CHAPTER II. The Problem of Soul and Body

THE PROBLEM OF SOUL AND BODY

§ 1. The problem of psychophysical connection has to do with the correlation of scientific abstractions, not of given facts of experience. § 2. The “consciousness” of Psychology is thus not the same thing as the finite individual subject of experience, and Reality must not be said to consist of “minds” in the psychologist’s sense. Again, we must not assume a priori that there can be only one working hypothesis of psychophysical connection. § 3. The possible hypotheses may be reduced to three, Epiphenomenalism, Parallelism, and Interaction. § 4. Epiphenomenalism is legitimate as a methodological principle in Physiology; it is untenable as a basis for Psychology because it implies the reduction of psychical facts to mechanical law. § 5. Parallelism. The arguments for Parallelism as necessarily valid to Psychophysics because of its congruity with the postulates of mechanical Physics, are fallacious. We cannot assume that Psychology must necessarily conform to these postulates. § 6. As a working hypothesis Parallelism is available for many purposes, but breaks down when we attempt to apply it to the case of the initiation of fresh purposive reactions. A teleological and a mechanical series cannot ultimately be “parallel.” § 7. We are thus thrown back on the hypothesis of Interaction as the only one which affords a consistent scheme for the correlation of Physiology and Psychology. We have, however, to remember that what the hypothesis correlates is scientific symbols, not actual facts. The actuality represented by both sets of symbols is the same thing, though the psychological symbolism affords a wider and more adequate representation of it than the physiological.

§ 1. Few questions have more constantly attracted the attention of philosophers, especially perhaps of those philosophers who have lived since the establishment of Christianity as the religion of the Western world, than that of the relation between the soul or mind and the body; and perhaps no question has given rise to graver misconceptions for want of a correct insight into the true logical character of the problem under discussion. Both in the half-scientific speculations of ordinary persons and in the more systematic theories of metaphysicians and psychologists, the subject is constantly approached under the totally erroneous preconception that the dualistic separation of human life into a bodily and a mental part or aspect is a datum of immediate experience which we can directly verify in ourselves, and that the task of philosophy is by ingenious but unverifiable hypothesis to transcend this chasm between given realities. From the standpoint of our previous chapter we can easily see that such a view fundamentally misrepresents the real philosophical problem.

So long as we are concerned with human existence as we directly find it in our immediate experience, or assume it in our practical social relations with our fellows, no question of the relation between body and mind can arise, because neither term of the relation is as yet before us. For my own immediate experience I am neither a body nor a soul, nor yet a composite of the two, but simply an individual subject of experiences in direct intercommunion with other individuals. Under the influence of conscious or unconscious dualistic prepossessions, we often speak as if it were a directly experienced fact that I can communicate with my fellow-subjects only indirectly through the medium of an alien “material” body, and we sometimes contrast this supposed restriction with an imagined higher state of existence, in which “disembodied spirits” may conceivably have direct intercourse with each other. But the truth is, that this direct intercourse and influence of one intelligent and purposive individual on another is no privilege reserved for our enjoyment in “a better world than this”; it is, as we can see if we will only forget our dualistic prepossessions, the very truth about our actual life. In actual life, before we have contaminated our direct enjoyment of it with psychological prejudices, we know nothing of the interposition of an inert “material” organisation between ourselves and the members of our social environment. The severance of the original unity of experience into a physical and a psychical aspect is entirely a product of our own abstraction-making intellect. “Body” and “soul” are not given actualities of experience, but artificial mental constructions of our own derived from the actual “facts” of life by the elaborate processes which we have just been studying.

As we have seen in constructing our concept of a mechanical physical order, we abstract certain elements of our direct experience from the whole, and consider them under the name of our “bodies” as if they had a separate existence; we then, by the aid of the hypothesis of “introjection,” represent those elements of direct self-experience which were omitted from the physical order as forming by themselves a second distinct whole or system called the “soul.” When we have reached this point, we are, of course, compelled to raise the question how these two systems, the bodily and the mental, must be supposed to be connected. But the important fact to remember is that the two systems are not facts of experience, but products of abstraction. Our task in discussing their relation is not to transcend a given dualism, but to get rid of one which we have manufactured for ourselves by the manipulation of experience in the interests of certain special scientific problems. Hence, as Münsterberg well puts it, we have not to find the connection which subsists, as an actual fact, between body and soul, but to invent a connection in keeping with the general scheme of our artificial physical and psychological hypotheses.

§ 2. As far as the interests of Metaphysics are concerned, this recognition that the problem of soul and body has to do solely with highly artificial products of scientific abstraction, and not with anything which can be called a “given” actuality, is the one principle of supreme importance which emerges from the discussion of the subject. Two very significant inferences may at once be drawn from it. (1) We clearly must not call the finite subjects of experience, of whom we saw reason to hold that ultimate Reality is exclusively constituted, “minds” or “souls” in the psychologist’s sense. To call them so would inevitably be to imply that exclusion from the physical order of “bodies” apart from which the psychological concept of the “soul” or “mind” has no significance. Or, in other words, it would identify them not with what they are for their own direct experience, but with what they become for one another’s theoretical reflection under the influence of “introjection.” As we have seen, it is legitimate and necessary for special scientific purposes to treat ourselves and other individuals as if we were such series of “mental states,” but it is never legitimate to forget that, when we do this, we are substituting a highly unreal symbolism for directly experienced facts.

