The dispute as to whether reason and conception or perception and sense are the source and test of ultimate knowledge is one of the most enduring in the history of thought. It has affected philosophy from the side of both the nature of the object of knowledge and the mental faculty operating to obtain it. From the side of the object those who put forward the claims of reason have placed the universal higher than the individual; those who have held to perception have reversed the order. From the side of mind, one school has emphasized the synthetic action of conceptions. The other school has dwelt upon the fact that in sensation the mind does not interfere with the action of objects in writing their own report. The opposition has extended to problems of conduct and society. On one hand, there is emphasis upon the necessity of control by rational standards; on the other hand, the dynamic quality of wants has been insisted upon together with the intimately personal character of their satisfaction as against the pale remoteness of pure thought. On the political side, there is a like division between the adherents of order and organization, those who feel that reason alone gives security, and those interested in freedom, innovation and progress, those who have used the claims of the individual and his desires as a philosophical basis.

The controversy is acute and longstanding. In consequence of it, philosophers have expended energy in controversy with one another, and the guidance they have given to practical affairs has been largely by way of support to partisans of contending forces. The situation raises a further point in our inquiry: What is the bearing of the experimental theory of knowing upon the rival contentions? The first point which presents itself is that the object of knowledge is eventual; that is, it is an outcome of directed experimental operations, instead of something in sufficient existence before the act of knowing. The further point to be presented is that, along with this change, sensible and rational factors cease to be competitors for primary rank. They are allies, coöperating to make knowledge possible. Isolation from each other is an expression of the isolation of each from organic connection with action. When theory is placed in opposition to practice, there is ground for dispute as to whether primacy in theory shall go to sense or intellect. Directed activity demands ideas which go beyond the results of past perceptions, for it goes out to meet future and as yet unexperienced situations. But it deals, both in origin and outcome, with things which can be had only directly, through immediate perception and enjoyment.

The three chief contending doctrines in this field are sensational empiricism, rationalism and Kantianism, with its compromise of the factors isolated in the two other schools. The doctrine of Kant has a superficial resemblance to the one just stated; it insists upon the necessity of both perception and ideas if there is to be knowledge. It is convenient, accordingly, to begin discussion with it. The element of similarity is suggested by Kant’s well known saying that perception without conception is blind, conception without perception empty. His doctrine none the less is fundamentally different from that which results from an analysis of experimental knowing. The fundamental difference lies in the fact that, according to the latter, the distinction of sense and thought occurs within the process of reflective inquiry, and the two are connected together by means of operations overtly performed. In the Kantian scheme, the two originally exist in independence of each other, and their connection is established by operations that are covert and are performed in the hidden recesses of mind, once for all. As to their original difference, sense-material is impressed from without, while connective conceptions are supplied from within the understanding. As to connection, synthesis takes place not intentionally and by means of the controlled art of investigation, but automatically and all at once.

From the experimental point of view, the art of knowing demands skill in selecting appropriate sense-data on one side and connecting principles, or conceptual theories, on the other. It requires a developed and constantly progressive technique to settle upon both the observational data and the idea that assist inquiry in reaching a conclusion in any particular case. But in Kant’s view, the distinction and the connection between the two, while necessary to anything which may be termed cognition, have nothing to do with the validity of any particular enterprise of knowing. Illusion and error exemplify the synthesis of sense and understanding quite as much as does the soundest instance of scientific discovery. In one case, the heart of the whole matter is the exercise of a differential control which makes the difference between good and bad knowing. In Kant’s scheme the blessings of the categories descend upon the material of sense without reference to making a distinction between the true and the false.

We summarize the differences as follows. 1 In experimental knowing, the antecedent is always the subject-matter of some experience which has its origin in natural causes, but which, not having been controlled in its occurrence, is uncertain and problematic. Original objects of experience are produced by the natural interactions of organism and environment, and in themselves are neither sensible, conceptual nor a mixture of the two. They are precisely the qualitative material of all our ordinary untested experiences. 2. The distinction between sense-data and interpretive ideas is deliberately instituted by the process of inquiry, for sake of carrying it forward to an adequately tested conclusion, one with a title to acceptance. 3. Hence each term of the distinction is not absolute and fixed, but is contingent and tentative. Each is subject to revision as we find observational data which supply better evidence, and as the growth of science provides better directive hypotheses to draw upon. 4. Hence the material selected to serve as data and as regulative principles constantly check one another; any advance in one brings about a corresponding improvement in the other. The two are constantly working together to effect a rearrangement of the original experienced material in the construction of a new object having the properties that make it understood or known.

