors at least as much as the actual content and implications of Mr. Dewey's philosophy. The danger in such cases is that what is presented as a critical analysis will in fact amount to little more than a translation of what, from the standpoint of an opposing philosophy of questionable validity, Mr. Dewey really must have meant and ought to have said. This risk is not altogether avoidable, for one can only criticize what he takes to be confused and inadequate by reference to clarity and adequacy as he sees them. It can, however, be minimized if the critic states at the outset the standpoint from which his analysis is to be made and the interest that directs it. The reader should then be in a position to consider the interpretation offered explicitly as an hypothesis, to be tested by its success in clarifying a theory which he has, on his own account, been trying to understand.

In the account here presented of Mr. Dewey's theory of knowledge and of nature, primary importance is attached to his insistence on the interpretation and criticism of statements made in terms of their use and testable validity in the contexts in which, prior to either epistemological or metaphysical analysis, they have a discoverable use and meaning. I shall term this reference of ideas to and their testing in specific situations "contextual analysis." The name does not particularly matter, but the procedure it calls attention to does, and the name suggested will serve, I think, as well as any other. It appears to me that a contextual analysis and testing of the ideas used in philosophical discussion is the indispensable basis for any adequate theory of either knowledge or nature and that Mr. Dewey in insisting on this fact and sometimes in carrying out the analysis or criticism required, has contributed very substan tially to the progress of inquiry in these fields. I propose, therefore, in the first place, to indicate what this method is and what, by its consistent use, we could reasonably expect to find out about knowledge and its place and function in the natural world.

It will be apparent, however, on further investigation, that the results he has actually reached are in many respects not consistent with the theory thus suggested. The non-philosophical inquirer who studies Dewey's theory of "inquiry" in the Logic or of "experience" in Experience and Nature will not find what, on the basis of the prospectus offered, he may well feel he had a right to expect. Instead he will discover that the context in terms of which Mr. Dewey interprets knowledge is one not appropriate to the actual procedures and claims of scientific or practical inquiry, and that his theory of nature is compromised and confused by this fact. It becomes necessary, therefore, to discriminate between those elements in the total theory which can be justified by the method recommended and those which tend to impede its satisfactory use. I believe that this discrimination can be made, that the complications and inconsistencies can then be accounted for by reference to the conditions under which Mr. Dewey's theory developed and the controversies in which it was involved, and that, finally, it can be shown that their elimination leaves the method originally outlined more clearly and reliably usable than it has so far been.

The interpretation thus presented is not offered as an account of what Mr. Dewey really meant, or as a rival theory which might increase its prestige by disposing argumentatively of so formidable a contender for epistemological or metaphysical supremacy. I have not been able to understand Mr. Dewey's philosophy as having any single clear or unequivocal meaning, and I do not suggest that the elements I have selected for favorable attention are any more "really" or "ultimately" what the author had in mind than much else that seems inconsistent with them. And it is certainly not my purpose to "refute" Mr. Dewey. On the contrary, I have learned so much from him that any criticism made will be based, in large part at least, on what, from his own writings, I have come to believe that philosophy ought to be and can be if philosophers take pains to know what they are talking about and to test their theories by reference to situations in which they have a testable meaning. It is in the interest of that sort of philosophy that I have here tried to discriminate those factors in Mr. Dewey's own theories which are permanently useful from those which, as it seems to me, have tended to impede its development.

I

I shall first discuss Mr. Dewey's theory of knowledge since, as will be shown later, it is only in terms of it that the more puzzling features of his empirical metaphysics are to be understood. And here some examination of the use of the term "epistemology" is at once required. In a common philosophical usage "epistemology" simply means "theory of knowledge" or "philosophical analysis of the nature of knowing and the meaning and criteria of truth." In this sense, Mr. Dewey not only has an "epistemology" but gives it a quite fundamental place in his philosophy. He himself, however, has habitually reserved the word "epistemology" as the designation for a particular sort of theory of knowledge of which he disapproves. This has led to some confusion, and has in particular left him open to the specious objection that he has not been able to avoid "epistemology," since he, too, has a theory about knowledge.

The difficulty is easily resolved. There are, on Dewey's view, two main sorts of philosophical accounts of knowledge, the epistemological and the contextual. And he has said quite plainly what the essential difference between these is. For him, as he explains in the first chapter of the Logic, "knowledge" has no meaning independent of inquiry so that "that which satisfactorily terminates inquiry, is by definition, knowledge," while for theories of the type to which he objects "knowledge is supposed to have a meaning of its own apart from connection with and reference to inquiry. The theory of inquiry is then necessarily subordinated to this meaning as a fixed, external end. The opposition between the two views is basic."¹

This distinction seems to me quite fundamental. Any student of the theory of knowledge is acquainted with the interminable and inconclusive controversies which arise when the disputants start with incompatible, but presumably quite evident, notions of what knowing must really be and then proceed to test the validity of all knowledgeclaims by reference to or derivation from what is thus really known. If the datum in knowledge must always be other than the "real" object which is the true objective of knowing, since all knowing involves transcendence, then of course all knowing is indirect and we are "epistemological dualists." If, on the other hand, it is quite evident to us that real knowing must be a grasp of being itself as this is directly "present to the mind," then epistemological monism, according to which the real object is itself the given, is the doctrine to be accepted. Such theories are admirably suited to dialectical elaboration and defense against opponents, since the partisan of either view has only to assume the validity of his own definition of knowing in his criticism of his opponent in order to show that the theory criticized either denies the possibility of "knowledge," so defined, and thus reduces to scepticism or else, when its "real" meaning is seen, reduces to his own theory and thus reluctantly and against its will testifies to the essential truth. Since this procedure is open to each of the contestants, if he is sufficiently tenacious in his insistence on his original stipulation, the controversy is in principle endless. Its defect is that it has, as experience has shown, very little connection with what reliable knowing shows itself to be outside the limits and stipulations of the debate, and in terms of the methods by means of which, in the sciences and in practical life, grounded knowledge is distinguished from unsubstantial and unsubstantiated opinion. Hence if we are to get any light on the nature of knowing as it operates in these cases, and to provide the sort of philosophical clarification that is urgently needed of some of the notions that are involved in it, we shall have to turn our attention away from the debate and find out in the first place what such knowing is, what the claims made for it mean in use, and how they are tested in the context of their primary and reliable application. And since it is by inquiry and investigation that most reliable knowledge is acquired, it is to inquiry that we must go to find out what it is.

"Epistemology," in Dewey's usage, is a designation for theories of knowledge that neglect this essential reference.