One consequence of confusing the symbolism with the fact may be noted in passing: when we have substituted the series of mental states for the felt unity of actual conscious life, we go on to ask ourselves how the fact and its symbol—the symbolic nature of which we have forgotten—are related. And thus arise all the unanswerable, because fundamentally unmeaning, questions as to the way in which the “self” has or owns the succession of “states.” Failing to see that the succession of states is simply the unitary subject itself, as it appears from the point of view of the “introjection” hypothesis, we then find ourselves confronted by the alternatives of foisting upon our Psychology the useless and unthinkable fiction of a changeless “substratum” of mental states—the soul-substance of the pre-Kantian psychologists—or resolving real life into a succession of discontinuous “mental images.” With the recognition that Psychology never deals directly with experienced reality, but always with the hypothetical products of an abstraction which is only justified by its usefulness for the special purposes of the psychologist, all these difficulties disappear.

(2) Another important consequence of our principle is that we cannot dogmatically assert that there can be only one legitimate theory of the “connection between mind and body.” If “mind” and “body” were really given as distinct but connected in direct experience, it might well be that there could only be one account of their connection answering to experienced fact. But since the separation is itself of our own intellectual manufacture, as we are dealing throughout with artificial creations of our own abstraction, any theory of their connection which is desirable for the solution of a special problem or class of problems will be legitimate for that particular class of problems. Thus the physiologist may legitimately, if it answers his special purposes, adopt a working hypothesis which the psychologist may find untenable, and again different types of psychological problem may legitimately assume different working hypotheses. I shall aim at showing in the immediately following paragraphs that there is one typical psychophysical hypothesis which, on the whole, lends itself better than its rivals to the general purposes of both Physiology and Psychology, but we shall see, as we proceed, that the hypotheses we reject are also legitimate for the solution of important special problems. In fact, our chief interest, as students of Metaphysics, in the further discussion of psychophysical connection will be to point out the fallaciousness of the metaphysical arguments which are commonly used to establish some one hypothesis as necessarily and exclusively true.

§ 3. Turning now to consider the chief types of hypothesis which have been, or are at present, actually put forward by metaphysicians and psychologists, we may perhaps group them under the five main heads of (1) Pre-established Harmony, (2) Occasionalism, (3) Epiphenomenalism, (4) Psychophysical Parallelism, (5) Interaction. For our purpose in the present chapter the number of alternatives may be further reduced by the omission of the first two. Neither the Pre-established Harmony of Leibnitz nor the Occasionalism advocated by Geulincx and Malebranche, and in a one-sided form by Berkeley, is likely to find much support from the philosophy of the present day. Both doctrines are, moreover,—that of Leibnitz avowedly and that of the Occasionalists by implication,—much more than special psychophysical hypotheses. They are in principle attempts to get rid of all transeunt causality, and have been discussed in their general bearings in our chapter on the Causal Postulate, where we satisfied ourselves that any science which recognises, as Psychology has to do, the existence of finite things must also admit the principle of transeunt causality, at any rate as a working hypothesis.

Each of the three remaining types of view has its supporters among contemporary students of science and philosophy. The epiphenomenalist theory is largely adopted by the workers in the physical sciences, and though not much countenanced by psychologists and metaphysicians, has the explicit support of Dr. Shadworth Hodgson, while some versions of the parallelist doctrine, notably that of Münsterberg, approach it very closely. The parallelist hypothesis is perhaps at present the most popular among the psychological specialists, and is represented by writers of such eminence as Wundt, Münsterberg, Ebbinghaus, Höffding and Stout. Finally, Interaction has powerful champions in Bradley, Ward, and James; to say nothing of its adoption by so sound a physiologist as Mr. McDougall. Both the latter doctrines, again, have historical connections with the great philosophical systems of the past, Parallelism with that of Spinoza, and Interaction with those, to mention no other names, of Descartes and Locke. In the philosophy of the ancient world the psychophysical issue can hardly be said to appear in a well-defined form, but we may perhaps state that Plato’s psychological doctrine is decidedly one of Interaction, while the view of Aristotle, though too complex to admit of very precise formulation, inclines rather towards Parallelism.

§ 4. Epiphenomenalism. Of the three hypotheses which remain for discussion, the theory of Epiphenomenalism has the least to recommend it, and is open to the most serious objections. According to this view, all causal connections are exclusively between physical states. Bodily changes succeed one another in accord with uniform laws of sequence, which it is the province of the physiologist to discover, and every bodily change is completely determined by bodily antecedents. Certain bodily conditions are further attended by corresponding “states of consciousness,” but those states stand in no causal connection with subsequent bodily states, nor yet with one another. They are thus consequences or effects, but are never causes. The whole series of physical changes, from birth to death, which makes up the history of the human body, goes on precisely as it would if “consciousness” were entirely absent. This is what is meant by the assertion that all mental states are epiphenomena, superfluous accessories, which arise in the course of the connected series of bodily changes, but are entirely without any determining influence upon it. The doctrine may be diagrammatically represented thus

a——α |╭──╮ b——β |╭──╮ c——γ,

where the italic letters symbolise physical and the Greek letters psychical states, the vertical lines indicating the course of causal sequence.

