Nowell-Smith cites R. E. Hobart, who had refined Hume's compatibilism in his landmark 1934 essay Free Will As Involving Determination And Inconceivable Without It.

Hobart correctly identified the need for some determination in the act of willing, without which there could not be moral responsibility. But Nowell-Smith demanded strict determinism.

Unlike Hobart, who admitted the possibility of some indeterminism and even absolute chance in the courses of action which "present themselves" to our will as alternative possibilities, Nowell-Smith makes the classic error of confusing indeterminism with total chance. He thinks that any real chance would render the world a complete chaos. He does not see that only adequate determinism is required.

The fallacy of [Incompatibilism] has often been exposed and the clearest proof that it is mistaken or at least muddled lies in showing that I could not be free to choose what I do unless determinism is correct. For the simplest actions could not be performed in an indeterministic universe. If I decide, say, to eat a piece of fish, I cannot do so if the fish is liable to turn into a stone or to disintegrate in mid-air or to behave in any other utterly unpredictable manner. (p.47)

Nowell-Smith's desire for a determined choice is satisfied by a two-stage model for free will, where the choice is adequately determined, even if the generation of alternative possibilities has involved indeterminism.

Six years later, Nowell-Smith again visited the problem (or as the determinists called it, the pseudo-problem) of free will in his essay in Mind, "Libertarians and Determinists." In it he says puzzlingly, "It is an old criticism of libertarianism that it makes out voluntary and deliberate actions to be 'chance events ', and libertarians are as anxious to deny this 'indeterminist' theory as they are to deny determinism." (p.329)

One of the puzzling features of the controversy is the fact that it is, on the whole, the libertarians who insist that there is such a thing as the Problem of free will, while the so-called determinists tend to call this problem a pseudo-problem. (They admit, of course, that there are plenty of problems in this obscure area.) This is puzzling because, in a way, there can be no such thing as the Problem for the libertarian. He thinks that it is obvious, that we have some sort of immediate insight into the fact that men are sometimes responsible for what they do and that they would not be responsible if they were not free in some 'contra-causal' sense; and if this is so, it is obvious that men are free in this sense and no problem arises. Nor does he even require to use this argument; for, he says, we have an immediate and indefeasible awareness of freedom. Yet the very fact that he insists on there being a problem here- would seem to show that there is something unsatisfactory about his immediate awareness. (p.318)

The immediate awareness of freedom. Could it be by introspection that we know of the existence, at the moment of choice, of open possibilities? Do we, as it were, see these possibilities? When we talk of 'possibilities' in connexion with natural phenomena, we always say something that could be expressed by the use of the modal word 'may'. We say "it may rain or it may not" ; to add "both possibilities exist" is to add a rhetorical flourish. In the case of human choice we say "I can (could) do this; but on the other hand I can (could) do that". In many cases it would be absurd, at a common sense level, to doubt that this is so or that the speaker knows it to be so; but philosophers are rightly puzzled by such platitudes and rightly ask "Just what is it that you know?" The philosopher's doubts are not allayed by the answer "What you know is that two genuinely open possibilities exist". Nor, if he is puzzled by the question "How do I know that I can do either this or that?", is he satisfied by the answer "You know this because you 'see' that both possibilities exist". Moreover, even it it were true that he knows that he could choose either this or that by some kind of introspection, it could not be true that part of what he so knows is "and no one could predict which I shall choose". That he could only know by testing people's powers of prediction. (And, of course, he knows both that he can choose and to what extent people can predict by experience; not by inspecting his experience at the moment of choice, but in the sort of way that he knows his own name or the way home.)

Exactly the same argument applies to the interpretation of the phrase 'I could have acted otherwise'. This is something which the plain man would certainly say of many of the things he does. And, contrary to the libertarian thesis, be would say this of actions that involve neither obligation nor an effort of will. No doubt there are many important differences between making an effort to do one's duty and choosing a peach; but at the common sense level none of the phrases that we use to indicate a free choice ('open possibilities', 'alternative courses', 'could have chosen otherwise') apply to the one rather than to the other. In the last resort it is not the libertarian's belief in freedom in moral contexts, but his denial of freedom in other contexts that is paradoxical. (p.324)