These statements are formal, but their meaning is not recondite. Any scientific investigation illustrates their significance. The astronomer, chemist, botanist, start from the material of gross unanalyzed experience, that of the “common-sense” world in which we live, suffer, act and enjoy; from familiar stars, suns, moons, from acids, salts and metals, trees, mosses and growing plants. Then the process of investigation divides into two kinds of operations. One is that of careful and analytic observation to determine exactly what there is which is indubitably seen, touched and heard. An operation takes place to discover what the sure data of the problem are, the evidence which theoretical explanation must reckon with. The other operation consists in searching through previous knowledge to obtain ideas which may be used to interpret this observed material and to suggest the initiation of new experiments. By these latter, more data are had, and the additional evidence they supply suggests new ideas and more experiments until the problem is resolved. The investigator never makes the division between perceptual and conceptual material at large or wholesale. He is careful at each stage of inquiry to discriminate between what he has observed and what is a matter of theory and ideas, using the latter as means of directing further observations, the results of which test the application of the ideas and theories employed. Finally, the original material is reorganized into a coherent and settled form capable of entering integrally into the general system of science.

A physician, for example, is called by a patient. His original material of experience is thereby provided; it requires a stretch of useless imagination to fancy that the ill man is a mass of sense data organized by categories. This experienced object sets the problem of inquiry. Certain clinical operations are performed, sounding, tapping, getting registrations of pulse, temperature, respiration, etc. These constitute the symptoms; they supply the evidence to be interpreted. The philosopher or logician, looking on, sees they are that part of the original object which is capable of being presented in observation as that is sensibly present. The results are not all that is or can be observed, but are those phases and portions of the experienced whole that are judged to be relevant to making an inference as to the nature of the ailment. The observations mean something not in and of themselves, but are given meaning in the light of the systematized knowledge of medicine as far as that is at the command of the practitioner. He calls upon his store of knowledge to suggest ideas that may aid him in reaching a judgment as to the nature of the trouble and its proper treatment. The analytic philosopher, looking on, notes that the interpreting material, by means of which the scattered data of sense are bound together into a coherent whole, is not itself directly sensibly present. So he calls it ideational or conceptual.

Sense data are signs which direct this selection of ideas; the ideas when suggested arouse new observations; the two together determine his final judgment or diagnosis and his procedure. Something is then added to the store of the clinical material of medical art so that subsequent observations of symptoms are refined and extended, and the store of material from which to draw ideas is further enlarged. To this process of coöperation of observation and conceptual or general ideas there is no limit. In no case are the data the whole of the original object; they are material selected for the purpose of serving as evidence and signs. In no case do general ideas, principles, laws, conceptions, determine the conclusion—although just as some men collect fragmentary observations without trying to find out what they mean, so in other cases an unskilled worker permits some preconceived idea to control his decision instead of using it as a hypothesis.

The case seems simple enough, so simple indeed that it may be supposed that we have overlooked the conditions which have created perplexity and controversy. But the source of these complications is that theories about the mind, about sensation and perception, about reason, the intellect, conceptions and perception, were framed and established in philosophy before the rise of experimental knowing. It is difficult to break loose from habits thus engendered so as to turn attention in a whole-hearted way to actual inquiry. While it may seem presumptuous to set up the case of the physician or some other concrete inquirer over against the elaborate machinery of the Critique of Pure Reason and the countless tomes of commentary it has called forth, our picture has behind it the whole weight of the experimental practices by which science has been actually advanced.

More specifically, it may be asserted that the Kantian theory went wrong because it took distinctions that are genuine and indispensable out of their setting and function in actual inquiry. It generalized them into fixed and wholesale distinctions, losing sight of their special rôles in attainment of those tested beliefs which give security. Consequently artificial complications were engendered, and insoluble puzzles created.