If a psychophysical hypothesis were ever directly applicable to the actualities of experience, we might, of course, dismiss Epiphenomenalism at once as inherently absurd. For nothing is more certain than that in the actual life of direct experience our knowledge and our interests do determine the course of our actions. That what we believe and desire does make all the difference in the world to the way in which we behave, is one of those elementary verities out of which no scientific hypothesis can claim to reason us. Hence, when the defenders of the theory attempt to draw practical moral and juristic consequences from their doctrines, we are within our rights in simply declining to concern ourselves with so absurd a travesty of the simplest facts of experience. So long, however, as the hypothesis is put forward simply as a working hypothesis for the correlation of our physiological and psychological theories, the case is different. Its validity as a psychophysical theory must be estimated solely by the degree in which it renders this systematic correlation feasible, and is not necessarily impaired by the manifest absurdities which result from mistaking the doctrine for a description of actual life.

Now, if we look at the hypothesis from this point of view, we can at once see that it is really legitimate for some purposes. For the purpose of physiological science it is obviously to our interest that we should be able to deduce the later from the earlier stages of a physiological process. We have thus an interest in treating physiological changes, if we can, as unconditioned by any but physiological antecedents. And every actual success in establishing a uniformity or “law” of Cerebral Physiology is proof that the assumption that, for the process in question, the only determining conditions which count are physiological, is equivalent to the truth. The physiologist, then, is clearly justified in treating the psychical series as epiphenomenal, if he means no more by this than that he intends to deal, as a physiologist, only with processes which can be successfully resolved into uniform sequences on the assumption that they involve only physiological terms. Though whether any processes in the nervous system can be successfully treated as purely physiological sequences, nothing but the physiologist’s actual success in obtaining results from his initial postulate can decide.

If, however, the physiologist should go on, as he sometimes does, to make the assertion that not only can some nervous processes be treated as if their psychical accompaniments made no difference, but that they really are what they would be without those accompaniments, or even that all nervous process is what it would be without “consciousness,” he commits a gross logical fallacy. It is a mere blunder in logic to argue that because the presence of certain circumstances makes no difference to the special result which follows on a given antecedent, the result would equally follow in their absence. For it might be that in their removal the very antecedents in which we are interested would disappear. We are not at liberty to infer that, because the course of certain physiological processes can be computed without taking their mental correlates into account, they could occur apart from those correlates.

Even more serious are the consequences which follow when it is assumed that all mental processes without exception may be regarded as epiphenomenal, i.e. that all human action, if only our Physiology were sufficiently advanced, might be brought under laws of purely physiological sequence. Such an assumption would lead at once to the following dilemma: Either our Physiology must remain rigidly faithful to the fundamental postulates of mechanical science, or not. If it is faithful to them, its descriptions of human action must rigidly exclude all reference to teleological determination by reference to conceived and desired ends. I.e. we must treat human conduct as if it were fatally determined apart from any possible influence of human choice and intention, and thus stultify that whole work of historical and ethical appreciation which we have already seen to be the principal raison d’être of Psychology as a science. We must revert, in fact, to a theory of life which is identical with the extremest forms of Pagan or Mohammedan fatalism in everything except the name it gives to its ineluctabile fatum. Or, if we are not prepared to do this, we must allow Physiology itself to use the psychological categories of desire, selection, and choice, and thus covertly admit that human action, after all, cannot be described without the introduction of factors not included in the physical order. It is no doubt due to their realisation of this dilemma that psychologists are all but universally agreed to reject the epiphenomenalist hypothesis, while its popularity with physiologists may be explained by observing that physiological uniformities can manifestly only be successfully established for those processes which can be treated as if they were only physiologically conditioned.

§ 5. Parallelism. The hypothesis of Parallelism attempts, while preserving some of the characteristic features of the cruder view just described, to avoid its unsatisfactory consequences. Agreeing with Epiphenomenalism in the doctrine that physiological changes must be treated as determined only by physiological antecedents, Parallelism denies that the events of the psychical series are mere “secondary” effects of their physiological correlates. According to it, the series of physical and that of psychical events are strictly “parallel,” but not causally connected. Each event in either series has its precise counterpart in the other, but the physical events do not cause the psychical events, nor vice versâ. The successive members of the physical series form a connected causal sequence, independent of their psychical concomitants, while these latter, it is generally assumed, form a similar chain of causally connected psychical states. Thus every nervous change is determined solely by precedent nervous changes, and the corresponding psychical change by the corresponding antecedent psychical changes. In diagrammatic shape our hypothesis now takes the form

╭──╮ a α | | b β | | c γ

Usually it is further added that the ultimate metaphysical explanation of this parallelism without mutual dependence must be found in the (Spinozistic) doctrine of Identity, i.e. the doctrine that the physical and psychical series are two different “sides” or “aspects” of a single reality. Some supporters of Parallelism (e.g., Ebbinghaus) conceive this single reality as a tertium quid, equally adequately expressed by both the series, others (e.g., Stout) hold that its real nature is more adequately revealed in the mental than in the physical series.