Theories of knowledge that constitute what are now called epistemologies have arisen because knowledge and obtaining knowledge have not been conceived in terms of the operations by which, in the continuum of experiential inquiry, stable beliefs are progressively obtained and utilized. Because they are not constructed upon the ground of operations and conceived in terms of their actual procedures and consequences, they are necessarily formed in terms of preconceptions derived from various sources, mainly cosmological in ancient and mainly psychological (directly or indirectly) in modern theory.²

The alternative proposed is simply, in the first instance, a reference back to the specific situations in which a difference between true and false beliefs can reliably be made out, and a careful account of what, in these situations, the meaning and validity of various truth claims is found to be. We shall only then be in a position to indulge in philosophical criticism or synthesis.

I shall follow Dewey, in this paper, in describing theories of the first type as "epistemological" and shall designate as "contextual" those that belong in the second group. I do not want to suggest, as Dewey seems at times to do, that all accounts of knowledge prior to his own have been primarily of the epistemological type. There is much careful observation and substantial wisdom in Locke's theory of knowledge, in Spinoza's, and in those of Aristotle and Aquinas, and much that the most up to date devotee of "contextualism" might with much profit learn from them. There is, however, a persistent tendency in theories of knowledge to lose touch with the situations in which knowledge-getting occurs and to substitute a dialectical elaboration of the "real" nature of knowledge for a contextual examination of the specific manner of functioning and meaning in use of the processes and claims of inquiry. In the period during which Mr. Dewey's own philosophy was developed this tendency reached an unhappy maximum in idealistic theories of "thought" and its relation to "reality" and in the great debate among the realists about the real or true object of knowledge and the essential nature of the mind's relation to it. To have insisted on the essential sterility of these theories and the need for philosophy to get into touch once more with the facts of knowledge-getting and testing was a needed contribution to the subject and one that Mr. Dewey has made in quite decisive fashion.

Stated summarily, the standpoint for a philosophical analysis of knowledge by reference to the contexts in which various knowledge-claims have meaning and testable validity in use is substantially this: (1) The philosopher has no special access to the nature of knowledge or of reality, or to the circumstances of their relation to each other. The knowledge-claims that it is his business to examine, understand, and unify so far as possible into a comprehensible whole, occur in the first instance in nonphilosophical activities and are to be understood by reference to their rôle in such activities. (2) These activities are various and are concerned not with reality as such or as a whole, but with those aspects or features of the world which are relevant to the particular requirements of the activity in question. (3) A reference to the circumstances under which, in the course of such activities, truth-claims are made and tested, is essential for an understanding of their meaning. To suppose that the statements in which the results of inquiry are summed up can retain their meaning when they are used for other purposes and in other contexts than those of their primary application is quite unwarranted, and leads to confusing results, as when, for instance, the validity of common sense statements is philosophically "criticized" as though they were claims to a sort of knowledge with which, in their ordinary and reliable use, they are not at all concerned, or scientific knowledge is treated as a highly inadequate attempt to characterize reality as a whole and is thus, in its own nature, essentially defective. (4) It is, none the less, philosophically legitimate and important to discover how far what can be found out by any one method of inquiry is relevant to what, on other grounds, we have reason to believe, and especially to our beliefs about what, in the sort of world we live in, is humanly possible and desirable. If "metaphysics" were simply an inclusive and necessarily provisional estimate of the place and prospects of human experience and aspirations within the natural world, then an empirical metaphysics would be a meaningful and useful investigation.

In some aspects of his philosophy, celebrated elsewhere in this volume, Mr. Dewey has given us this sort of theory. In the theory of knowledge, however, his position is much less clear. To see its significance we had best turn directly to his latest and most comprehensive account of the nature of knowledge, in the Logic.

Knowledge, as already noted, is to be understood as the appropriate outcome of inquiry. What, then, is inquiry? The answer to this question does not take us, as it should, to such specific sorts of inquiry as serve in practice as our means of finding out about the "antecedent" environment or the consequences of human behavior in it. It refers us instead to a theory about the rôle of ideas as instruments to be used in so altering a present indeterminate situation that an enjoyed future experience, itself non-cognitive but worth while on its own account, will reliably ensue, through the use of procedures which have proved their instrumental value in this capacity. The ultimate objective of knowing is held in all cases to be such an existential transformation of the subject-matter of knowledge, and the only object to which ideas ultimately refer is the experienced outcome of this transformation, as this is later to be "had" or immediately experienced. This does not seem, prima facie, to correspond to the intent of knowing or the manner of its use and validation in many sorts of inquiry, and a closer examination will tend to confirm the suspicion that the discrepancy is a radical one. I propose to examine this discrepancy as it first enters into and complicates Dewey's theory of "inquiry," and to show how the confusions it engenders are explained by the antecedent epistemological entanglements in which his notion of the relation of ideas to experience and of cognitive inquiry to practice is involved.

"Inquiry," Mr. Dewey tell us, "is the controlled or directed transformation of an indeterminate situation into one that is so determinate in its constituent distinctions and relations as to convert the elements of the original situation into a unified whole."³ The appropriate outcome of inquiry is a judgment "warrantably assertible" as following validly from the correct use of appropriate methods, but the objective of this judgment is simply the je-constitution of the situation in which thinking arose as response to the indeterminate or doubtful, in such wise that a final state of determinate resolution and unification is achieved.⁴ Propositions about matters of fact or possible courses of action are used in the process of reaching the final judgment and resolution. But these

are neither self-determined nor self-sufficient. They are determined with reference to an intended future issue and hence are instrumental and intermediate. They are not valid in and of themselves, for their validity depends upon the consequences which ensue from acting upon them—as far as these consequences actually ensue from the operations the propositions dictate and are not accidental accretions.⁵

All thought contains a practical factor, "an activity of doing and making which reshapes antecedent existential material which sets the problem of inquiry,"⁶ and since the ultimate reference of the ideas involved in this reshaping is to the reconstruction to be achieved, "The ultimate ground of every valid proposition and warranted judgment consists in some existential reconstruction ultimately effected."⁷

Mr. Dewey's development of this theory is accompanied by many illustrations of the way in which inquiry and investigation do alter antecedent situations by bringing to light facts and suggesting hypotheses relevant to the problem being investigated and are thus instrumental to finding out whatever it was that the investigator was previously in doubt about. Equally important in his exposition is the manner in which knowledge once attained is instrumental to further and non-cognitive interests, not least among them the interest in enjoying the outcome of knowing in a cleared-up situation as something worth having on its own account. Of the antecedents in preliminary analysis, hypothesis and the like, and of the consequences of knowing for many other humanly desirable ends Mr. Dewey has a great deal that is enlightening to say. But at the center of the theory there remains an ambiguity that is likely to puzzle even those most anxious to profit by its teachings. One s natural tendency is to suppose that what the ideas used and the analyses performed in the course of inquiry are instrumental to is finding out whatever it is that the particular inquirer was investigating, which might be immediate experience, or unperceived antecedent existence, or the structure of some purely hypothetical logical system, or anything else, existent or non-existent, which can in any way be investigated. And similarly we are inclined to suppose that the worth of knowledge for improving man's estate, or ushering in immediate experiences non-cognitively enjoyed, though enjoyed as the fruit of previous cognition, is essentially distinct from its worth as knowledge, as true belief about its own intended object in the sense that what it asserts to be so is so and is based on adequate evidence or arrived at by a method which leads reliably to true conclusions. Something of this sort seems to be presupposed in inquiry as ordinarily pursued. If we look for its equivalent, however, in Mr. Dewey's description of inquiry, we shall be at a loss to find it.