Take for example the fragmentary and isolated character of sense-data. Taken in isolation from a context in a particular inquiry they undoubtedly have this character. Hence when they are generalized into a character at large, the result is the doctrine of the disconnected “atomicity” of sense-data. This doctrine is common to sensationalism and to some forms of the new realism, along with Kantianism. As a matter of fact, smells, tastes, sounds, pressures, colors, etc., are not isolated; they are bound together by all kinds of interactions or connections, among which are included the habitual responses of the one having the experience. Some connections are organic, flowing from the constitution of the subject. Others have become engrained in habit because of education and the customary state of culture. But these habitual connections are obstacles rather than aids. Some of them are irrelevant and misleading. In any case, they fail to provide the clews, the evidence, which is wanted in the particular inquiry in hand. Consequently, sense qualities are artificially isolated from their ordinary connections so that the inquirer is free to see them in a new light or as constituents of a new object.

Since the very need for inquiry shows that there is a problem set by the existing situation, there can be no understanding of it achieved until there are new connections established. The fragmentary and isolated character of sense-data does not therefore describe anything belonging to them intrinsically, but marks a transitory, although necessary, stage in the progress of inquiry. The isolation of sense-data from their status and office in furthering the objective of knowing is responsible for treating them as a kind of isolated atomic existence. If we keep an eye on the actual enterprise of knowing, it is clear that only sense-data can supply evidential subject-matter; ideas of what is not presented in sense interpret evidence, but they cannot constitute it. The whole history of science shows, however, that material directly and originally observed does not furnish good evidential material; as we saw, it was the essential mistake of ancient science to suppose that we can base inference upon observed objects without an artificial prior analytic resolution. Hence there is need of a distinctive type of experimental operations which detach some qualities of the object; these form sense-data in the technical meaning of the word.

Traditional empiricism was accordingly right in insisting that no amount of conceptions, of thought material, could by itself deliver any knowledge of existence, no matter how elaborate be the conceptual system and how internally coherent. We cannot derive existence from thought—pace idealism. Observed material is necessary to suggest ideas and it is equally necessary to test them. The senses are, existentially speaking, the organs by which we obtain the material of observation. But, as we have previously noted, this material is significant and effective for purposes of knowing only as it is connected with operations of which it is the product. Merely physical interactions, whether of external things or of the organism, yield observations that form the material of inquiry; a problematic material. Only operations intentionally performed and attentively noted in connection with their products give observed material a positive intellectual value, and this condition is satisfied only by thought: ideas are the perception of this connection. Even non-scientific experience, as far as it has meaning, is neither mere doing nor mere undergoing, but is an acknowledgment of the connection between something done and something undergone in consequence of the doing.

In its later history, empiricism tended to identify sensory consequences with “mental” or psychical states and processes; this identification was the logical conclusion of taking the object of science, in which these qualities are not found, as the real object. But the insistence, as by contemporary realists, that sense-data are external and not mental does not remedy the logical error. It repeats the isolation of sense-data from the intentional operations by which they are supplied and from the purpose and function of these operations. Hence it makes it necessary to call in the supplement of logical objects, now termed essences. What is even more important, no light is thrown upon the control of the course of actual inquiry. For there is still failure to see that the distinction between sense-data and objects of rational apprehension is one which occurs within reflective investigation, for the sake of regulating its procedure.

The history of the theory of knowledge or epistemology would have been very different if instead of the word “data” or “givens,” it had happened to start with calling the qualities in question “takens.” Not that the data are not existential and qualities of the ultimately “given”—that is, the total subject-matter which is had in non-cognitive experiences. But as data they are selected from this total original subject-matter which gives the impetus to knowing; they are discriminated for a purpose:—that, namely, of affording signs or evidence to define and locate a problem, and thus give a clew to its resolution.

If we recur to the instance of the patient and the inquiries of the physician, it is evident that the presence of a man who is ill is the “given,” and that this given is complex, marked by all kinds of diverse qualities. Only the assumption—such is made by Kant and is common to the traditional theories—that all experience is inherently cognitive leads to the doctrine that perception of the patient is a case of knowledge. In reality the original perception furnishes the problem for knowing; it is something to be known, not an object of knowing. And in knowing, the first thing to be done is to select from the mass of presented qualities those which, in distinction from other qualities, throw light upon the nature of the trouble. As they are deliberately selected, being discriminated by special technical operations, they become data; these are called sensible simply because of the rôle of sense organs in their generation. They may then be formulated as the subject-matter of primitive existential propositions. But even so, there is no class of such propositions in general. Each inquiry yields its own primitive existential propositions, even though they all agree in having for their objects qualities which investigation reveals to be connected with the use of organs of sense. Moreover, these primitive propositions are such only in a logical sense as distinct from being empirically primitive, and they are only hypothetical or conditional. This statement does not imply that their existence is hypothetical; perception, as far as it is properly conducted, warrants their existence. But their status in inquiry is tentative. Many, perhaps most, errors in physical inference arise from taking as data things that are not data for the problem in hand; they undoubtedly exist, but they are not the evidence that is demanded. In some respects, the more undoubted the existence of sensory qualities, the less certain is their meaning for inference; the very fact that a quality is glaringly obvious in perception exercises an undue influence, leading thought to take its evident presence as an equivalent of evidential value. The reader of detective stories is aware that it is a common device to have the inquirer misled by the too patent character of given “clews”; genuine clews are usually obscure and have to be searched out. The conditional character of sense-data in inferential inquiry means, then, that they have to be tested by their consequences. They are good clews or evidence when they instigate operations whose effect is to solve the problem in hand.