The grounds commonly adduced in favour of the parallelistic view as the most satisfactory psychophysical theory, are of two kinds. As a positive argument it is urged that cerebral anatomy has already to some extent confirmed the doctrine of correspondence between definite physical and psychical processes by its successful “localisation” of specific sensory and motor processes in various cortical “centres,” and may reasonably be expected to accomplish further such “localisations” in the future. Stress is also laid upon the formal analogy between the psychological laws of retentiveness, association, and habit, and the physiological theories of the formation of “conduction-paths” in the brain. These positive contentions do not, however, take us far. The correspondences upon which they rest, so far as they are ascertained experimentally and are not mere deductions from the principle of Parallelism itself, would be equally natural on a theory of Interaction, or of one-sided dependence of either series on the other. The real strength of the case for Parallelism rests upon certain negative assumptions which are widely believed to exclude the hypothesis of causal dependence of either series on the other. These negative assumptions appear to be in the main three.

(1) It is said that, while we can without difficulty conceive how the later stages of a continuous physical or psychical process can be connected by causal law with its earlier stages, we are entirely unable to conceive how psychical events can arise from physical antecedents, or vice versâ, because of the utter disparateness of the physical and the psychical. The physical process, it is urged, is continuous, and so, on the other side, is the psychical, but when we attempt to think of a cerebral change conditioning a mental change, or vice versâ, there is a complete solution of continuity which we cannot bridge by any causal formula.

(2) The doctrine of Conservation of Energy is sometimes supposed to be incompatible with the admission of psychical states among the antecedents or consequents of physical states. It is said that if psychical states can influence the course of nervous change, there will be “work” done in the organism without the expenditure of energy, and if the total effect of nervous change is not exclusively physical there will be loss of energy without “work” being done by the organism, and in either case the principle of Conservation will be contravened.

(3) Finally, it is maintained that it is a fundamental postulate of the physical sciences, that every change of configuration in a material system such as the living organism is assumed to be, is due to exclusively physical antecedents, and that this postulate must therefore be respected in Psychophysics. These are, so far as I can gather them from the works of the psychologists who adopt the parallelist view, the principal arguments by which their case is supported.

It is clear that if all—or any—of these contentions are valid, it must follow that Parallelism is not only a legitimate but the only legitimate hypothesis for the co-ordination of physical and psychical science. I believe, however, that every one of them is fallacious, and that for the following reasons:—

(1) The argument from the inconceivability of causal relation between the physical and the psychical is perhaps the most effective of the alleged grounds for denying interaction between the psychical and the physical. Yet its force is not really so great as it might appear. It is not denied that we can, in simple cases, assign the conditions under which a mental state follows on a physical state (e.g., we can assign the physical conditions of the emergence of a given sensation). But, it is argued, we cannot show why those conditions (e.g., the stimulation of the retina, and indirectly of the “optical centres” in the brain by light of a given wave length) should be followed by this particular sensation (e.g., green, and not some other colour). This means that we cannot construct a mathematical equation connecting the character of the sensation with that of the stimulus, as we can to connect the earlier with the later stages of a purely physical process. This is, of course, obvious enough. It is only by making complete abstraction from the appearance of new qualities in the course of a process, and by treating it as a purely geometrical and quantitative transformation, that we can render it amenable to our equations.

As we saw in our discussion of Causality, mathematical Physics only succeeds in its constructions on the condition of excluding all qualitative change, as “subjective,” from its purview. But we also saw there that the origination of the qualitatively new is an essential part of the idea of Causality, and that in reducing all change in the physical world to quantitative transformation, mathematical Physics really does away with the causal concept. We are, in fact, in precisely the same logical position if we speak of physiological changes as causing sensation, as when we speak of a quantitative change in the proportions of a chemical compound as the cause of alteration in its qualities. The objection that the psychical effect cannot be connected by an equation with its alleged cause, would hold equally in any case of the production of the qualitatively new, i.e. in every case where we use the category of causality at all. And for that very reason it has no force when urged as an objection to psychophysical causality in particular.

(2) The argument from the Conservation of Energy may be more briefly dismissed, as its fallacious character has been fully recognised by the ablest recent exponents of the parallelistic view, such as Dr. Stout and Professor Münsterberg. As Dr. Stout points out, the argument involves a formal petitio principii. The principle of Conservation of Energy has only been established for what are technically known as conservative material systems, and no absolute proof has been given, or seems likely to be given, that the human organism is such a conservative system. Further, as has been urged by many critics, and notably by Professor Ward, the principle of conservation, taken by itself, is simply a law of exchanges. It asserts that the quantity of the energy of a conservative system remains constant under all the transformations through which it passes, but, apart from the rest of the postulates of mechanical science, it affords no means of deciding what transformations of energy shall occur in the system, or when they shall occur. Hence there would be no breach with the special principle of Conservation of Energy if we were to assume that psychical conditions can determine the moment at which energy in the organism is transformed e.g., from the kinetic to the potential state, without affecting its quantity.