The most puzzling feature of this instrumental theory of knowledge is that, in the picture of inquiry it offers, knowing, in the sense in which it was understood in the preceding paragraph, seems not to occur at all. There are steps that would ordinarily be thought of as leading up to it—the "jam tomorrow" stage in which experienced events are not simply known on their own account but are used as signs, or instruments, or evidence for something else that is to be known. And there are steps leading away from it, the "jam yesterday" stage, at which the use of what is already known as a means for the attainment of some further satisfaction is stressed. But what, on the ordinary view, ought to occupy the central place between these two processes and to lend its significance to both, is just not there. It was Mr. Lovejoy, I believe, who observed that "I am about to have known" is the appropriate pragmatic equivalent for "I know." The comment, as applied to Mr. Dewey's theory, is enlightening.

It must not, of course, be supposed that this result is due to any inadvertence on Mr. Dewey's part. It means simply that if he is right, what I referred to as the "ordinary" view of knowing is in need of further analysis, and that when this is supplied the apparent reference to antecedent existence, to a cognitive validity of truth-claims essentially distinct from their efficacy in reconstructing experience, is replaced by a reference of ideas to future experience and to the means for so altering a present situation that a desired and anticipated future will reliably ensue. I am not at present concerned to deny that this is what knowledge "really" is, or that the ultimate objective of knowing may be what he takes it to be. Any decision on these points depends so largely on antecedent epistemological commitments that it is not, on the whole, a matter for fruitful discussion. It is important, however, to observe that this is not what knowing is "known as" when we take it for what it shows itself to be in physical, historical or sociological research and that the attempt to understand these activities in terms of the theory that Mr. Dewey has offered is more fruitful of epistemological controversy, in the disparaging sense in which he uses that term, than of philosophical enlightenment.

The particular issues that will serve best to illustrate the contrast between the theory of knowledge which a contextual analysis would appear to require and that to which Mr. Dewey, as a result of previous epistemological commitments, has been led, are those which concern the true or ultimate object of cognition and the relation of cognition to other and non-cognitive modes or access to reality. It will appear, I believe, that his discussion of these issues is not intelligible until we refer it back to the idealistic and realistic philosophies in relation and opposition to which it was developed, and that while it can be controversially justified as an alternative to them, its fruits in contextual application are of questionable value.

In The Quest for Certainty, Mr. Dewey has insisted at length that the assumption that "the true and valid object of knowledge is that which has being prior to and independent of the operations of knowing" is unwarranted and that on the contrary, "the true object of knowledge resides in the consequences of directed action."⁸ This is held to be particularly true and important in the case of the sciences, where it must be seen that "scientific conceptions are not a revelation of prior and independent reality,"⁹ but that, on the contrary, "scientific men accepted the consequences of their experimental operations as constituting the known object,"¹⁰ and cared nothing for an antecedent archetypal reality.

What precisely do these statements mean? In the procedure of inquiry, whether into the structure of the atom, the cause of infantile paralysis, or, if anyone is interested to investigate, the batting averages of all members of the New York Yankees baseball team in 1921, the true object of knowledge is surely just whatever it is that the inquirer wants to find out about. That antecedent being, as it existed prior to the operations of inquiry, can in this sense be a true and legitimate object of knowledge and even an archetype in so far as knowledge of it must conform to what it was, if it is to be the truth about it, seems not really doubtful. It is of course true that antecedent existence is not the true and valid object of knowledge, the only or exclusive one. On this point we should be inclined to say, if we came at the matter directly, that the question as to the true object of knowledge is a puerile one, since anything whatever can be an object of knowledge if there is any humanly possible way of finding out about it and if anyone is interested in finding out, and that the attempt to set up any such object as preeminently the true or genuine article is inspired by such extraneous moral or epistemological considerations as, e.g., that it is something important or desirable to know about or that it is what we must really be knowing if somebody's theory of knowledge is the true one. The objects of science are any objects about which the sciences can give us reliable information; and that some scientific conceptions are revelations of a prior and independent reality in that, by the use of them, true and warranted statements can be made about events and objects that existed before cognitive situations ever occurred and independent of such situations, is, I should think, as sure as anything can be in this uncertain world.

The inquirer interested in understanding what scientists are talking about, how their various theories and conceptions are instrumental to finding out, in an approximate but on the whole reliable way what is going on, has gone on and is likely to go on in the world, and what sort of evidence is available for testing statements made on these matters, will get no light from a discussion of the "true" object of knowledge. And when he discovers that Mr. Dewey, in order to emphasize the experimental and operational nature of scientific concepts, is forced to assert that these conceptions do not "ultimately" refer to past events, or indeed to anything but the empirically observable consequences of acting upon them, he is likely to be more confused than enlightened.

Why should Mr. Dewey have introduced such claims into his theory of knowledge? The answer is to be found not in the nature of scientific procedure but in that of Mr. Dewey's antecedent quarrel with a "spectator theory of knowledge." He defined the issue quite clearly in his Essays in Experimental Logic.

The new realism finds that it [thinking] is instrumental simply to knowledge of objects. From this it infers (with perfect correctness and inevitableness) that thinking (including all the operations of discovery and testing as they might be set forth in an inductive logic) is a mere psychological preliminary, utterly irrelevant to any conclusions regarding the nature of objects known. The thesis of the essays is that thinking is instrumental to a control of the environment, a control effected through acts which would not be undertaken without the prior resolution of a complex situation into assured elements and an accompanying projection of possibilities—without, that is to say, thinking.¹¹

Whether or not this characterization of the new realist's position is adequate is not here important. The essential fact is that when thinking is held to be instrumental to a knowledge of "reality" this "reality" is thought of by Dewey as something that is supposed to be known antecedently to and independently of the processes by means of which scientific investigation takes place. The "spectator" theory of knowledge thus stands in his mind for the view that we can know things by passively contemplating them or accepting preconceived ideas about them as adequate bases for conclusions as to their essential natures. It follows of course that scientific method, with its experimental manipulation of given data and constant modification of antecedently accepted ideas, is out of harmony with any such theory of knowledge. If this be knowledge of antecedent being, then neither scientific method nor experimental logic gives such knowledge.

