It is chiefly on the Libertarian side that I find a tendency to the exaggeration of which I have just spoken. Some Libertarian writers maintain that the conception of the Freedom of the Will, alien as it may be to positive science, is yet quite indispensable to Ethics and Jurisprudence; since in judging that I “ought” to do anything I imply that I “can” do it, and similarly in praising or blaming the actions of others I imply that they “could” have acted otherwise. If a man’s actions are mere links in a chain of causation which, as we trace it back, ultimately carries us to events anterior to his personal existence, he cannot, it is said, really have either merit or demerit; and if he has not merit or demerit, it is repugnant to the common moral sense of mankind to reward or punish--even to praise or blame--him. In considering this argument, it will be convenient--for clearness of discussion--to assume in the first instance that there is no doubt or conflict in our view of what it is right to do, except such as may be caused by the present question. It will also be convenient to separate the discussion of the importance of Free Will in relation to moral action generally from the special question of its importance in relation to punishing and rewarding; since, in the latter species of action, what chiefly claims attention is not the present Freedom of the agent, but the past Freedom of the person now acted on.

As regards action generally, the Determinist allows that a man is only morally bound to do what is “in his power”; but he explains “in his power” to mean that the result in question will be produced if the man choose to produce it. And this is, I think, the sense in which the proposition “what I ought to do I can do” is commonly accepted: it means “can do if I choose,” not “can choose to do.” Still the question remains “_Can_ I choose to do what in ordinary thought I judge to be right to do?” Here my own view is that--within the limits above explained--I inevitably conceive that I _can_ choose; however, I can suppose myself to regard this conception as illusory, and to judge, inferring the future from the past, that I certainly shall not choose, and accordingly that such choice is not really possible to me. This being supposed, it seems to me undeniable that this judgment will exclude or weaken the operation of the moral motive in the case of the act contemplated: I either shall not judge it reasonable to choose to do what I should otherwise so judge, or if I do pass the judgment, I shall also judge the conception of duty applied in it to be illusory, no less than the conception of Freedom. So far I concede the Libertarian contention as to the demoralising effect of Determinism, if held with a real force of conviction. But I think the cases are rare in which it is even on Determinist principles legitimate to conclude it to be certain--and not merely highly probable--that I shall deliberately choose to do what I judge to be unwise.[62] Ordinarily the legitimate inference from a man’s past experience, and from his general knowledge of human nature, would not go beyond a very strong probability that he would choose to do wrong: and a mere probability--however strong--that I shall not will to do right cannot be regarded by me in deliberation as a reason for not willing:[63] while it certainly supplies a rational ground for willing strongly--just as a strong probability of any other evil supplies a rational ground for special exertions to avoid it. Indeed, I do not see why a Libertarian should not--equally with a Determinist--accept as valid, and find it instructive to contemplate, the considerations that render it probable that he will _not_ choose to do right in any particular circumstances. In all ordinary cases, therefore, it does not seem to me relevant to ethical deliberation to determine the metaphysical validity of my consciousness of freedom to choose whatever I may conclude to be reasonable, unless the affirmation or negation of the Freedom of the Will somehow modifies my view of what it would be reasonable to choose to do if I could so choose.

The Methods of Ethics — Part 75