The refutation of indeterminism is rendered difficult, or, at least, deficient in persuasiveness, by the fact that, like all rationalist theories, it necessarily embodies a certain amount of empirical fact. Thus the indeterminist will meet accusations of ignoring facts by saying, “But I have also admitted that that is so”; he will adopt such devices as the “in so far as”, so that any demonstration of causal necessity will hold of things only in so far as they are necessitated, and will leave it possible for them to proceed quite otherwise in so far as they are free. While, however, it would be too laborious to pursue the indeterminist through all the shifts by which he tries to defend his position, the consideration of some of them may help to bring out the main issues. When, for example, in supporting man's freedom, he “also” admits necessitation in nature, he prompts the determinist to bring up the important point that there is no distinction whatever between man and nature (and hence no question of a false or forced analogy between the two), that “nature” means no more and no less than what is, and that a theory of the conditions of existence, embodying a general theory of causality, will apply indifferently to men and any other existing things.

It is to be noted, at least, that, in default of the presentation of such a general theory, any discussion of supposed peculiarities of human causality is quite beside the mark. Moreover, this general theory will not be affected by the appeal to special cases in the attempt to find a “negative instance”. The determinist may quite well admit that he does not know the determining cause of a particular mental event; but this will no more lead him to doubt the truth of determinism than a similar ignorance in the case of a non-mental event would do. In fact, if causal necessitation were an “unproved and unprovable assumption” in respect of mental events, it would be equally so in respect of other events. It is not surprising that indeterminists should avoid this extreme position, the unscientific and mythological character of which is only too apparent. But without giving some ground for belief in causal necessitation in “nature”, they cannot show that this ground is lacking in the case of man; and if they do give such a ground, it will be seen to cover the case of any “humanity” with which we are acquainted — apart from the point, previously noted, that humanity is in any case included in the subject-matter of logic (“what is”) and comes under the logical theory of causality.

Something may be said here about the kind of “proof” that is possible

  ― 123 ―
on a point of logic. As regards direct argument, one may attempt to show, in the manner of Alexander (largely following Kant), that a thing as spatio-temporal exhibits a certain character, e.g., that it occupies a definite place in a regular sequence of a certain type. To speak of a thing, it may be said, is to speak of certain “ways of working”, the continuance and the development of which are, of course, affected by the other ways of working by which the thing is surrounded. It would be argued, in this way, that it is a condition of a thing's existence that it determines and is determined by other things, and that to investigate or “give an account of” it involves consideration of such determinations. Thus, to give an account of any actual thing that could be called “initiative” would be to exhibit certain regularities, to present it, in the common phrase, as “subject to laws”, including those which “govern” its relations to other sorts of things which, in any particular instance, may or may not be present. In other words, discounting metaphysical notions of “governing” and restricting ourselves to a positive account of interrelated ways of working, we should treat the occurrence of “initiative” in the human mind in exactly the same logical way as we should treat the occurrence of magnetism in a pin.

The above remarks suggest a less direct treatment of logical problems, viz., by considering what is involved in the recognition of a thing as a subject of investigation — more generally, in the very possibility of “discourse”. This is, of course, the traditional approach (the “Socratic” approach) to logic; it leads up to and does not abrogate the consideration of what is involved in the recognition of the thing as existing; it is only in terms of existence that we can, in the end, criticise discourse. But this way of expressing the matter brings out the point that, in rejecting a particular logical theory, we should be able to show that the exponent of it not merely has a false view of existence but implicitly, in his own statement of the case, admits the view that we are upholding against him (as when a person argues against objective implication or denies objective truth). A particularly important instance is that of the demonstration, by the upholder of a spatio-temporal logic, of the fact that those who argue that certain things (e.g., minds) are not spatio-temporal, cannot avoid implying that they are. Again, indirect “proof” of a logical position may take the form of showing that our opponent's view involves him in insoluble problems — though this amounts to the same as contradicting the possibility of discourse. At any rate, while bringing out the insoluble problems of indeterminism is not the most rigorous proof of determinism, it may well be the most effective and would seem to be an essential preliminary.

The question of complete indeterminism need scarcely be argued here; it will perhaps be admitted that such a position cannot consistently give any description of anything — not even that of being “undetermined”. The position confronting us is that which upholds determination and also indetermination. And this position can be met by the regular arguments against any attempt to divide reality into “realms” (in effect,

  ― 124 ―
to have more than one logic or theory of being) — and particularly, as I have indicated in a number of papers, by the demonstration of the impossibility of finding any relation between the different realms; so that there will be a “universe” of complete determination and another “universe” of complete indetermination. If it were true, as Dr Loughnan says,note that “the doctrine of free will has nothing to say about the realm of physical science, neither explicitly nor by logical implication”, this would mean that free will had no physical effects, that it “subsisted” in entire separation from the physical — and so, apart from other difficulties, we should be back, on the side of the free, at complete indetermination: in fact, nonentity.

But if it is contended that the free acts on the determined (if the usual “interactionist” position is taken up), then the upshot is that there is no determination, no “law”, anywhere. For any physical “uniformity” is to the effect that a certain set of physical antecedents gives place to a certain set of physical consequents; but, with the intervention of a free agent, the very same set of physical antecedents will have a different set of physical consequents. It is surely clear that if, with no physical difference in the antecedents, there is a different physical sequence, then there can be no physical uniformity. This is simply an illustration of the impossibility of combining the free and the determined in any situation. For the determinist, of course, there is no difficulty. For, while in any case he holds the mental to be physical, the recognition of the occurrence of a certain sequence except when some other factor intervenes is a commonplace of the theory of physical interactions. To deny “interactionism” (interaction between different realms or levels of reality) is not to deny interaction. To state, e.g., that two substances combine in a certain way except when a third substance is present is still to state a uniformity, and leaves it possible to determine the “uniform” action of the third substance. But, as has been shown, the operation of a non-uniform factor would destroy all uniformities.

The argument is not affected by the adoption of a theory of a partly determined and partly free mind or, again, a theory of what we may call “inclining” but not necessitating causes. In the former case, the same problem as before arises in connection with the relation of the mind's free activities to other activities, mental or non-mental, and the additional difficulty of the same thing's having free and unfree parts is scarcely worth considering. In the latter case, the difficulty is the same; for, if we have a condition which is necessary but not sufficient for a certain result and this result nevertheless occurs, we simply have an arbitrary and occult factor contributing the remaining part of what is necessary — and, as before, this really means giving up uniformity. Either, in fact, we have necessary and sufficient conditions all the way, or we have something “causa sui”, which, mysteriously, is also the cause of something else — and it is no more possible to connect the self-causing

  ― 125 ―
with the other-causing of such an entity than to connect the determined and the undetermined in general.

In conclusion, it may be said that the selection of mind as a bearer of freedom is not due to any special interest in mind. Those who are interested in mind's workings will naturally take up a determinist position. The indeterminists are those with an axe to grind, with certain “values” to defend, with the view that certain things ought to be or are to be done. Theoretical concern with what is the case is, it seems to me, coextensive with determinism.