It is clear, I think, that if we want to understand Mr. Dewey's puzzling denial of what seems, on the face of it, the obvious import of some reliably tested knowledgeclaims, we must think of him as referring to theories of this sort. His statements are simply not comprehensible apart from such reference. That is why, after burying the spectator theory of knowledge in one volume after another, he has been obliged to dig it up again in subsequent works to justify by contrast his own insistence thatthe true object of knowledge is not only got at by experimental methods but simplyis the observable outcome of such experimental procedures. But while this is intelligible, in terms of the controversy in question, it is not at all helpful as a characterization of the actual aims of experimental science, where it is not simply what will happen when an experiment is performed that is in question, but also what evidence this supplies about the nature and behavior of other objects which may themselves be beyond the range of observation but can be known about by means of experimental evidence. Mr. Dewey evidently does not mean to deny that in some sense we have such knowledge, but he does insist on stating the nature and aims of knowing in such fashion as to make any clear analysis of it impossible. And this is done because a straightforward statement suggests to him, though it does not by any means imply, an epistemological theory which he is extremely anxious to avoid. The epistemological controversy has thus impeded and confused the contextual analysis that was wanted.

A related matter on which Mr. Dewey finds it very important to insist is the falsity of the assumption "that knowledge has a uniquely privileged position as a mode of access to reality in comparison with other modes of experience."¹² Actually, things can be "had" in immediate experience, as well as known, and in such "having" "we experience things as they really are apart from knowing" while knowing is that special mode of experiencing things "which facilitates control of objects for purposes of non-cognitive experiences."¹³ Cognition thus is not, as philosophers have in the past assumed, "the measure of the reality found in other modes of experience," and an insistence on this fact is held to be of great significance.

The difficulty one finds in understanding these passages, and many others like them in Dewey, is that knowledge or cognition appears to be used in them in two different but not adequately distinguished senses. Referred to ordinary operations of investigation, knowledge consists of true statements or beliefs arrived at by a reliable method. The cognitive interest is that in the acquisition and adequate testing of such beliefs; and nothing can be directly relevant to the cognitive goodness of a belief or the satisfaction of the interest in knowing except that which tends to confirm or confute the belief in question by serving as evidence of the nature of its object.

It is only because some connection with this meaning for knowledge and the tests of its validity is carried over into Mr. Dewey's discussion that it retains its appearance of pertinence to inquiry in the sciences and practical affairs. But at the same time this is not what Mr. Dewey himself, in his specific epistemological analyses, takes knowing to be. Instead, as we have already seen, he regards it as a use of ideas as signs of possible future experiences and means for effecting the transition to such experiences in a satisfactory manner. These future experiences, in so far as they terminate inquiry, will not be cases of "knowing," i.e., of the use of given experiences as signs of something else. Hence what justifies cognition is not anything in the same sense "known" at all, but the occurrence of a noncognitive satisfaction, and the goodness of cognition in its own primary aim or intent, is determined by its use in bringing about such experiences.

Each of these accounts is intelligible enough in itself, but when we try to apply to non-philosophical inquiry the results reached on the epistemological level, confusions arise which have always surrounded the "instrumental" theory of knowledge. Are we actually to suppose that the validity of ordinary truth-claims as true is to be determined by something else than what we can find out as to the nature of their objects, or that we are to regard as evidence of their truth the fact they are instrumental to non-cognitive satisfactions? And does the claim that we have non-cognitive access to "reality" mean that we have any way of finding out what really exists or is the case, that is not, just in so far as it is a way of finding out, a way of knowing or of cognition? Of course we are related to our environment in many other ways than framing and testing beliefs about it. We "grasp reality" in seeing and handling perceptual objects, in enjoying good health and in doing all manner of things that are not knowing, and may be more satisfactory than knowing. But if the question arises as to what, in any of these ways of experiencing or behaving we find out about "reality," the only possible answer is a "cognitive" one, namely, in the first place a statement of what these experiences are, what, other than themselves they are evidence for and how they are relevant to the rest of what, in our various relations with it, we find the world to be. If "practice" reveals the nature of "reality" in any significant sense, it is surely because we find out something through our practical relations with things that we should not otherwise have known. And similarly, if, as Dewey maintains, we experience things as they really are apart from knowing, this is itself something that can be known, where "knowing" is not a reference to some future experience later to be noncognitively enjoyed as a result of present intellectual operations, but a true belief, tested by inspection of the very experiences in question, as to what these experiences are. Mr. Dewey himself makes many statements about immediate experience that he evidently regards as verifiably true in this sense.

In all this, however, we are putting "cognition" in a context other than that to which his own analysis refers. If cognizing is only what, for epistemological purposes, he takes it to be, then there are ways of finding out about objects, e.g., observing our own immediate experiences and making a true report of them, which are not "cognitive," since they are not ways of using the experiences in question as signs of future experiences. We then have "access to reality" which is not "cognitive" but nevertheless is a source of information about its intended object. And we have a goodness of ideas in the instrumental sense which bears no clear relation to their goodness as evidence for or information about the objects of which they purport to supply knowledge, but which nevertheless is somehow intended as an equivalent for truth in the more usual sense.

The controversy about pragmatism or instrumentalism has always been a particularly unrewarding one, in which neither party seemed at all able to understand what the other was saying, and any criticism offered from a non-pragmatic standpoint was rejected by the pragmatists as a misrepresentation. It is not my purpose to revive that controversy. What I want to point out is that its inconclusiveness and the misunderstandings on both sides were due to the fact that what Mr. Dewey says about cognition is true of it as he defines it, and false of it as more ordinarily understood, and that the attempt to interpret what he has to say in terms of the ordinary use of the term "truth" leads only to ambiguity. The truth-relation on his view is that of an idea to a future experience, when the idea is intended to suggest a way of behaving that will lead the thinker, if he acts upon it, to enjoy that future experience, and the goodness of an idea in that connection is its capacity to serve reliably for the purpose intended. The question of its truth in any other sense does not arise, and it is the essence of this extremely ingenious theory to see to it that it shall not arise. Once admitted into Mr. Dewey's epistemological universe of discourse, the critic will find himself quite unable to make the objection he had intended. It will be quite impossible to "know" immediate experience as it is in its own qualitative being, since to know anything is not to attend to it on its own account, but to use it as a sign of something else. Yet immediate experience will provide "access to reality," indeed, our only first-hand approach to it. Hence the claims of non-cognitive experience to epistemological primacy. And when ideas are used cognitively their intent as cognitive, as instruments for the resolution of an indeterminate situation, will be to eventuate and find their justification in something not in the same sense known at all. The subtlety with which this theory has been developed is of the highest order.

