The subtitle of this book is suggestive, for one can avoid debating moral realism in several ways. Hume and Kant might first be understood as having avoided debating moral realism because the term was not invented for close to two centuries after they wrote. But of course what we call the ‘moral realism’ debate does apply to their theories despite the novel terminology. Or secondly one might invoke a claim that moral realism is a meta-ethical issue on a distinct plane from normative ethical considerations, and hence the subject of an entirely different kind of analysis that abstracts from, or is even unrelated to, particular normative theories. But this claim, although promoted by the initial twentieth-century advocates of what came to be called meta-ethics, is certainly implausible because some issues, such as the status of reason as a source of moral law or the role of values that depend on human nature, are not exclusively metaethical or normative, particularly in a Kantian theory. Or thirdly one might avoid the debate about realism by employing a narrow focus in reconstructions of political philosophy proper, discussing the foundational questions that involve realism somewhere other than in that specific discussion of political philosophy. In that case only the narrow normative claims and not the complete theory are presented without debating moral realism. Those who are concerned with political philosophy might in this case simply take for granted an answer to the realism question, or bracket the debate for the purpose of answering specific questions in political philosophy first. The issue of realism is then handled in the more extensive general theory, or perhaps in some foundational discussion. I take this third option to describe the way that Hume and Kant might be understood to present political philosophies that do not get entangled with issues about realism.

Kenneth Westphal has none of these in mind but instead holds that what he takes to be Hume and Kant’s shared framework – ‘Natural Law Constructivism’ – is completely worked out in a way that resolves all the issues about identification, knowledge and justification of moral principles without any dependence on whether value or moral facts or standards are real or not, and so is compatible with both moral realism and non-realism. Traditional natural law theory relied on a moral realism and moral teleology that both Hume and Kant reject with their constructivist version. Yet they retain the content of natural law. In the course of arguing for this conclusion Westphal presents many novel and sound assessments, but the whole does not serve to prove his main claim.

The first chapter sets up the main issue as showing how Hume and Kant share a particular approach to moral philosophy: ‘Their form of moral constructivism identifies and justifies strictly objective basic moral principles and the practices they structure (social institutions), whilst maintaining neutrality about moral realism and its alternatives’ (p. 1). ‘Moral realism’ is defined, without further explanation or justification, as ‘views according to which there are human mind-independent moral truth-makers, of whatever kind’ (p. 1). Such a definition should make Kant a moral realist by default, since Kant insists that the moral truth-maker is the reason shared by every possible rational being, not merely human beings. One should instead focus on the moral agent in general, or ‘finite rational being’, in which case the issue of realism becomes the more nuanced one of analysing the nature of moral agency and the extent to which it can be seen as real or to ground the real. While Westphal does assess the conditions for moral agency when looking at Kant’s moral philosophy in general, he also seems to insist more strongly on the particular human situation when focusing on ‘natural law’ as the set of principles of justice necessary for individual human action under conditions of a plurality of persons with moderately limited resources. These conditions are specifically human, even though given the conditions and human nature, the resulting principles of justice are not arbitrary although constructed for that situation and thus artificial (p. 5). Presumably this leaves space for a moral realist claim that a moral fact can be true for just such contingent circumstances, and equally space for a moral anti-realist claim that allows one to accept for purposes of argument the contingent circumstances and assumed ends of agents that allows one to deduce the principles. But both might object that Natural Law Constructivism would not be fully justified without resolving their further questions. So can Westphal show that Natural Law Constructivism does provide justification?

Seemingly not for Hume. Two chapters discuss how Hume’s theory of justice exhibits this form of constructivism. While moral virtues are natural, justice is an artificial virtue; some take this to mean that Hume is a moral realist regarding moral virtues. Westphal rightly takes Hume’s reliance on moral sentiments to be insufficient for moral judgement, so rejects Hume’s main approach to moral philosophy (pp. 43–5). Westphal instead finds a plausible theory in the way that Hume in the Treatise of Human Nature identifies a need for social coordination, given certain circumstances such as limited scarcity, and provides three principles of justice related to the recognition and stability of possessions, their transferability and the keeping of promises. These are artificial, yet not arbitrary, principles necessary for the end of a functioning society; hence Natural Law Constructivism. One might think that by insisting on the artificiality and contextual basis for these principles, Hume falls quite clearly on the side of an anti-realist theory of justice. Leaving that aside, there is a more important objection to seeing Hume as successfully providing a complete theory: he may use this process to identify principles of justice, but cannot justify them without justifying the ends for which social coordination is useful. The justificatory element can be provided only by some view on the value of human sentiments, needs, aims and the like, and that discussion involves elements of the moral realism debate. In places later in the book Westphal seems to concede that Hume’s theory is more of an incomplete (p. 168) or fragmentary (p. 195) step toward the proper Natural Law Constructivism provided by Kant.