It is hardly necessary to repeat the criticisms of the rationalistic doctrine of conceptions that have been brought out in previous chapters. The doctrine stood for a positive truth:—the necessity of relations, of connectivity in existence and knowledge, and it noted the fact of the connection of relations with thought. For while some connections are always found in the material of experienced things, the fact that as experienced these things are problematic and not definitively known, means that important relations are not presented in them as they stand. These relations have to be projected in anticipation if the reactions of the inquiries are not blind fumblings—if they are genuinely experimental. Such relations must be thought; they are present conceptually, not sensibly. They represent possible consequences of operations, and the possible and the conceivable are one. Just as sensationalism ignores the functional rôle and hypothetical status of sensible qualities in an inquiry, so rationalism makes a fixed and independent matter out of the utility of conceptions in directing inquiry to solve particular problems.

The object of this criticism of historical theories of knowledge is not just to cast discredit upon them. It is to direct attention to the source of their errors. As soon as and whenever it is assumed that the office of knowledge is to lay hold of existence which is prior to and apart from the operations of inquiry and their consequences, one or other of these errors or some combination of both of them is inevitable. Either logical characters belonging to the operations of effective inquiry are read into antecedent existence; or the world as known is reduced to a pulverized multiplicity of atomically isolated elements, a Kantian “manifold”; or some machinery is devised, whether of an “idealistic” or a “realistic” sort, to bring the two together.

When, on the other hand, it is seen that the object of knowledge is prospective and eventual, being the result of inferential or reflective operations which redispose what was antecedently existent, the subject-matters called respectively sensible and conceptual are seen to be complementary in effective direction of inquiry to an intelligible conclusion.

There is another way of discussing the fundamental issue which does not involve so much going over topics worn threadbare by previous discussion. In effect, traditional theories treat all reflective or inferential knowing as cases of “explanation,” and by explanation is meant making some seemingly new object or problem plain and clear by identifying its elements with something previously known, ultimately something said to be known immediately and intuitively, or without inference. In traditional theory, “discursive” knowledge, that involving reflection, must always be referred for its validation back to what is immediately known. It cannot bring its credentials with it and test its results in the very process of reaching them. There is postulated identity implicit or explicit of the results of inference with things known without inference. Making the identity explicit constitutes proof.

There are many different and opposed theories regarding the way in which this identification takes place. There is a doctrine that the operation is one of subsumption of given particulars under given universals; that it is classificatory definition; that it is a kind of Platonic reminiscence in which perceptual material is cognized by being identified with a priori forms; that it is a case of schematization à la Kant; that it is an assimilation of present sensations to images that revive previous sensations. These theories differ widely among themselves; they are irreconcilable with one another. But they all have one premise in common. They all assume that the conclusions of reflective inference must be capable of reduction to things already known if they are to be proved. The quarrel between them is strictly domestic, all in the family. The differences between them concern the character of the original immediately known objects with which the conclusions of reflection must be identified in order to be really known. They all involve the supposed necessity that whatever is a product of inference must, in order to be valid knowledge, be reducible to something already known immediately. Thus they all take the element of knowledge found in inferential conclusions to be simply a matter of restatement.1

The especial significance of the experimental procedure is that it scraps once for all the notion that the results of inference must be validated by operations of identification of whatever sort. When we compare the premise which underlies all the different theories that assume a primitive mode of direct knowledge (knowledge which does not include reflection) with the practice of experimental science, according to which only the conclusion of reflective inquiry is known, we find three marked points of contrast. The first difference is that the traditional theories make all reflective knowledge to be a case of recognition going back to an earlier more certain form of knowledge. The second is that they have no place for genuine discovery, or creative novelty. The third point concerns the dogmatic character of the assumption regarding what is said to be immediately known, in contrast with the experimentally tested character of the object known in consequence of reflection.

