Elements of metaphysics

by A. E. Taylor

Cause, as currently understood, is thus identical not with the whole true logical ground, but with the ground so far as it can be discovered in the train of temporally antecedent circumstances, i.e. cause is incomplete ground. This point is important, as it shows that the principle of Causation is not, like the principle of Sufficient Reason, axiomatic. It is no necessary logical consequence of the knowability or systematic character of the Real that an event should be completely determined by temporally antecedent events; for anything that is implied in the systematic character of the Real, the event may be equally dependent on subsequent occurrences. Again, the principle of Causation cannot be empirically established by an appeal to the actual course of experience. Actual experience is certainly not sufficient to show that every event is absolutely determined by its antecedent conditions; at most the success of our scientific hypotheses based upon the assumption of causality only avails to show that events may be inferred from their antecedents with sufficient accuracy to make the causal assumption practically useful.

Regarded as a universal principle of scientific procedure, the causal assumption must be pronounced to be neither an axiom nor an empirical truth but a postulate, in the strict sense of the word, i.e. an assumption which cannot be logically justified, but is made because of its practical value, and depends upon the success with which it can be applied for confirmation. In the sense that it is a postulate which experience may confirm but cannot prove, it may properly be said to be a priori, but it is manifestly not a priori in the more familiar Kantian sense of the word. That is, it is not a necessary and indispensable axiom without which systematic knowledge would be impossible. For, as we have already seen and shall see more fully in the immediate sequel, it may not be, and indeed in the last resort cannot be, true.

§ 5. This last statement will possibly appear startling to the reader who is unacquainted with the history of metaphysical investigations into Causality. But it is easy to show that it is really the expression of an obvious truth. For the causal principle, as we have just seen, is an imperfect expression of the really axiomatic principle of Sufficient Reason or Ground and Consequence. And it is readily seen that the expression it gives to that principle, because imperfect, must be partially false. What the principle of Ground and Consequence says is, that the whole of existence is a single coherent system in which every part is determined by the nature of the whole as revealed in the complete system. But if this is true, each constituent of the system can only be completely determined by its connections with all the rest. No constituent can be entirely determined by its relations to a lesser part of the whole system, in the way presupposed by the notion of one-sided causal dependence. The “cause” must, if the principle of Ground and Consequence be valid, be determined by the “effect” no less than the “effect” by the “cause.” And therefore the causal postulate cannot be the whole truth.

How this fatal logical defect in the principle of Causation makes itself felt in the logic of the inductive sciences, and how logicians have sought without success to avoid it, we shall incidentally see as our discussion proceeds. At present we must be content to note that, owing to this flaw, Causation, wherever it is asserted, can only be Appearance and never complete Reality, and that no science which works with the concepts of cause and effect can give us the highest truth. Of course, the logical defects of the concept need not impair its practical usefulness. Though it can never, for the reason given already, be ultimately true that any event is absolutely determined by antecedent events, the assumption may be sufficiently near the truth to yield useful deductions as to the course of occurrences, precisely as a mathematical approximation to the value of a surd quantity may, without being the exact truth, be close enough for practical use. Also, it might well be the case that the causal postulate approximates more nearly to the truth in some spheres of investigation than in others, a consideration which is not without its bearing on the ethical problems of freedom and responsibility.

If we ask how the causal postulate, being as it must be only imperfectly true, comes to be made, the answer is obvious. The whole conception is anthropomorphic in origin, and owes its existence to our practical needs. To take the latter point first, logically there is no better reason for treating an event as determined solely by antecedents, than for treating it as solely determined by subsequent events. Yet when the latter supposition is made, as it is by all believers in omens and presages, we all agree to condemn it as superstitious. Why is this? Two reasons may be assigned. (a) Even granting that an event may be determined by subsequent events, yet, as we do not know what these events are until after their occurrence, we should have no means of inferring by what particular events yet to come any present event was conditioned, and thus should be thrown back upon mere unprincipled guess-work if we attempted to assign its, as yet future, conditions.

