It is undeniably the case, as Matravers insists, that normative political philosophy has been dominated for three decades or more by theories of justice. But these theories, including John Rawls's signal achievement, have concentrated on distributive justice and questions of retributive justice have been excluded. Why is this the case? Matravers has an answer: The separation occurred because “contemporary theories of justice cannot explain the relationship of justice and morality more broadly conceived” (p. 2). This seems a way to restate the problem rather than to offer an answer. But the problem itself is stated clearly: Absent an overall moral context within which to situate both distributive and retributive accounts, such accounts become less connected to, or intelligible to, one another. Going their own separate ways impoverishes each.

Matravers builds his case by exploring, first, what is at stake in the debate between consequentialism and retributivism—the dominant ways of construing the punishment task as an autonomous enterprise. But must one choose one or the other? Are these approaches mutually exclusively? Matravers argues that, no, they are not and that each captures insights that any comprehensive theory must take into consideration. Indeed, second, any account, to be compelling, must incorporate what he calls a “correct understanding” of what is due to responsible moral agents. Is there something like what Simone Weil would call a right to punishment? (Surprisingly, Matravers does not reference Weil, this despite the fact that punishment features centrally in her major treatise, The Need for Roots.)

Contemporary punishment theory is a failure because, no matter how the question is refracted, the need to integrate and to justify an account of punishment within a larger overall theory is ignored. Matravers' first five chapters are devoted to the critical task of showing the internal weaknesses and flaws in current accounts, whether consequentialist or retributivist; whether justice is construed as impartial or mutually self-interested. Chapters 6 through 9 offer his constructivist theory of moral norms, going on to locate coercion and punishment within the framework of what he calls a comprehensive account of the moral community.

This is a very tall order and Matravers hasn't quite met his goal. His meticulous unpacking of the accounts he finds wanting is carefully done. But his arguments suffer from a fair bit of repetition as he states and restates the central dilemma and his assessment of that dilemma. Moving to what he calls “constructivism” rather than “contractualism,” and it is less clear than it ought to be precisely what hangs on his rather idiosyncratic use of the term “constructivism,” Matravers builds up a moral context within which punishment theory can be located. The elements of such an account must be complex and multiple. Any such account must not only take seriously questions of reciprocity, fair play, and mutual advantage, but also incorporate matters of “sufficient security” and “justified coercion,” else punishment theory will wither on the vine.

Especially important here is Matravers' explicit recognition of “the educative function of punishment” whereby a community “expresses its abhorrence at the offender's action” and reaffirms the choice all persons must make between the requirements of membership and the yearning for some hypothetically independent existence. Despite the importance and complexity of its many parts, Justice and Punishment doesn't add up to the compelling, comprehensive moral theory within which Matravers claims, rightly, any account of punishment must be imbedded. His conclusions restate the problem, this time as a claimed accomplishment. But a few quite essential ingredients are missing—including a clearer articulation of precisely what Matravers means by “the moral community” in light of the fact that the conditions of contemporary pluralistic societies are such that we find ourselves members of a number of overlapping moral communities whose requirements for membership in good standing and standards of what counts as an infraction and what as acceptable punishment may conflict with, rather than reinforce, one another.

One minor note: In his determination to right the wrongs of male pronouns, “she” becomes the new universal in Matravers book. The upshot is that all murderers are “shes.” This is a bit unsettling.