We begin with the last point. When it is stated that the conclusions of knowledge involving inference must be subordinated to knowledge which is had directly and immediately, and must be carried to the latter for proof and verification, we are at once struck by the multitude of theories regarding what is immediately and infallibly known. The diversity and contradictions give ground for a suspicion that in no case is the “knowledge” in question as self-evident as it is asserted to be. And there is good theoretical ground for the suspicion. Suppose a man “explains” the eclipse of the moon by saying it is due to the attempt of a dragon to devour it. To him the devouring dragon is a more evident fact than is the darkening of the moon. To us the existence of an animal capable of such a feat is the doubtful matter. It will be objected that it is unfair to take such an absurd case as an instance: dragons are not the sort of thing which any philosopher has asserted to be the object of direct and certain non-inferential knowledge. But the illustration still serves a purpose.

The thing to be known is “explained” by identification with something else. What guarantees this something else? If it too has to be guaranteed by identification with something else, there is an infinite regress. To avoid this regress, we stop short and assert that this or that object or truth is directly known, by sense intuition, by rational intuition, as a direct deliverance of consciousness, or in some other way. But what is such a procedure except the essence of what Bentham called ipse dixitism? What is it but arbitrary dogmatism? Who guards the guardians? The theory which places knowledge in rounded out conclusions is in no such dilemma. It admits the hypothetical status of all data and premises and appeals for justification to operations capable, when they are repeated, of yielding like results. The antecedents do not have to be substantiated by being carried back to earlier antecedents and so on; they are good and sound if they do what is wanted of them: if they lead to an observable result which satisfies the conditions set by the nature of the problem in hand.

The significance of this point comes out more clearly in dealing with the genuineness of discovery or new knowledge. By terms of the traditional theories, this is impossible in the case of inference and reflective inquiry. We know, according to them, only when we have assimilated the seemingly new to something previously known immediately. In consequence, all distinctive individual, or non-repeated, traits of things are incapable of being known. What cannot be treated as a case of something else stays outside knowledge. Individualized characteristics are unknowable surds.

According to this doctrine, reflective inquiry may hit upon new instances of laws, new specimens of old truths, new members of old classes, but not upon intrinsically new objects of knowledge. As far as empiricism is concerned the case of Locke is instructive. His Essay on Human Understanding is one continued effort to test all reflective beliefs and ideas whatever by reduction to original “simple ideas” that are infallibly known in isolation from any inferential undertaking—a point in which many of the new realisms are still Lockeian.

If one looks at the course of science, we find a very different story told. Important conclusions of science are those which distinctly refuse to be identified with anything previously known. Instead of having to be proved by being assimilated to the latter, they rather occasion revision of what men thought they previously knew. The recent crisis in physical science is a case in point. The experimental discovery that the velocity of light remains the same when measured either with or against the direction of the earth’s movement was totally unaccountable on the basis of previous knowing. But scientific men accepted the consequences of their experimental operations as constituting the known object, rather than feeling under obligation to “prove” them by identification with what was said to be antecedently known. Inferential inquiry in scientific procedure is an adventure in which conclusions confound expectation and upset what has been accepted as facts. It takes time for these new facts to be assimilated: to become familiar. Assimilation of the new to the familiar is doubtless a precondition for our finding ourselves at home in the new and being able to handle it freely. But the older theories virtually made this personal and psychological phase of assimilation of new and old into a test of knowledge itself.

The third point, that cognition is recognition, only presents the same difficulty in another way. It presents it in a light which brings out a distinctive point. The theory that knowledge due to reflection consists in identifying something with what is already known or possessed confuses the psychological trait of familiarity, the quality of finding ourselves at ease in a situation, with knowledge. The conception originated when experimental knowing occurred only occasionally and as if by accident; when discoveries were regarded as gifts of the gods or as special inspirations: when men were governed by custom and were uneasy in the presence of change and afraid of the unknown. It was rationalized into a theory when the Greeks succeeded in identifying natural phenomena with rational ideas and were delighted with the identification because their esthetic interest made them at home in a world of such harmony and order as that identification involved. They called the result science, although in fact it fastened wrong beliefs about nature upon Europe for well nigh two thousand years.