The crucial question is, however, for what was it devised, and for what is it useful? As an analysis of the interest and criteria of knowing in non-philosophical research it is not really helpful, for the terms in which it is stated and the assumptions on which it is based are not comprehensible except in relation to epistemological controversies on a quite different level, and when applied directly to the ordinary business of truth-seeking, result in endless misunderstanding. But if one recalls the idealism out of which Dewey's theory developed, the situation is altered. The limitation that Mr. Dewey puts on "cognition" and his insistence that it is to be justified by its furtherance of interests and satisfactions not in the same sense cognitive are the direct result of his rejection of his idealistic antecedents, and find their explanation in their relation to it.

In objective idealism "thought" had long since been divorced from the ordinary business of acquiring information. For reasons with which every student of epistemology will be familiar, thought had come to be regarded as a kind of construction, having its point of departure in immediate experience but transcending such experience in its search for an object that would fully satisfy thought's own demand for completeness and consistency. The cognitive interest was just this interest in systematic completeness and coherence, and proofs were not lacking that "Reality" must correspond to our ideas, satisfy the demands of thought, and the like.

The pragmatic revolt against this idealism was a thoroughly salutary one. Thinking, we were told, was to have its test not simply in meeting its own demands for consistency, but in meeting the demands of the situations in which it arose. This reference to "actual situations," which the Studies in Logical Theory brought to the center of philosophical discussion, was intended to correct the arbitrariness and isolation of a "thought" "absolute and self-inclosed" and to place the tests of thinking in its capacity to serve other ends than its own. That "cognition," in the sense in which the idealists had understood it, is "mediate" essentially and finds its justification in its relation to specific situations, is surely true. But that what justifies cognition is its relation to the objects, or facts, or events or whatever else, by its means, we can find out about was not, for Dewey, a live alternative. He was still too much of an idealist to refer directly from thought to its object. The reference was instead to immediate experience, to "practice" and, in general, to the satisfaction of other interests than knowing. The result was that, in denying the right of "thought" to lay down its own laws as to what "reality" must be and insisting on its essential responsibility to something beyond mere thinking, he was impelled to maintain the essential dependence of "cognition" for its validity on its capacity to satisfy "non-cognitive" demands. Such is the primary basis of "instrumentalism."

A position arrived at as a modification of idealism proved further useful in eliminating all reference to "antecedent," "trans-empirical," and otherwise undesirable objects on which the realists were by this time insisting. In the reflective situation as Dewey defines it no such reference occurs, for the only object an idea can be about is that empirically attainable future to which it serves as a guide. As a means of avoiding questions which, given this definition of knowledge, it would be difficult to answer, it is a model of its kind. But that kind is precisely the kind called by Mr. Dewey "epistemology," the kind that stipulates on grounds (mainly psychological in modern philosophy) what knowing is to be, and acknowledges only those operations which can be fitted into its pattern: the preliminary operation in which data are manipulated for use as evidence and the consequent operation in which the results of knowledge are used for the benefit of man's estate, but not the central and primary operation which is the finding out, on the basis of evidence, of those reliably ascertainable conclusions which can be used in a subsequent practical reconstruction because, on their own account, they constitute knowledge of the "situation" to which they refer.

We have now to assess the consequences of this view for an understanding of "knowledge," and an adequate theory of its nature. These will have to be judged on two levels. So far as the epistemological controversy is concerned, the theory is a formidable one. Once inside the "reflective situation" as Dewey describes it, there is no escape from his conclusions, and there are persuasive considerations that recommend such an analysis as opposed to those of its rivals. It is "empirical," "practical," even "operational" and these are all terms of praise in the marketplace of current discussion. The view can be "refuted," of course, as it has been many times, but the refutations proceed from assumptions about knowledge which have their own difficulties and whose results are in many respects out of harmony with facts to which Mr. Dewey and his followers can legitimately refer for support.

When the test is made, however, by reference to the measure in which the theory enables us to understand better and test more justly the knowledge-claims that are made in the course of non-epistemological inquiries, the result is a less favorable one. This does not mean that Mr. Dewey has not made valuable contributions to the subject. No one, certainly, has been at more pains to draw from the procedures of experimental science material illustrative of the thesis he is defending, or to insist that the test of any such theory must be found in its application to the actual subject. This interest in application, and in experimental procedures gives Mr. Dewey's Logic a solid content which more orthodox treatises on the subject rarely possess. He has shown us the environment, physical, biological, and social, in which such inquiry has to operate, and thus provided useful and philosophically relevant information about it. There are very few epistemological theories in the current crop of which as much can be said.

The difficulties that persist in any consistent attempt to apply his theory in contextual analysis or criticism of knowledge, however, are two. The first is an inescapable vagueness, a tendency to see all around the specific object, but never to focus clearly on the object itself. The suggestion that this is the effect of Mr. Dewey's manner of expressing himself is quite misguided. On the contrary, he has said just what, in terms of his epistemology, he ought to say, with great skill. The point rather is that his theory focuses attention exclusively on the antecedents and consequences of knowing. Hence when we try to fix our gaze on the object, logical, scientific, or qualitative, as itself an object of knowledge, what at a distance seemed substantial enough becomes diffused and spreads out over its immediate neighborhood. And when we insist on locating it more exactly, we are met by statements not about it at all, or our means of finding out about it, but rather about the inaccessibility of an "antecedent" or archetypal reality, the futility of "self-enclosed" thought, and the value of thinking as a means of enriching the life of men. It then becomes apparent that the theory is afflicted with epistemological strabismus, that one eye has been fixed all the time on the defects of opposing theories, and that the specific nature of cognitive inquiry has, in consequence, been blurred.