In the same way, when we ask whether someone can do or could have done something, we are asking whether he will or would have been successful if he tries or had tried to do it; and his trying to do it presupposes his wanting to do it. It is absurd to speak of success or failure or to ask whether any- thing prevents a man from doing something without presupposing a desire to do it on his part. And it is this presupposition that we are expressly excluding when we ask "Could he still do it, even if he did not want to?" This question makes sense if it is construed as being analogous to "Is it a conductor although it is not now conducting?" For example, I do not now want to multiply 567 by 479; I have no earthly reason for making the attempt; yet I have no doubt that I can do this, that is to say that if I had a reason for trying, I should try and succeed. But to interpret it to mean "Can I do it, even granted that I have no reason for trying? is to construe it as being analogous to " Would it still be a conductor if the conditions under which it actually carries a charge are not fulfilled?" (p.327)

Prima facie, then, statements containing ' can ' and ' could' are not reports of what occurs or does not occur at a particular moment; their modality in itself suggests this,' and the onus of proof is on those who would construe them differently in the special case of being able to choose the path of duty in the face of temptation. And even if it were true that, in this special case, the modal sentence was used to report an intro- spected datum, it would be difficult to see how 'no one could have predicted which I would choose' or 'I created a rupture in causal continuity' could be either a part of or entailed by what was reported. Moreover, if we do construe the modal sentence in this way, we ought, I think, to be more puzzled than we are by the fact that questions of responsibility, of the propriety or impropriety of blaming someone, turn on the truth or falsity of 'he could have acted otherwise'.

I should like to emphasize again the fact that I am not grinding any kind of behaviourist axe here. To say that ' I could have acted otherwise' is not a report of what one 'sees ' when he looks within is not to say that it is in all cases true that a spectator knows better than a man himself what he could or could not have done if he had tried. (p.328)

The Concept of Forces. So far I seem to have been defending a form of determinism; but this is not the case. I have simply refrained from raising the familiar objections to determinism, which are just as fatal. For the fact is that ' determined ' and 'contra-causal? have not been analysed and until their meanings are made clear, until we know to what we are committed when we say that choice is or is not contra-causal, the disputants must be fighting in the dark. There are at least three different bogeys that go under the name of 'determinism' and they must be kept distinct because each requires a different technique of exorcism.

(i) There is logical determinism or fatalism, according to which we cannot help doing what we do since things are what they are and our actions will be what they will be. I shall forbear to comment on this brand of determinism, since I doubt whether anyone is now seriously deceived by it.

(ii) There is the theory that we might call 'literal mechanical determinism'. Every human action, no matter what else it may be, is a complex but determinate movement of a determinate body. Suppose that this body were, quite literally, a complex machine? Suppose that we could predict its every movement in accordance with the laws of classical mechanics if only we knew enough about the particles concerned? This is the eighteenth-century bogey, and about it one can say little except that it might be true, that there is nothing in the present trend of natural science to warrant it (indeed it seems to presuppose a misunderstanding of explanation and prediction even within natural science) and that, if it were true, our talk about human conduct would be quite different from what it actually is.

(iii) There is also the obscure, elusive and much more important theory that desires (and other motives) are somehow like mechanical forces. The mechanical analogy is recognized as an analogy, but accepted without criticism as a good one. The idea seems to be that a man (or a 'self') is like a billiard- ball on which certain forces (his desires) act. The determinists say that all his actions are determined by the strength and direction of the forces, as movements are in classical mechanics, so that he 'must' do what he does in the same sense that a billiard ball 'must' move as it does.

Here Nowell-Smith describes Robert Kane's "Self-Forming Actions" exactly

The libertarian agrees that this is true in the case of most human actions but wants to make a special exception in the case of those situations that involve a conflict between desires and the sense of duty. The interplay of the forces, he says, is liable to be upset by the operations of a special sort of force, to which there is no analogue in mechanics, called "will-energy". Of this force little can be said except that we all feel it within us (and that is why little needs to be said), that we can use it or fail to use it, that it is unpredictable in its operations and that our status as free, responsible beings depends on its existence. Just why an observer is supposed to be able to discover the exact nature of the other forces but not of "will-energy" is not made clear; but this must be the case, since it is held that all actions other than those involving the sense of duty are, in principle, predictable in a para-mechanical way.