Newtonian science, as we have seen in another connection, in effect only substituted one set of identifying objects, the mathematical, for those previously employed. It set up permanent substances, the particles or atoms having inherent mathematical properties, as ultimate realities, and alleged that reflective thought yields knowledge when it translates phenomena into these properties. Thus it retained unimpaired the theory that knowing signifies a process of identification. It required over two centuries for the experimental method to reach a point where men were forced to realize that progress in science depends upon choice of operations performed and not upon the properties of objects which were alleged to be so antecedently certain and fixed that all detailed phenomena might be reduced to them. This conception of knowledge still dominates thinking in social and moral matters. When it is realized that in these fields as in the physical, we know what we intentionally construct, that everything depends upon determination of methods of operation and upon observation of the consequences which test them, the progress of knowledge in these affairs may also become secure and constant.

What has been said does not imply that previous knowledge is not of immense importance in obtaining new knowledge. What is denied is that this previous knowledge need be immediate or intuitive, and that it provides the measure and standard of conclusions obtained by inferential operations. Inferential inquiry is continuous; one phase passes into the next which uses, tests and expands conclusions already obtained. More particularly, the conclusions of prior knowledge are the instruments of new inquiries, not the norm which determines their validity. Objects of previous knowledge supply working hypotheses for new situations; they are the source of suggestion of new operations; they direct inquiry. But they do not enter into reflective knowing by way of providing its premises in a logical sense. The tradition of classic logic persists in leading philosophers to call premises what in effect are regulative and instrumental points of view for conducting new observations.

We are constantly referring to what is already known to get our bearings in any new situation. Unless there is some reason to doubt whether presumptive knowledge is really knowledge, we take it as a net product. It would be a waste of time and energy to repeat the operations in virtue of which the object is a known object unless there were ground for suspecting its validity. Every adult, irrespective of whether he is a man of science or not, carries in his head a large store of things known in virtue of earlier operations. When a new problem comes up, one habitually refers to what is already known to get a start in dealing with it. Such objects, until we have occasion to doubt them, are settled, assured; the given situation is dubious, but they are secure. Hence we take them for granted, we take them as a matter of course. Then if we question them, we tend to fall back upon something else already known. What is too easily overlooked (especially in quest for certainty by attachment to the fixed) is that the objects we thus fall back upon are themselves known in virtue of previous operations of inferential inquiry and test, and that their “immediacy” as objects of reference marks an assured product of reflection. It is also overlooked that they are referred to as instruments, rather than as fixed in and of themselves. The case is similar to the use of tools previously manufactured when we are dealing with the conditions of a new situation; only when they prove defective does invention of new tools demand recurrence to the operations by which they were originally constructed.

This act of taking and using objects already known is practically justified; it is like eating a fruit without asking how it was grown. But many theories of knowledge take this retrospective use of things known in virtue of earlier operations as typical of the nature of knowledge itself. Being reminded of something we already know is taken as the pattern of all knowing. When the thing of which we are now retrospectively aware was in process of being known, it was prospective and eventual to inquiry, not something already “given.” And it has cognitive force in a new inquiry whose objective and ultimate object is now prospective. Taking what is already known or pointing to it is no more a case of knowledge than taking a chisel out of a tool-box is the making of the tool. Because some theories of knowledge have taken the operations that yield the known object to be merely mental or psychical instead of overt redispositions of antecedent subject-matter (and thus have terminated in some form of “idealism”) is no reason for denying the mediated character of all known objects.

Thus we are led by another road to the conclusion that the basic error of traditional theories of knowledge resides in the isolation and fixation of some phase of the whole process of inquiry in resolving problematic situations. Sometimes sense-data are so taken; sometimes, conceptions; sometimes, objects previously known. An episode in a series of operational acts is fastened upon, and then in its isolation and consequent fragmentary character is made the foundation of the theory of knowing in its entirety.

Reflective knowing certainly involves identification. But identity itself has to be defined operationally. There are as many meanings of identity and identification as there are types of operation by which they are determined. There is identification of an object as a member of a class, of a plant as belonging to a certain species: taxonomic identity. The classic theory of definition took this to be the sole valid type of logical definition. There are identifications that are historic, that are concerned with individuals as such. They define the identity of an individual throughout a series of successive temporal changes, while the other type is purely static. This kind of identity is secured by operations that introduce temporal continuity into what is otherwise discrete: it yields genetic and generative definitions. For the identity of an individual is constituted by continued absorption and incorporation of materials previously external—as in the growth of a person, a nation or a social movement. It demands operations that redispose and organize what antecedently exists. Identifications effected by inferential operations are of this type. They are not reductions of the new object or situation to terms of something already known. Traditional theories treat them as if they were of the static and subsumptive type.

