If we confine our attention to such connexion between means and ends as is scientifically cognisable, it does not appear that an act now deliberated on can be less or more a means to any ulterior end, because it is predetermined. It may, however, be urged that in considering how we ought to act in any case, we have to take into account the probable future actions of others, and also of ourselves; and that with regard to these it is necessary to decide the question of Free Will, in order that we may know whether the future is capable of being predicted from the past. But here, again, it seems to me that no definite practical consequences would logically follow from this decision. For however far we may go in admitting Free Will as a cause, the actual operation of which may falsify the most scientific forecasts of human action, still since it is _ex hypothesi_ an absolutely unknown cause, our recognition of it cannot lead us to modify any such forecasts: at most, it can only affect our reliance on them.
We may illustrate this by an imaginary extreme case. Suppose we were somehow convinced that all the planets were endowed with Free Will, and that they only maintained their periodic motions by the continual exercise of free choice, in resistance to strong centrifugal or centripetal inclinations. Our general confidence in the future of the solar system might reasonably be impaired, though it is not easy to say how much;[67] but the details of our astronomical calculations would be clearly unaffected: the free wills could in no way be taken as an element in the reckoning. And the case would be similar, I suppose, in the forecast of human conduct, if psychology and sociology should ever become exact sciences. At present, however, they are so far from being such that this additional element of uncertainty can hardly have even any emotional effect.
To sum up: we may say that, in so far as we reason to any definite conclusions as to what the future actions of ourselves or others will be, we must consider them as determined by unvarying laws: if they are not completely so determined our reasoning is _pro tanto_ liable to error: but no other is open to us. While on the other hand, when we are endeavouring to ascertain (on any principles) what choice it is reasonable to make between two alternatives of present conduct, Determinist conceptions are as irrelevant as they are in the former case inevitable. And from neither point of view does it seem practically important, for the general regulation of conduct, to decide the metaphysical question at issue in the Free-will Controversy: unless--passing from Ethics into Theology--we rest the reconciliation of Duty and Interest on a theological argument that requires the assumption of Free Will.