(3) It is, however, true that it is inconsistent with the postulates of mechanical Physics, taken as a whole, to admit the determination of physical sequences by non-physical conditions. To admit such determination would be to stultify the whole procedure of the mechanical sciences. For, as we have seen in our Third Book, the primary object of mechanical science is to reduce the course of events to rigid laws of uniform sequence, and thus to facilitate the formulation of practical rules for our own interference with it. It is therefore a legitimate postulate of mechanical science that—for its special object—desire and will shall be excluded from our conception of the conditions which determine events, and the whole course of nature treated as if conditioned only by physical antecedents. If there is any department of experienced reality which cannot be successfully dealt with according to these postulates, then the formulation of rigid laws of uniform sequence is, in principle, impossible for that department, and it must be excluded from the “world” which mechanical science investigates.

But the fact that mechanical science can only attain its end by treating all physical events as independent of non-physical conditions, does not afford the slightest presumption that they must be treated in the same way for all purposes and by every branch of inquiry. Whether Psychology, in particular, is under the logical necessity of conforming to the mechanical postulates, will depend upon our view as to whether the object subserved by Psychology is the same as that of the mechanical sciences, or different. If our purpose in psychological investigation is not identical with the purposes of mechanical science, there is no sense in demanding that we shall hamper our procedure as psychologists by adherence to postulates based upon the special nature of the interests to which mechanical science has to minister.

Now, we have already contended that the aims of Psychology only partially and temporarily coincide with those of the mechanical sciences. If we were right in holding that the principal object of Psychology is to provide a general terminology of which History and Ethics can avail themselves in their appreciations of life, it follows at once that Psychology imperatively needs the recognition of that very teleological aspect of human action which is excluded on principle, and rightly so for the special purpose of mechanical Physics, by the fundamental mechanical postulates. Thus the argument that the parallelistic hypothesis must be the most suitable for the psychologist, because it conforms to the mechanical postulates of sciences which deal with experience from a different standpoint and in a different interest, loses all its cogency.

Now that we have, as I trust, sufficiently disposed of the a priori arguments for the parallelistic view, we are in a position to estimate it, as a psychological hypothesis, purely on its merits as evinced by its actual success. But first we must point out once more that the whole question is not one as to actualities, but purely as to the most satisfactory way of bringing two sets of abstractions, originally devised for divergent purposes, into touch with one another; and further, that if the hypothesis were put forward as a final metaphysical truth about the constitution of the real world, it would be manifestly self-contradictory.

In the first place, Parallelism, taken for anything more than a convenient working hypothesis, would involve a flagrant breach of logic. It is obvious that, as Mr. Bradley has urged, you cannot infer from the premisses that one total state, containing both a physical and a psychical element, causes another complex state of the same kind, the conclusion that the physical aspect of the first, by itself, has caused the physical, and the psychical the psychical aspect of the second. To get this conclusion you need a “negative instance,” in which either the physical or the psychical state is found apart from its correlate, but followed by the same consequent as before, and Parallelism itself denies the possibility of such an instance. From the premisses that a α is always followed by b β, it attempts to infer, without any “dissection of nature,” that a by itself was the necessary and sufficient condition of b, and α of β. And this is, of course, logically fallacious. Dr. Ward expresses the same point differently when he urges that unvarying and precise concomitance without causal connection is a logical absurdity.

That the supporters of the hypothesis themselves are conscious of the difficulty, is shown by their unanimous assertion that the psychical and physical series are ultimately manifestations of one and the same reality. What they do not explain is how, if this is so, the two series can be phenomenally so utterly disparate as to exclude mutual influence on one another. The difficulty becomes insuperable when we reflect that on the parallelistic view the physical series must be rigidly mechanical, as otherwise we shall have a breach with those mechanical postulates which are supposed to require the exclusion of psychical states from the determining conditions of physical occurrences. Thus, if teleology is to be recognised anywhere in our scientific constructions, it must be in our conception of the psychical series. And on the whole the supporters of Parallelism admit this in practice by the free use of teleological categories in their Psychology. But it ought by now to be clear to us that the nature of the identical reality cannot be expressed with equal adequacy in a teleological series, and in one which is, by the principles of its construction, purely mechanical. Here, again, most of the parallelists are really in agreement with us, for they usually in the end call themselves “Idealists,” and assert that the “mental” series is a more faithful representation of Reality than the physical. But if the two series are not on the same level in respect of their nearness to Reality, it is hard to see how there can be exact correspondence between them. This is a point to which we shall immediately have to return.

§ 6. When we ask, however, whether Parallelism, apart from these questions of ultimate philosophy, is legitimate as a working hypothesis in Psychology, the answer must be that, in certain departments of psychological investigation, it certainly is so. In practice, the doctrine of the parallel but independent series amounts, for the most part, to little more than a methodological device for the division of labour between the physiologist and the psychologist, the physiologist restricting himself to the formulation of such uniformities as can be established between nervous processes, considered as if independent of external influence, and the psychologist doing the same for their psychical accompaniments. As a principle of methodical procedure, therefore, in those parts of Psychology which deal with the more passive and, as we may say, routine-like aspects of mental life, Parallelism is a useful and therefore a legitimate working hypothesis.