Hence these theories have no way of accounting for the discrimination and differentiation, the novel elements, involved in the conclusions of inferential knowing. They must be viewed as mere surds, cognitively speaking. Identifications through processes of temporal growth are, on the contrary, differentiations; new and previously external material is incorporated; otherwise there is no growth, no development. All reflective inquiry starts from a problematic situation, and no such situation can be settled in its own terms. It evolves into a resolved situation only by means of introduction of material not found in the situation itself. Imaginative survey, comparison with things already known, is the first step. This does not eventuate in complete knowledge, however, until some overt experimental act takes place by means of which an existential incorporation and organization is brought about. Merely “mental” revisions remain in the status of thought as distinct from knowledge. Identification through operations that rearrange what is antecedently given is a process of additive discrimination; it alone is synthetic in the true sense of that word, involving likeness-and-difference.

Objective idealisms have insisted upon the conjoint presence of identity and difference in objects of knowledge, as in the doctrine of the “concrete universal.” But they have ignored the phase of temporal reconstruction with its necessity for overt existential interaction.

A further implication of the experimental determination of the known object concerns its office in the verification of hypotheses. It is often supposed that the value of experiment lies merely in the fact that it confirms, refutes or modifies a hypothesis. From the standpoint of the personal interest of the inquirer such an interpretation often holds good. He is interested in a theory, and views the eventually disclosed state of facts solely in its bearing upon the theory he is entertaining. To him, at the time, the cognitive value of the results of experimental operation lies in the test they afford of the claims of his hypothesis. Even so, however, verification, or the opposite, is attained only because experimentation effects a transition of a problematical situation into a resolved one. In this development new individual objects with new features are brought to light. As far as the objective course of knowledge is concerned, as distinct from the personal interest of the investigator, this result is the important one; in comparison with it the verification of a hypothesis is secondary and incidental. The institution of a new object of experience is the essential fact. It would not occur to any one surveying the body of scientific knowledge as a whole to think that its value lay in the corroboration it provides for a number of hypotheses. Taken in the large, the significance of the body of subject-matter as a whole clearly resides in the fact that it marks an added depth, range and fullness of meaning conferred upon objects of ordinary experience.

This consequence is the only intelligible end that can be assigned to processes of reflective inquiry. It marks a gain that during their course hypotheses have gained increased solidity. But the eventual object of activity with tools is not to perfect tools, but is found in what tools accomplish, the products they turn out: When a person working on the basis of a certain idea succeeds in making an invention, his idea is verified. But verification was not the purpose of making the invention, nor does it constitute its value when made. The same may be said of physicians working upon a certain hypothesis in cure of a disease. Only an ultra-specialist would regard a successful outcome simply as verification of a theory. Since a hypothesis is itself instrumental to inquiry, its verification cannot constitute the whole significance of inquiry.

Hypotheses which have later been rejected have often proved serviceable in discovery of new facts, and thus advanced knowledge. A poor tool is often better than none at all. It has even been doubted whether any hypothesis ever entertained has not turned out later to have been erroneous in important respects. It is still questioned whether many of the objects of the most valuable and indispensable hypotheses in present use have actual existence; the existential status of the electron is still, for example, a matter of controversy. In many cases, as in the older theory of the nature of atoms, it is now clear that their worth was independent of the existential status imputed to their subject-matter; that indeed this imputation was irrelevant and as far as it went injurious. As we have seen, progress beyond the Newtonian scheme was made possible when the ascription of antecedently existing inherent properties was dropped out, and concepts were regarded as designations of operations to be performed.

These considerations have a practical importance with respect to the attitude of disdain often affected—usually in behalf of preservation of some dogma—toward the course of science. It is pointed out that scientific men are constantly engaged in furbishing and refurbishing their theories, rejecting those to which they have been devoted, and putting new ones in their place only in time to reject these also. Then it is demanded why we should put our trust in science self-confessed to be unstable rather than in some old dogma which men have continued to believe without change. It is overlooked that the instability affects the intellectual apparatus which is employed, conceptions which are frankly hypothetical. What remains and is not discarded but is added to is the body of concrete knowledge and of definite controls constructed by conceptions no longer tenable. No one would dream of reflecting adversely upon the evolution of mechanical inventions because the sickle had been discarded for the mowing machine, and the mechanized tractor substituted for the horse drawn mower. We are obviously confronted with betterment of the instrumentalities that are employed to secure consequences.

