This article asks whether justice is binary, whether matters are either-or with respect to it. This question has been inexplicably neglected, and the elementary conceptual work has not been done. We consider this question through exploring the implications of free-will-related justice. We see that there are actually two questions of very different scope here, and that two distinct notions of binarity need to be distinguished. In the process, the plausibility of considering justice as a binary notion is evaluated.

The question whether libertarian free will is 'worth wanting' was emphasized (and given a negative answer) by Dennett in his recent Elbow Room. I argue that this question is seriously ambiguous, and includes two different sorts of questions, which I attempt to elucidate. I then go on to argue that once the two questions are clearly perceived, the compatibilist case is hardly as strong. Even if libertarian free will is impossible (and is thus in one sense 'not worth wanting'), in another sense libertarian free will is 'worth wanting'.

I have argued that compatibilism has difficulties resisting prepunishment, and that it is thus a much more radical view than is typically presented and perceived. Helen Beebee presented two counterarguments, which I examine.

In "On the Alleged Shallowness of Compatibilism: A Critical Study of Saul Smilansky: 'Free Will and Illusion'" ('Iyyun' vol. 51, 2002; 63-79) James Lenman mounts a strong attack on the notion of 'ultimate injustice', a mainstay of common dissatisfaction with compatibilism. I survey two broad ways in which we can understand justice. Under the first, the notion of ultimate injustice is indeed questionable, but under the second it makes good sense. Then I show that in the free will issue it is the second type of justice, that allows and indeed affirms ultimate injustice, which is most salient.

In his recent monumental book On What Matters, Derek Parfit argues for a hard determinist view that rejects free will-based moral responsibility and desert. This rejection of desert is necessary for his main aim in the book, the overall reconciliation of normative ethics. In Appendix E of his book, however, Parfit claims that it is possible to mete out fair punishment. Parfit?s position on punishment here seems to be inconsistent with his hard determinism. I argue that Parfit is mistaken here, in a way that leads him to unjustified optimism about the possibility of fair penalization. Insofar as we take the free will problem seriously, we cannot reconcile a belief in the absence of desert with a belief in the fairness of penalization.

In 'Free Will, Agency, and Meaning in Life', Derk Pereboom proposes an optimistic model of life that follows on the rejection of both libertarian and compatibilist beliefs in free will, moral responsibility, and desert. I criticize his views, focusing on punishment. Pereboom responds to my earlier argument that hard determinism must seek to revise the practice of punishment in the direction of 'funishment', whereby the incarcerated are very generously compensated for the deprivations of incarceration. I claimed that funishment is a practical 'reductio' of hard determinism. Pereboom replies, but I claim that he misses a key component of my 'reductio', the idea that moving in the direction of funishment will considerably weaken the deterrence of potential criminals so that hard determinism becomes self-defeating in practice. Beyond the challenge of funishment, I raise various other difficulties with Pereboom's model, concerning its deeply unintuitive implications, the harm it does to the motivation of potential criminals, its weakness in resisting utilitarian-like dangers, and more. Our conclusions should lead to a re-evaluation of the compatibilist interpretation of moral life, as a richer, more plausible, and safer interpretation than hard determinism. This needs to be combined with a true hard determinist acknowledgment of the deep injustice and tragedy involved in punishment in light of the absence of libertarian free will. Such a complex view will come closer to doing justice to notions of justice, morality, and decency.

I have argued that compatibilism cannot resist in a principled way the temptation to prepunish people, and that it thus emerges as a much more radical view than is typically presented and perceived; and is at odds with fundamental moral intuitions. Stephen Kearns has replied, and in the present paper I examine his arguments.

In his 1995 'Mind' paper, Fred Feldman claims that our central philosophical way of thinking about the connections between responsibility and desert and between time and desert is misguided. I believe that we have the resources to handle Feldman's doubts, and maintain the conceptual connection between responsibility and desert. My detailed interpretation of the connection between responsibility and desert satisfies our intuitions concerning Feldman's examples, and accords with the general structure of our thought concerning justice, insofar as it involves desert. I also briefly consider the issue of time and desert.

TBA

In 'Mind' of January 1996 I proposed an outline of a theory of responsibility and desert, which I claimed both a) enables us to see responsibility as a condition for desert even in the major apparent counter-examples such as those proposed by Fred Feldman; and b) represents the ordinary way of seeing the connection between responsibility and desert better than previous formulations. Behind this proposal lies a crucial distinction between 'two ways' in which responsibility can be a condition for desert. From Feldman's reply in the same issue it seems that this crucial distinction was not sufficiently brought out in my paper. This I attempt to do here.

Hard determinism is commonly thought to have grave moral consequences. Without attempting to fully deny this, I explore a previously unnoticed apparent ethical advantage of a belief in hard determinism. The lack of concern with personal desert which accompanies belief in hard determinism seems to make possible a form of purer morality' than ordinary morality. In the process I briefly consider some perplexities of Kantianism, the personal' difficulty of being a hard determinist, and the question as to whether non- utilitarian ethical considerations are possible with hard determinist premises.

The paper points out a crucial difficulty in Van Inwagen's An Essay on Free Will--his claim for the direct obviousness of the existence of libertarian moral responsibility. Van Inwagen's attempt to discredit the beliefs of the people who doubt the existence of libertarian moral responsibility is shown not to succeed. And since virtually all of Van Inwagen's case for the existence of libertarian free will depends on his case for the "obviousness" of libertarian moral responsibility, the failure of the latter undermines the former.