The question by which its claim to be the best hypothesis must be decided is, to my mind, that of its applicability to the case of the fresh initiation of new purposive adaptations to changes in the organism’s environment. For it is just in dealing with these cases that Psychology, if it is to fulfil the purpose we have ascribed to it, must most obviously discard mechanical for teleological categories. Hence it is here, if anywhere, that a difficulty of principle must make itself felt when we attempt to treat the psychical and the physical series as exactly parallel and corresponding. It seems to follow necessarily from the conception of physical science as based upon the mechanical postulate, that a teleological and a mechanical series cannot possibly run “parallel” in all their details in the fashion presupposed by the hypothesis under consideration.

If Psychology is to be of any use in supplying Ethics and History with the subject-matter of their appreciations, it is manifest that it must make the assumption that desire and choice are operative in determining the course of human action, and thus must—at certain points at least—explicitly employ the categories of teleology. These categories, again, cannot possibly be translated into the rigidly non-teleological symbolism of a physical science, based upon the mechanical postulates, as every science of “general laws” must be. It follows that “exact parallelism without mutual interference” cannot, consistently with the purpose which Psychology subserves, be employed, even as a working hypothesis throughout the whole field of psychological investigation itself. When the attempt to extend its employment to the whole sphere of psychical processes is seriously made, it leads inevitably to the crude fatalism of the doctrine that there is no such thing as choice or action (free or otherwise) in the universe. In actual practice, the supporters of Parallelism, who reject this doctrine when it is explicitly avowed under the name of Epiphenomenalism, only succeed in doing so because they do not really insist on carrying out the parallelistic hypothesis in their Psychology. They commonly make their hypothesis prominent, while they are dealing with the comparatively passive and routine-like aspects of mental life, association, habituation, etc., but allow themselves to lose sight of it as soon as they come to treat of such explicitly teleological concepts as attention and choice. Their procedure is also rendered easier for them by the liberal use which evolutionary biologists, even while professing with their lips fidelity to the mechanical postulates, allow themselves to make of teleological categories which are really purely psychological.

It would be an easy task, if space permitted, to show in detail how the fundamentally different principles underlying the construction of the mechanical and the teleological series involve the presence, in the individual members of each series, of characters to which nothing corresponds in those of the other. Thus we might ask, with Dr. Ward, what corresponds in the psychical scheme to the composition of the units of the physiological scheme out of their various chemical components, and of these, again, out of more elementary physical “prime atoms”? or, from the opposite side, we might ask, what is the cerebral equivalent, in terms of a rigidly mechanical Physiology, of the psychological character of “meaning” or “significance”? But the multiplication of these problems becomes superfluous if the reader has once grasped our principle, that exact correspondence is only possible between series which are either both mechanical or both—and both in the same degree—teleological. Between a genuinely teleological and an honestly mechanical series such correspondence is logically impossible, because of the fundamental difference between their types of construction.

§ 7. For the reasons just produced, it is, I think, necessary to hold that the oldest and simplest hypothesis of the connection between body and mind, that of Interaction, is after all the most satisfactory. According to this view, the two series cannot be thought of as presenting an exact correspondence, and must be thought of as causally influencing each other at different points, precisely as any two sets of physical events do. If we adopt it we shall recognise in sensation a psychical state which has physical processes among its immediate antecedents, and in motor reaction similarly a physical process with psychical antecedents. It is scarcely to be denied that this conception of body and mind, as two things which stand in causal relation, is the hypothesis which most naturally presents itself, when once we have artificially broken up the unity of immediate experience into a physical and a psychical side, and so created the problem of psychophysical connection. So natural is it, that even psychologists who accept one of the other hypotheses are to be found constantly speaking of voluntary movement in terms which, if they mean anything, imply causal determination of bodily by mental process, while no psychologist of any school has ever succeeded in expressing the relation of sensation to stimulus in any other phraseology than that of Interaction. Probably the hypothesis would never have been exposed to hostile criticism at all, but for the metaphysical objections, already dismissed by us as fallacious, founded upon the notion that the mechanical postulates with which Interaction conflicts are ascertained truths about the actual structure of the reality with which we are in touch in immediate experience.

It is clear that, from the nature of the problem to be solved, we cannot be called upon to prove the actual occurrence of psychophysical interaction. As a working hypothesis for the interrelation of two sets of scientific abstractions, the theory is in principle incapable of direct establishment by the “appeal to facts.” All that is requisite for its justification is to show that it is (a) not in principle at variance with any fundamental axiom of scientific procedure, and (b) enables us to co-ordinate our scientific results in the manner most suitable for the uses to which we propose to put them. Both these conditions are fulfilled by the hypothesis of Interaction, if our foregoing arguments are sound. We have seen the fallacious nature of the objections brought against it on a priori grounds of logical method, and have also seen that it is positively demanded if we are at once to be faithful to the mechanical postulates upon which physical science depends for its successes, and to recognise in our psychological constructions that teleological character of human action which is all-essential for History and Ethics. In substance this is the whole case for the Interaction hypothesis, and no further accession of strength would result from its elaboration in detail.

