Again, Dewey says, perception may be translated as the effect of a cause that produced it. But the cause does not ordinarily appear in experience, and the perceptions, as effects, remain isolated from the system of things. Truth and error then become matters of the relation of the perception to its cause. The difficulties attendant upon this view can be avoided by taking sense perceptions as terms in practical judgments. Here the 'other term' which is sought is the action proposed by the perception. "To borrow an illustration of Professor Woodbridge's: A certain sound indicates to the mother that her baby needs attention. If there is error it is not because the sound ought to mean so many vibrations of the air, while as matter of fact it doesn't even suggest air vibrations, but because there is wrong inference as to the act to be performed."[268] The idea is tested, not by its correspondence with some formal reality, but by its ability to lead up to the experience to which it points.
From the consideration of error as cognitive, Dewey passes on to consider its status as primitive sense data. He draws a distinction between sensation as psychological and as logical. Ordinary sensation, just as it comes, is often too confused to serve as a basis for inference. "It has often been pointed out that sense qualities being just what they are, it is illegitimate to introduce such notions as obscurity or confusion into them: a slightly illuminated color is just as irretrievably what it is, as clearly itself, as an object in the broad glare of noon-day."[269] But when a confused object is made a datum for inference, its confusion is just the thing to be got rid of. It is broken up by analysis into simple elements, and the psychologist's sensations are logical products, not psychological facts. "Locke writes a mythology of the history of knowledge, starting from clear and distinct meanings, each simple, well-defined, sharply and unambiguously just what it is on its face, without concealments and complications, and proceeds by 'natural' compoundings up to the store of complex ideas, and the perception of simple relations of agreement among ideas: a perception always certain if the ideas are simple, and always controllable in the case of the complex ideas if we consider the simple ideas and connections by which they are reached. Thus he established the habit of taking logical discriminations as historical or psychological primitives—as 'sources' of beliefs and knowledge instead of as checks upon inference."[270] This way of treating perception found its way into psychology and into empirical logic. The acceptance of the doctrine that all sense involves knowledge, Dewey believes, leads to an epistemological logic; but all perception must involve thought if the 'given' is the simple sensation.
There is nothing especially new in this critique of sensationalism. Historically, sensationalism had been displaced by idealism, and the idea that reality is a construct of ideas held together by logical relations was given up long before functionalism arrived on the scene. But if inference, or rationality, is not present in all experience as the combiner of simple into complex ideas, it may be present in some other form, even more vital. Dewey, however, does not consider such possibilities.
Finally, in an article of slightly earlier date than the studies which have just been considered, Dewey returns to a consideration of metaphysics, and the possibility of a metaphysical standpoint in philosophy. This article, entitled "The Subject-Matter of Metaphysical Inquiry,"[271] deserves careful notice.
The comments of a number of mechanistic biologists on vitalism furnish the point of departure for Dewey's discussion. These scientists hold that, if the organism is considered simply as a part of external nature, as an existing system, it can be satisfactorily analyzed by the methods of physico-chemical science. But if the question of ultimate origins is raised, if it be asked why nature exhibits certain innate potentialities for producing life, science can give no answer. These questions belong to metaphysics, and vitalistic or biocentric conceptions may be valid in the metaphysical sphere.
This raises the question of the nature of metaphysical inquiry. Dewey says that the ultimate traits or tendencies which give rise to life need not necessarily be considered ultimate in a temporal sense. On the contrary, they may be viewed as permanent, 'irreducible traits,' which are ultimate in the sense of being always present in reality. The inquiry and search for these ultimate traits is what constitutes valid metaphysics. "They are found equally and indifferently whether a subject-matter in question be dated 1915 or ten million years B. C. Accordingly, they would seem to deserve the name of ultimate, or irreducible, traits. As such they may be made the object of a kind of inquiry differing from that which deals with the genesis of a particular group of existences, a kind of inquiry to which the name metaphysical may be given."[272]
The irreducible traits which Dewey finds are, in the physical sciences, plurality, interaction, and change. "These traits have to be begged or taken in any case," for wherever and whenever we take the world, we must explain it as "a plurality of diverse interacting and changing existences."[273] The evolutionary sciences add another trait; that is, evolution, or development in a direction. "For evolution appears to be just one of the irreducible traits. In other words, it is a fact to be reckoned with in considering the traits of diversity, interaction, and change which have been enumerated as among the traits taken for granted in all scientific subject-matter."[274]
The doctrine that plurality, interaction, change, and evolution are permanent traits of reality gains in clearness when contrasted with the opposed theories which involve creation, absolute origins, or temporal ultimates. The term 'ultimate origins' may be taken in a merely relative sense which is valid. The French language has an origin in the Latin tongues, which is an ultimate origin for French, but this is not an absolutely ultimate origin, since the Latin tongues, in their turn, have origins. It is, for instance, meaningless to inquire into the ultimate origin of the world as a whole; and it is equally futile to trace any part of the world back to an absolute origin. "That scientific inquiry does not itself deal with any question of ultimate origins, except in the purely relative sense already indicated, is, of course, recognized. But it also seems to follow from what has been said that scientific inquiry does not generate, or leave over, such a question for some other discipline, such as metaphysics, to deal with."[275]
Theories like that of Laplace, for instance, trace the world back to an origin in some undifferentiated universe; or, in Spencer's terms, some state of homogeneity. From this original state the world is said to evolve. But the undifferentiated mass lacks the plurality, interaction, and change which are presupposed in all scientific explanation. These traits must be present before development can occur. "To get change we have to assume other structures which interact with it, existences not covered by the formula."[276] In short, although Dewey only implies this, all scientific explanation presupposes a system of interacting parts; nothing can be explained by reference to an undifferentiated world which lacks such traits.
Dewey is particularly interested in the origin of mind or intelligence. In dealing with mind, he says, we must begin with the present, and in the present we find that the world has an organization, "in spots," of the kind we call intelligence. This existing intelligence cannot be explained by any theory which reduces it to something inferior. The "attempt to give an account of any occurrence involves the genuine and irreducible existence of the thing dealt with."[277] Mind cannot be explained by being explained away, nor can it be explained as a development out of an original source in which the potentiality, or direction of change towards mind, was lacking.