The second difficulty in application arises not in understanding the various knowledge-claims made but in estimating their validity. Since all thinking is "practical," according to Dewey, and all is justified by the reconstitution of experience it is instrumental in achieving, we should expect to judge of the "worth" of thinking by its results. But here it is important to distinguish between the goodness of knowing for its own primary objective as knowing and its goodness as a means to other ends. It is, again, of the essence of Dewey's theory that it prevents us from making this distinction in any clear way. Intellectual inquiry is practical in that it involves making choices, manipulating materials, testing hypotheses, and so far as the inquiry is successful, changing the situation in which the investigator is doubtful into one in which his mind is at rest on the point at issue. Hence "the conduct of scientific inquiry, whether physical or mathematical, is a mode of practice; the working scientist is a practitioner above all else, and is constantly engaged in making practical judgments: decisions as to what to do and what means to employ in doing it."¹⁴

Yet it is quite evident that the outcome of all this "practice" may be a well grounded theory which is not "practical" in the sense that it has the least use in or relevance for any further ends, and, in particular, for those that are socially significant or regarded by wise and good men as of primary importance. Whether all knowledge that a man with a social conscience ought to permit himself to pursue must be in this further sense "practical" is a genuine moral question, but it is not the same at all as the question whether the "practice" involved in his inquiry itself has justified, i.e., verified or given grounds for believing in the truth of, the hypothesis with which he was working. There is, then, a practice intrinsic to knowledge-getting, and a "practice" to which it may or may not be relevant, and which is of much moment for its moral or social worth but need be of none at all—unless the hypothesis happened to be about social uses and worth—in testing its truth.

This distinction is, in our own time, of some importance. For earnest men in a hurry are likely to be impatient of inquiries whose relevance to their own socially reconstructive aims is not apparent. If they are encouraged to confuse the practical value of such inquiries for the ends which they take to be of primary importance, with their cognitive validity, they will inevitably judge them unfairly and make utility for the furtherance of their preferred interests the final measure of what is actually true about the world.

That Mr. Dewey himself intends to subordinate scientific method to more immediate social interests is not even plausible enough as a suggestion to make its denial important. No one in this generation has done more than he to celebrate the value of scientific thinking or to discredit attempts to subordinate the pursuit of truth by its means to ulterior interests. But that is precisely because he believes that scientific method is "practical" in the social and moral sense as well. When he comes to deal with intellectual inquiries of whose relevance to the interests he regards as important he is less convinced, his position is a more dubious one. In a recent contribution to the International Encyclopedia of Unified Science, he has drawn a sharp line between those types of inquiry that manifest a scientific attitude and are therefore, in his view, to be approved, and those that do not. And one of the bases for distinguishing a scientific from an unscientific attitude is stated as follows:

Above all, it [the scientific attitude] is the attitude which is rooted in the problems that are set and questions that are raised by the conditions of actuality. The unscientific attitude is that which shuns such problems, which runs away from them, or covers them up instead of facing them. And experience shows that this evasion is the counterpart of concern with artificial problems and alleged ready-made solutions. For all problems are artificial which do not grow, even if indirectly, out of the conditions under which life, including associated living, is carried on.¹⁵

Now, even artificial problems have presumably grown somehow, since some people are concerned with them, and if the conditions of living are not accountable for them, this must be because "living" is here used in a eulogistic sense, to refer to such conditions as Mr. Dewey thinks are important. Either all problems whatever arise in conditions of life and actuality since they do actually arise in the life history of human beings, or else "life" and "actuality" stand for the conditions in which important problems arise, or those we ought to take account of. And when we recall that it was the cognitive goodness of scientific method that was in question here, as opposed e.g. to that of the sort of metaphysics and compensatory religious belief of which Mr. Dewey disapproves, we can see how fatally easy it is to make relevance to conditions felt to be important a primary criterion of the goodness of knowledge as knowledge.

With Mr. Dewey's valuations on this subject I have no quarrel. But with his tendency to make such valuation the test of what is valid as knowledge, I have. The essential fact is that where the distinction between the value of an idea as a means for discovering the truth has been confused with its value as a means for subserving interests felt on other grounds to be important, there is simply no basis left for an independent estimate of truth as such. Mr. Dewey's theory, however liberal its intentions, does involve this confusion and does therefore in practice leave all claims to knowledge at the mercy of ulterior preconceptions about what is "actual," "living," or "socially significant." The theory would be sounder and more useful with this epistemological confusion eliminated.

II

The term "metaphysics," like "epistemology," has often in recent years been used in a derogatory sense. But on this point Dewey does not as a rule adopt the current fashion. In Experience and Nature, he takes metaphysics as "a statement of the generic traits manifested by existences of all kinds without regard to their differentiation into physical and mental,"¹⁶ and maintains that an adequate metaphysic can supply a ground-map of the province of criticism. To note, for example, that contingency is a pervasive trait of natural events, and to bring this fact into connection with concrete situations of life, is to provide a metaphysical basis for value judgments. And, in general,

the more sure one is that the world which encompasses human life is of such and such a character (no matter what his definition), the more one is committed to try to direct the conduct of life, that of others as well as of himself, upon the basis of the character assigned to the world.¹⁷

The understanding of man with his wants and hopes and limited capacities as a factor in the natural world out of which the human organism has developed and with which, in even its loftiest flights, the human spirit remains essentially continuous, is then the primary task for this metaphysics, and one which, as Mr. Dewey rightly observed, would provide a sound basis for that criticism of values and meanings with which philosophy in his view is primarily concerned.

In one sense this is a very modest project. "This is the extent and method of my metaphysics: the large and constant features of human sufferings, enjoyments, trials, failures and successes together with the institutions of art, science, technology, politics and religion which mark them, communicate genuine features of the world within which man lives."¹⁸ No more "transcendent" reality than the world of natural events is referred to, and the situation of human experience within nature provides the limited but reliable basis on which this "empirical naturalism" is to be built.

In another sense, however, the task that Mr. Dewey has set himself is more arduous than that of traditional metaphysics. For he proposes to use an empirical method throughout, and this has by no means usually been the procedure in these matters. In concrete experience

things present themselves in characteristic contexts, with different savors, colors, weights, tempos and directions. Experience as method warns us to give impartial attention to all of these diversifications. Non-empirical method sets out with the assumption that some one of these groupings of things is privileged, that it is supreme of its own right, that it furnishes a standard by which to measure the significance and real quality of everything else.¹⁹

This seems to me a remarkably sound and important point. The word "empirical" is perhaps not happy in this connection, but there is no need to argue over terminology. The fact, which badly needed emphasis, and which in the passage quoted admirably receives it, is that we do come at things in a variety of contexts and that while each of these reveals the "real" nature of things as thus discovered, there is no good reason to believe that any one among them provides a unique approach to "reality" as such, or a preferentially ultimate basis for metaphysics. The attempt to discover a reality thus ultimate and inclusive, to which all that in any fashion we find out about the world must be referred if we are to understand its kind and degree of "reality," whether this be "matter" or "mind," God or Nature, the Absolute or the inevitable dialectical development of history, has not proved, on the whole, an enlightening one. The "reality" discovered is always at best an aspect of the world for which it is supposed to provide the final explanation, and to attempt to unify all experience by reference to or derivation from this metaphysical ultimate will finally confuse our notion of the world and leave us vainly trying to connect this "reality" with what, on other and more substantial grounds we know to be the case. It is no wonder that "metaphysics," thus understood, has fallen into disrepute. Mr. Dewey's attempt to provide a basis for philosophical criticism by reference to the pervasive features of existence which are found alike in human experience and striving and in the world of events with which that striving is inevitably bound up, and to do this without recourse to metaphysical simplification of the sort he condemns as "non-empirical," is an uncommonly hopeful and promising one.