The adverse criticism of science just mentioned attaches only to some of the philosophic interpretations which have been advanced. If scientific conceptions were valid in the degree in which they are revelations of antecedent properties of real Being and existence (as the Newtonian scheme took them to be), there would be something disturbing in their continual revamping. The claim of any one of them to be valid would suffer discredit. Not so, if they are instrumentalities which direct operations of experimental observations, and if the knowledge-property resides in conclusions. Fruits remain and these fruits are the abiding advance of knowledge. Thus the breaking down of the traditional barrier between theory, supposed to be concerned with prior reality and practice concerned with production of consequences, protects the actual results of theory from cavil.

At the same time, it does away once for all with the grounds upon which wholesale skeptical and agnostic philosophies have rested. As long as theories of knowledge are framed in terms of organs assigned to mind or consciousness, whether sense or reason or any combination of the two, organs occupied, it is alleged, in reproducing or grasping antecedent reality, there will continue to exist such generalized skeptical philosophies. Phenomenalism, which holds that impressions and ideas come between the knower and things to be known, will have plenty of support as long as sensations and ideas are supposed to be valid only when they report to mind something prior to them. Phenomenalism may be objected to on the ground that data, ideas, essences, are means of knowing, not its objects. But as long as they are regarded as merely mental means rather than as means which through overt acts effect actual redisposition of antecedent things, the retort will have the character of an arbitrary tour de force; it will be a pious doctrine rather than a conclusion empirically verified.

It is always in place to be doubtful or skeptical about particular items of supposed knowledge when evidence to the contrary presents itself. There is no knowledge self-guaranteed to be infallible, since all knowledge is the product of special acts of inquiry. Agnosticism as confession of ignorance about special matters, in the absence of adequate evidence, is not only in place under such circumstances but is an act of intellectual honesty. But such skepticism and agnosticism are particular and depend upon special conditions; they are not wholesale; they do not issue from a generalized impeachment of the adequacy of the organs of knowing to perform their office. Theories which assume that the knowing subject, that mind or consciousness, have an inherent capacity to disclose reality, a capacity operating apart from any overt interactions of the organism with surrounding conditions, are invitations to general philosophical doubt.

The case stands radically otherwise when it is seen that “mental” states and acts are organs of knowing things not directly but through the overt actions which they evoke and direct. For the consequences of these acts constitute the object said to be known; and these consequences are public and open. Doubt and skepticism attach only to the adequacy of the operations used in achieving the issue which transforms a problematic situation into a settled or resolved one. Instead of being impotent and paralyzing, they are opportunities for bettering concrete methods of inquiry.

Once more, we recur to the problem raised concerning the possibility of carrying over the essential elements of the pattern of experimental knowing into the experience of man in its everyday traits. A statement that judgments about regulative ends and values, the creeds that are to govern conduct in its important interests, are upon the whole matters of tradition, dogma and imposition from alleged authorities, hardly requires argument in its support. It is equally patent that skepticism is rife as to the value of purposes and policies of life thus supplied; the skepticism often extends to complete agnosticism as to the possibility of any regulative ends and standards whatever. The course of human experience in such matters is supposed to be inherently chaotic. Even more precious than the special conclusions of scientific inquiry is its proof that intelligent experimental inquiry is possible which, when it is used, will develop expansion of ideas and regulation of securely tested consequences. It is, once more, a hypothesis rather than a settled fact that extension and transfer of experimental method is generally possible. But like other hypotheses it is to be tried in action, and the future history of mankind is at stake in the trial.

• 1.

  The logic of Stuart Mill is the classic logic of sensational empiricism. Yet he demanded “canons” of proof for induction as rigorous as those of Aristotle were for syllogistic reasoning. The essence of these canons is that proof consists in identification of the results of inference with particulars given in sense, just as with Aristotle demonstration consisted in subsuming them under independently given universals. That the latter was influenced by Euclidean geometry with its assumption of axioms as self-evident truths we have already noticed. Mathematicians now recognize that indemonstrables and indefinables are starting points of operations and that in themselves they have neither meaning nor “truth.”