(b) A more important consideration is that our search for causes is ultimately derived from the search for means to the practical realisation of results in which we are interested. We desire to know the conditions of occurrences primarily, in order to produce those occurrences for ourselves by setting up their conditions. It is therefore essential to us for our practical purposes to seek the conditions of an occurrence exclusively among its antecedents, and the causal postulate which asserts that the complete conditions of the event are comprised somewhere in the series of antecedent events is thus the intellectual expression of the demand made by our practical needs upon Reality. We postulate it because, unless the postulate is approximately realised, we cannot intervene with success in the course of events. We refuse, except as a pure speculation, to entertain the notion that an event may be determined by subsequent as well as by antecedent events, because that notion leads to no practical rules for operation upon our environment.

§ 6. As might be expected of a postulate so obviously originated by our practical needs, the concept of cause on examination reveals its anthropomorphic character. This is particularly obvious when we consider the concept of Causation as it figures in everyday unscientific thought. The various scientific substitutes for the popular notion of cause all exhibit traces of the endeavour to purge the conception of its more anthropomorphic elements. In the popular use of the concept this anthropomorphism comes out most strikingly in two ways. (a) A cause, as popularly conceived, is always a person or thing, i.e. something we can imagine as a whole, and into which we can mentally project a conscious life akin to our own. To the scientific mind it seems obvious that causes and effects are alike events and events only, but for popular thought, while the effect is always a quality or state (e.g., death, fever, etc.), the cause is regularly a thing or person (the bullet, the poison, the tropical sun, etc.).

(b) Closely connected with this is the emphasis popular thought lays upon what it calls the activity of the cause. The cause is never thought of as merely preceding the effect as an “inseparable antecedent”; it is supposed to make the effect occur, to bring it about by an exercise of activity. According to the most coherent expositions of this type of thought, in causation one thing is always active in producing a change in another thing which is passive. The origin of this notion is sufficiently obvious. As all philosophers since Hume have recognised, the “activity” of the cause results from the ascription to it of the characteristic feeling of self-assertion and self-expansion which accompanies our own voluntary interference in the course of events. Similarly, the “passivity” of the thing in which the effect is produced is only another name for the feeling of coercion and thwarted self-assertion which arises in us when the course of nature or the behaviour of our fellows represses our voluntary execution of our designs.

Science, in its attempt to extend the concept of causal determination over the whole domain of existence, has naturally felt these anthropomorphic implications as obstacles. From the effort to expel them arises what we may call the common scientific view of causation, as ordinarily adopted for the purposes of experimental investigation and formulated in the works of inductive logicians. The concept of a thing, except as the mode of interconnection of states, being unnecessary for the sciences which aim simply at the reduction of the sequence of occurrences to order, the notion of causation as a transaction between two things is replaced in the experimental sciences by the conception of it as merely the determination of an event by antecedent events. Similarly, with the disappearance of things as the vehicles of causal processes falls the whole distinction between an active and a passive factor. As it becomes more and more apparent that the antecedent events which condition an occurrence are a complex plurality and include states of what is popularly called the thing acted upon as well as processes in the so-called agent, science substitutes for the distinction between agent and patient the concept of a system of reciprocally dependent interacting factors. These two substitutions give us the current scientific conception of a cause as the “totality of the conditions” in the presence of which an event occurs, and in the absence of any member of which it does not occur. More briefly, causation in the current scientific sense means sequence under definitely known conditions.

Indispensable as this notion of the determination of every event by a definite collection of antecedents and by nothing else is for practice, regarded as a logical formulation of the principle of the systematic unity of existence, it is open to grave objections, most of which will be found to have made themselves felt in the logic of the inductive sciences quite independently of conscious metaphysical analysis. In dealing with these difficulties, we shall find that their general effect is to place us in the following dilemma. If we wish to state the causal principle in such a way as to avoid manifest speculative falsehood, we find that it has to be modified until it becomes identical with the principle of Ground and Consequence in its most universal form, but as thus modified it is no longer of any service for the purposes of the experimental sciences. You seem driven to take it either in a form in which it is true but practically useless, or in one in which it is useful but not true. To illustrate the way in which this dilemma arises, we may examine three of the main problems which have actually been created by the scientific use of the principle,—(a) the puzzle of continuity; (b) the puzzle of the indefinite regress, (c) the puzzle of the plurality of causes.