It may be added that it is one great recommendation of the hypothesis of Interaction, that it is quite consistent with the full recognition of the relative usefulness of the alternative theories, though they, as we have seen, are unable to do justice to those aspects of fact which can only be expressed in terms of Interaction. Thus the hypothesis of Interaction can readily afford to admit that, for certain purposes and up to a certain point, it is possible to treat physical or psychical processes as if they were determined solely by physical or psychical conditions respectively, and even to treat some physical processes as if the presence of their psychical concomitants made no difference at all to their occurrence. The reason of this is, that whereas a mechanical hypothesis can give no intelligible account of a purposive process at all, a teleological hypothesis can quite easily account for the apparently mechanical character of some of the processes which fall under it. As we have seen (Book III. chap. 3, § 6), a purposive reaction, once established, approximates to mechanical uniformity in the regularity with which it continues to be repeated, while the conditions are unchanged, and the end of the reaction is therefore still secured by its repetition.

Thus we can readily see that, even if we contented ourselves with the attempt to translate into the language of psychological science the processes which make up the life of an individual subject, many of them would appear to be going on with routine uniformity. And when we deliberately set ourselves to obtain uniformities by taking an average result, derived from comparison of a multitude of subjects, our results are, of course, always mechanical in appearance, because the element of individual purpose and initiative has been excluded by ourselves from our data in the very process of taking the average. Hence we can understand how, on the hypothesis of Interaction itself, all those mental processes which consist in the repetition of an already established type of reaction should come to appear mechanical, and thus to suggest that mechanical conception of psychical processes which is common to the epiphenomenalist and the parallelist view. Interaction, and Interaction alone, is thus a hypothesis capable of being applied to the whole field of psychological investigation.

I will conclude this chapter with some considerations on the bearing of our result upon the special problems of Metaphysics. We have explicitly defended Interaction as being no statement of actual experienced fact, but a working hypothesis for the convenient correlation of two scientific constructions, neither of which directly corresponds to the actualities of experience. This means, of course, that Interaction cannot possibly be the final truth for Metaphysics. It cannot ultimately be the “fact” that “mind” and “body” are things which react upon each other, because, as we have seen, neither “mind” nor “body” is an actual datum of experience; for direct experience and its social relations, the duality subsequently created by the construction of a physical order simply has no existence. Nor can it be maintained that this duality, though not directly given as a datum, is a concept which has to be assumed in order to make experience consistent with itself, and is therefore the truth. For the concept of Interaction manifestly reposes upon the logically prior conception of the physical as a rigidly mechanical system. It is because we have first constructed the notion of the “body” on rigidly mechanical lines that we have subsequently to devise the concept of “mind” or “soul” as a means of recognising and symbolising in our science the non-mechanical character of actual human life. And since we have already seen that the mechanical, as such, cannot be real, this whole scheme of a mechanical and a non-mechanical system in causal relation with one another can only be an imperfect substitute for the Reality it is intended to symbolise. In fact, we might have drawn the same conclusion from the very fact that the psychophysical hypothesis we have adopted is couched in terms of Transeunt Causality, since we have already satisfied ourselves that all forms of the causal postulate are more or less defective appearance.

The proposition that the psychophysical theory of the “connection” of “body” and “mind” is an artificial transformation, due to the needs of empirical science, of the actual teleological unity of human experience, is sometimes expressed by the statement that mind and body are really one and the same thing. In its insistence upon the absence of the psychophysical duality from actual experience, this saying is correct enough, but it perhaps fails to express the truth with sufficient precision. For, as it stands, the saying conveys no hint of the very different levels on which the two concepts stand in respect to the degree of truth with which they reproduce the purposive teleological character of real human experience. It would perhaps be nearer the mark to say that, while the physiologist’s object, the “body,” and the psychologist’s object, the “mind,” are alike conceptual symbols, substituted, from special causes, for the single subject of actual life, and may both be therefore said to “mean” or “stand for” the same thing, their actual content is different. For what in the language of physiology I call my “body” includes only those processes of actual life which approximate to the mechanical ideal sufficiently closely to be capable of being successfully treated as merely mechanical, and therefore brought under a scheme of general “laws” of nature. Whereas what, as a psychologist, I call my “mind” or “soul,” though it includes processes of an approximately mechanical type, includes them only as subordinate to the initiation of fresh individual reactions against environment which can only be adequately expressed by teleological categories. Thus, though “mind” and “body” in a sense mean the same actual thing, the one stands for a fuller and clearer view of its true nature than the other. In Dr. Stout’s terminology their intent may be the same, but their content is different.

Consult further:—R. Avenarius, Der Menschliche Weltbegriff; B. Bosanquet, Psychology of the Moral Self, lect. 10; F. H. Bradley, Appearance and Reality, chap. 23; Shadworth Hodgson, Metaphysic of Experience, vol. ii. pp. 276-403; William James, Principles of Psychology, vol. i. chaps. 5 and 6; H. Lotze, Metaphysic, bk. iii. chaps. 1 and 5 (Eng. trans., vol. ii. pp. 163-198, 283-517); H. Münsterberg, Grundzüge der Psychologie, i. chaps. 11. (pp. 402-436), 15 (pp. 525-562); G. F. Stout, Manual of Psychology,^3 Introduction, chap. 3; James Ward, Naturalism and Agnosticism, vol. ii. lects. 11 and 12 (art. “Psychology” in Supplement to Encyclopædia Britannica, p. 66 ff.).