The title of his principal work on metaphysics, Experience and Nature, indicates the point of departure for his theory. Its center is the human situation, as this is disclosed in the whole course of our experience of or commerce with the world. The essential fact is that "this human situation falls wholly within nature. It reflects the traits of nature, it gives indisputable evidence that in nature itself qualities and relations, individualities and uniformities, finalities and efficacies, contingencies and necessities are inextricably bound together."²⁰ Experience is continuous with the rest of nature in that it is both a consequence of purely natural (physical, biological and social) interactions and also a fair sample of what natural events really are. The pervasive traits of human experience are traits of nature itself and can be used in metaphysics as a guide to its character. "Man fears because he exists in a fearful, an awful world. The world is precarious and perilous."²¹ And so, as Dewey has so frequently insisted, the indeterminate situation which elicits thought and requires reflection for its adequate resolution is as objective as any other natural situation. A change in it is brought about in a satisfactory and reliable way only when we have altered factors in the environment, not merely our feelings or beliefs about it. Thought in its dealing with the doubtful or precarious is itself a development within the natural world, and the changes it initiates are as "real" as a thunderstorm or an earthquake and as genuinely, though not as a rule as catastrophically, effective in altering the world of nature.

In academic discussion, this view has sometimes been described as "objective relativism," in order to stress the fact that the experienced world is at once in some of its major features dependent on and conditioned by the special relations in which sentient (and more particularly human) organisms stand to their environment and also a direct presentation of that environment itself, or the order of natural events, as it is under such conditions. Nature is not something essentially beyond the range of perceptual inspection, having its exclusive being in characters independent of all relation to human responses to it. Far more in the natural world than we can ever experience there certainly is and must be. But unless what we experience also belongs to and is, under the special but entirely natural conditions of organic interaction, a sample of the nature to which we claim to refer, then our relation to this ulterior nature becomes problematic, and the conditions of interaction which are in fact our means of getting in touch with it are treated as barriers to knowledge of what it is. It is evident, I think, that if metaphysics is to unify our knowledge by stressing the connections between what we experience the world as being and what, more indirectly, we find out about its nature and behavior under conditions in which we cannot ourselves observe it, some such principle is essential. The alternative appears to be, as in fact it has proved to be, an essentially unplausible attempt to predicate of the natural world, as it exists independently of any sentient organism's response to it, the characters which it takes on, sofar as we know, only under these special conditions, or, failing this, to insist that the world as experienced and in its empirically discoverable connections is not the "real" world at all, but only an inadequate subjective counterpart of it. From that point on, the relation between experience and nature will be whatever the metaphysician chooses to take it to be; for his only real ground for supposing that there is a connection will be some remnant of those empirically discoverable interactions which his theory actually has rendered dubious but which for epistemological reasons or as a result of "animal faith" he insists on retaining. In asking philosophers to turn their attention rather to that relation between experience and nature which consists in the fact that what we experience is the outcome of natural and scientifically describable processes and that it is itself an instance and, within its contextual limits, a fair sample, of what nature is and thus a sound basis for a further exploration of it, Dewey has served his subject well and has brought the whole discussion back to a point from which a comprehensive estimate of man's place in nature might profitably proceed.

The "naturalism" of such a theory is founded partly on fact, in so far as it stresses the dependence of the "higher" or more spiritual aspects of human behavior on the physical, biological and social environment in which they are manifested, and partly on a decision that the whole of experience is to be interpreted as falling within the situation which this environment determines. No conclusive proof for such a comprehensive naturalism is possible, since demonstrations of the traditional type that "nature," whether identified with matter, or space-time, or the evolutionary process, is all that can be real, have been ruled out by the method of this empirical metaphysics. It does, however, provide a standpoint for the organization of experience as a whole in which everything on other grounds reliably verifiable finds a credible place, and in which "value" and "meanings" are freed from dubious speculative and supernatural entanglements without being robbed of any of their human validity in the process. As such many reasonable men will prefer it to any speculative alternative and will find it justified to the extent in which it gives order, proportion, and a basic and essential sanity to their total view of things. While it cannot be said that Mr. Dewey has worked out a fully satisfactory theory along these lines, he has made very notable contributions to it.

There is, nevertheless, a skeleton in the closet of this admirably planned metaphysics, which has seriously compromised its good repute and thus given aid and comfort to its speculative rivals. It consists in an unhappy discrepancy between experience as it ought to be if its place in the natural world is to be made intelligible, and experience as it must be if Dewey's epistemology is correct. In the former capacity, "experience" is the essential link between man and a world which long antedates his appearance in it. In the latter, "experience" is the terminus of all knowing, in the sense that all our cognitive claims refer ultimately to what experience will show itself to be in a "resolved" situation and to nothing else. If this latter account is true, all statements about a natural environment outside of these immediate experiences become on analysis simply means of facilitating cognitive transitions to such enjoyed immediacies and the world which should have provided the background for our experience, and the measure of its metaphysical significance, "collapses into immediacy," and Mr. Dewey's naturalism reduces, as Mr. Santayana has said, to a "philosophy of the foreground."²²

The difficulty does not arise from Mr. Dewey's laudable attempt to treat experience as continuous with the rest of nature. It is the result of a misinterpretation of the way in which experience functions in knowing, and its consequence is that the reference to "experience" cuts us off from, instead of connecting us with, the circumambient environment, in which experience must be placed if it is to retain its status as a natural event.

The precise nature of this difficulty can be specified by reference to the relation described in Experience and Nature, of scientific objects to qualitied events as these are immediately experienced. The latter are held to be "ends" both for knowledge and in nature, the former are relational and "instrumental," the conceptual means we use for establishing connections between qualitied events. These qualitied events are taken as the type of what a natural event, in its concrete individuality, is, and it follows from this that the objects with which the physical sciences deal, representing the statistical outcome of complex processes of measurement and comparison, are not "individual existential objects" in this primary sense at all. "The procedure of physics itself, not any metaphysical or epistemological theory, discloses that physical objects cannot be individual, existential objects. In consequence, it is absurd to put them in opposition to the qualitatively individual objects of concrete experience."²³ The difference between the "world of physics" and the "world of sense" is to be explained as that between objects of thought, when thought is essentially instrumental to the satisfactory reconstruction of experienced and "qualitied" situations, and objects of immediate enjoyment—the ends or termini of thinking, in their directly apprehended being.

Thus "the proper objects of science are nature in its instrumental characters"²⁴ and hence science "is not a final thing. The final thing is appreciation and use of things of direct experience."²⁵

This seems to me to violate the primary principle of Dewey's "empirical method" as previously laid down. The objects about which the physical sciences provide abstract and schematic but nonetheless reliable information are not in their own nature "instrumental" at all. Our knowledge of them is instrumental in so far as we can use it to make life in other respects more satisfactory and in this instrumental use science is, of course, not a final thing. But to suppose that the whole meaning of what science tells us about the physical environment is reducible to this instrumental function is to treat one context in which things come to us as ultimate for metaphysics, and this is an irreparable mistake.

For we actually need scientific information not merely in this instrumental capacity, but as information about the causes and conditions of human experience, if our naturalism is not to lapse into a hopelessly anthropocentric view of things. Of course the things we directly experience are more "concrete" than the objects of science, since they are the only things we can get at in terms of sensuous content and emotional associations and practical uses. They owe this special status, however, precisely to the fact that we stand in special relations to them. Whatever the "individuality" of an object beyond the range of direct experience may be it will always for us remain abstract and relational, since all we can know about it is what can be inferred from its relations to other things. To take "concreteness" in this sense as the measure of individual reality, and abstractness as evidence of a merely "relational" or instrumental character in nature is to make the special conditions under which a sentient organism gets into connection with things the measure of their reality. It is the basis for every sort of panpsychism and animism, but hardly for an empirical naturalism.

Since Mr. Dewey has no leanings toward animism or other such attempts to reduce the world to the human scale, it is in a different direction that we must look if this unhappy situation is to be explained. We have not far to look. If conceptual knowing as it functions in physics is a way of finding out by such means as we can what the physical environment is, then experience is not its exclusive object. Objects that cannot be come at directly must of course be known about through such report as other things give of their nature. Experience in such investigation is not the terminus of knowing but a means to knowing about something else. And it is what is thus indirectly known that gives "experience" the meaning that Dewey wants to place on it as a natural event. It would be quite illegitimate to suppose that the reference to such objects renders what we experience "unreal," or "bifurcates" nature. Neither scientific investigation nor empirical observation and enjoyment can be the measure of any other reality than that of its own appropriate object. But it is essential to acknowledge that we know both what the world is like under the conditions of observation and enjoyment and also what, in a much more general and approximate way, the unexperienced environment is by which experience is conditioned. If we could not know this there would be no sense in calling this environment "nature" and regarding experience as our means of finding out about it. On any contextual analysis of knowing this is a feasible and straightforward interpretation.

If, however, the meaning of conceptual inquiry is defined simply as its capacity to refer to and satisfactorily initiate future experiences, experience loses this "vehicular" significance altogether and the limits of what we can enjoy or immediately "have" become the limits of the world to which we can significantly refer. And since these limits are relatively narrow ones, we are left with a reality far too limited for what, outside metaphysics and epistemology, we find the world to be. Mr. Dewey himself warned us of where such a non-contextual metaphysics would lead and the result is an impressive verification of his warning.

The attempt to establish this has necessitated a somewhat ungrateful emphasis on the defects of a theory which, as compared with its rivals, has very much to recommend it and which stands today as the most significant contribution America has made to philosophic enlightenment. It would have been possible, of course, to stress this aspect of the matter more, and to write an essay, as many have been and will be written, in praise of this philosophy. But this outcome, though more laudatory, would have been less appropriate, I think, to the actual merits of its subject. For Mr. Dewey did not compose a philosophy to be appreciated, along with other speculative and literary monuments, as an impressive specimen of man's answer to the riddle of existence. He developed rather a method for clarifying ideas and testing theories in such fashion that men, less interested in even ultimate riddles than in knowing what their statements mean and what, in relation to their own more inclusive purposes; their various activities are worth, could see more clearly and judge more sanely in these matters. In this he has placed us all greatly in his debt. This is a continuing work, and one in which there is still very much to do. We need to know how best to do it, and to decide, more definitely than was possible when Dewey's own work began, what elements in the theory presented are reliable for the purpose in hand. It is because I believe this work is profoundly worth continuing that I have tried here to suggest the line along which work can now most profitably proceed.

The conclusion "warrantably assertible" as the outcome of the preceding analysis is, I believe, the following: Mr. Dewey's epistemology is not, either in its method or results, in harmony with the philosophical procedures he has recommended or the empirical metaphysics he proposes to develop. The procedure would be more directly applicable in non-epistemological contexts and far less open to misconstruction, the metaphysics freed from a serious and quite gratuitous difficulty, if this epistemology were abandoned. In so far as Mr. Dewey's philosophy is one of the rival "positions," developed in the last generation as competing accounts of the "true" object of knowledge and the "final" meaning of truth, this conclusion amounts to the claim that that position is in a fundamental respect untenable. In so far, however, as this philosophy has been, more than any other of the period, a project for the use of philosophic analysis and criticism for the clarification of basic ideas and the coördination of the various aspects of the world as experienced and known, the conclusion offered amounts to a suggestion for its wider and more consistent application. It is the enduring worth of Mr. Dewey's philosophy in this latter aspect that lends whatever significance it may have to this discussion.

NOTES

¹Logic, 8.

²Ibid., 534-5.

³Logic, 104-5.

⁴Ibid., 134.

⁵Ibid., 164.

⁶Ibid., 160.

⁷Ibid., 489.

⁸The Quest for Certainty, 196.

⁹Ibid., 165.

¹⁰Ibid., 185.

¹¹Essays in Experimental Logic, 30.

¹²The Quest for Certainty, 106.

¹³Ibid., 98.

¹⁴ Logic, 161.

¹⁵International Encyclopedia of Unified Science, Vol. I, No. I, 31.

¹⁶Experience and Nature, 412.

¹⁷Ibid., 413-4.

¹⁸Journal ïf Philosophy, Vol. XXIV, 59.

¹⁹Experience and Nature, 15.

²⁰Ibid., 421.

²¹Ibid., 42.

²² See Santayana's review of Experience and Nature in the Journal of Philosophy, Vol. XXII, 680ff., and Dewey's reply in Vol. XXIV, 57ff.

²³The Quest for Certainty, 241.

²⁴Experience and Nature, 137.

²⁵The Quest for Certainty, 221-2.