-----

Footnote 174:

Compare the following striking passage from Avenarius, Menschliche Weltbegriff, p. 75: “Let an individual M denote a definite whole of ‘perceived things’ (trunk, arms and hands, legs and feet, speech, movements, etc.) and of ‘presented thoughts’ as I, ... then when M says ‘I have a brain,’ this means that a brain belongs as part to the whole of perceived things and presented thoughts denoted as I. And when M says ‘I have thoughts,’ this means that the thoughts themselves belong as a part to the whole of perceived things and presented thoughts denoted as I. But though thorough analysis of the denotation of I thus leads to the result that we have a brain and thought, it never leads to the result that the brain has the thoughts. The thought is, no doubt, a thought of ‘my Ego,’ but not a thought of ‘my brain’ any more than my brain is the brain of ‘my thought.’ I.e. the brain is no habitation, seat, generator, instrument or organ, no support or substratum of thought. Thought is no indweller or commander, no other half or side, and also no product, indeed not even a physiological function or so much as a state of the brain.”

Footnote 175:

As elsewhere in this work, I am using the terms “mind” and “soul” as virtually interchangeable names for the object studied by the psychologist. So far as there is any definite distinction of meaning between the terms as currently used by English writers, “soul” seems to carry with it more of the implication of substantiality and relative independence than “mind.” It might not be amiss to adopt the term “soul” as a name for the finite subject of experience as he is for himself in actual social life, and to confine the name “mind” to the construction which symbolises this subject for psychological purposes. But the popular antithesis between soul and body is perhaps too strongly rooted to admit of this suggestion. In earlier passages, e.g., Book II. chap. 2, § 6, I have used the term “spirit” in the sense here suggested for “soul.”

Footnote 176:

So, in dealing with astronomical problems, we are free to adopt either the Copernican or the Ptolemaic scheme, whichever happens to be the more convenient for our special purpose. The superior truth of the Copernican system seems to mean no more than that the range of its utility is the wider of the two. I may observe that I do not here employ the term “utility” in the narrowly practical sense of those philosophers who, e.g., condemn all speculation about the “Absolute” on the ground of inutility. Whatever satisfies any human aspiration is for me, so far, “useful.” It follows that there is, for me, no such thing as the “useless knowledge” which “Pragmatism” denounces. Thus, if a man’s peace of mind depends upon speculation about the “Absolute”—on the habits of angels, or any other topic you like (and this is a matter in which every man must in the end decide for himself)—Pragmatism would appear to be false to its own principle in forbidding him to speculate.

Footnote 177:

The assumption is not always made, however. Professor Münsterberg, who classes himself as a supporter of Parallelism, holds on metaphysical grounds that all causal connection must be between physical states. Hence he denies that psychical states can be causally connected with one another, except indirectly through the causal relations of their physical correlates. His doctrine is thus hardly to be distinguished from Epiphenomenalism, except in terminology, though he avoids the consequence of practical Fatalism by his insistence upon the purely artificial nature of both the physical and the psychical series. (His reason for refusing to admit causal relation between psychical states is that causal connection can only be established between universals, whereas every psychical state is unique. Does not this argument imply a confusion between the actual experience and its psychological symbol?)

Footnote 178:

Most supporters of Parallelism, it may be noted, stultify their own case, so far as it rests on this special contention, by admitting the causal determination of psychical states by one another, though, as psychical states are essentially qualitative, the reduction of causation to quantitative identity is particularly inadmissible here. Professor Münsterberg is quite consistent, therefore, in denying psychical causality and reducing Parallelism to Epiphenomenalism.

Footnote 179:

The reader who has followed the argument of our Third Book will not need to be reminded that the world of purely mechanical processes is simply an ideal construction based on postulates which we make for their practical convenience, and in no sense a direct transcript of the world of actual experience.

Footnote 180:

The “neutral Monism” to which the doctrine of rigid Parallelism logically leads, when put forward as more than a working hypothesis, will, one may hope, in England at least, fail to survive the exposure of its illogicalities in the second volume of Professor Ward’s Naturalism and Agnosticism.

Footnote 181:

This case includes, as will be apparent on a little reflection, not only the initiation of new motor reactions upon a sensation or percept, but also that of sensation itself as a qualitatively novel reaction upon physiological stimulation, and thus includes both the processes in which supporters of Interaction have always recognised the causal interconnection of the physical and the psychical.

Footnote 182:

It is with great pleasure that I note the coincidence of my own view on the impossibility of reconciling Parallelism with the recognition of the psychological importance of “meaning” with that of Mr. Gibson (essay on “The Problem of Freedom,” in Personal Idealism, p. 150 ff.). Professor Münsterberg’s declaration, that the consciousness investigated by Psychology “knows nothing by its knowledge and wills nothing by its will,” seems to me a confession of the bankruptcy of Parallelism as a basal psychological hypothesis. Still more so his elaborate and brilliant demonstration that the “brain” with which my “mind” may be regarded as “parallel” is not the brain as studied and charted by the anatomist, i.e. not the brain as a physical object at all. See Psychologie, i. 415-428.

Footnote 183:

See his essay on “Error” in Personal Idealism.

-----