The discussion of Kant is more nuanced and very rewarding. Largely adopting and extending Onora O’Neill’s interpretation, Westphal stresses that the universalization test itself expresses the nature of rational willing. He defends the universalization test against critics in part by adding an element: he proposes that both a maxim and its opposite be tested. If one is unable to consistently will one’s own maxim and its universalized counterpart, and one is also unable to consistently will the opposite of one’s own maxim and its corresponding universalized counterpart, then the action is permissible. Further, if a maxim passes the test but its opposite fails, the maxim is required; if the maxim fails and its opposite passes, the maxim is forbidden. This addition is intended to allow for the permissibility of maxims such as ‘whenever I reach a door at the same time as another, I shall always pass through second’ (p. 83). The idea is fruitful, but one must still fiddle with the maxim to determine a proper opposite. Substituting ‘never’ for ‘always’ to obtain the opposite would have both fail, but using ‘not always’ for the opposite would seem to allow it to pass the test, making the original maxim forbidden. Some more guidance on determining opposites would have been helpful.

Westphal concedes here that Kant does offer a ‘general’ metaphysics of morals regarding rational willing in the Critical works and then provides the ‘specific’ metaphysics concerning how that is realized in human beings in the body of the Metaphysics of Morals (pp. 79–80). Given Westphal’s definition of moral realism in relation to ‘human mind-independent moral truth makers’ (p. 1), one would think that the ‘general’ metaphysics would be presenting a moral realist view, and even the ‘specific’, because dependent upon the general, would have to be morally realist (ignoring here the caveats I already gave about needing to assess the status of reason itself). But Westphal’s actual analysis takes the nature of rational willing itself to be of prime importance, with the formal process of willing requiring the possibility of universalization: ‘Kant’s method for identifying and justifying basic moral principles is constructivist because it makes no appeal to any antecedent source or kind of normative authority. … at its core [it] embodies the autonomy of reason’ (p. 97). Reason itself provides the principles; no independent source is needed. Without a justification for the authority of reason, however, Westphal cannot be providing a complete justification. In parallel, he argues that the requirement of universalizability ‘incorporates’ respect for persons, thus obviating the need for any independent, realist value of personhood (p. 97). This is a plausible way of showing how the value of persons is a product of reason’s own structure and activity rather than any independent and prior value, and so does provide an excellent Kantian response to value realists, but again it leads to the question of the authority of reason itself. Thus, Westphal’s analysis does not avoid the issue of moral realism yet offer a complete theory. A complete theory needs to fully explain justification, for even to say that reason can be self-justifying without recourse to independent truth is to provide an anti-realist meta-ethical theory of justification.

So much for the constructivism in Natural Law Constructivism. One might at this point wonder what role the focus on natural law is supposed to play. We are told that ‘Kant’s constructivism articulates the content of a natural law theory, though it moots the issues of ontology … and motivation that confront traditional natural law theories’ (p. 103, my emphasis). Westphal appears to have in mind the basic principles and institutions needed for cooperative social life but does not provide a single comprehensive list or explain why he associates them qua content with natural law theory rather than other approaches such as social contract theory or Hegelianism (he does argue that social contract theory is a failed approach to grounding principles for social life, but this does not prevent those principles from having the same content). Nor does Westphal explain why Hume and Kant’s rejection of any naturally good aims or ends in life – a traditional feature of natural law theory – precludes them from being considered natural law theorists. Certainly both Hume and Kant on occasion applied the term to their own theories; however, both are also seen as providing alternative and substantially different approaches (based on utility and the a priori, respectively). While both Hume and Kant (among others) share some principles in political philosophy, the overlap they have with one another and with traditional natural law seems insufficient for identifying the result as a natural law theory in content. And of course Westphal holds that in form they reject the need for traditional natural law’s realism.

Still, the principles Hume and Kant provide and their arguments for them are well presented and defended in this book. In general, Westphal considers the need for principles for social cooperation. When these are specified for Kant, as in the discussion of Hume, they become the basic ground for possession of objects. One of the most impressive chapters is devoted to an analysis of Kant’s arguments for the principles that support rightful possession. Westphal very carefully and persuasively shows that Kant’s argument in the Metaphysics of Morals establishes rightful use of objects but not full property rights, which would include, for example, a right to destroy an object merely to prevent others in need from any potential future use of it. The arguments that Kant provides are in part based upon empirical conditions specific to human nature and the human environment, such as relative scarcity, but at the same time Kant uses the a priori universal principle of justice (or right), that (roughly) an action is right if it can coexist with everyone’s freedom in accordance with universal law, so he is not relying merely on empirical premises. Westphal uses this principle to apply a contradiction test to a ‘maxim of arrogant willing’ that would allow one to use objects possessed by others (p. 160). Since this maxim cannot be universalized, a principle for mutual recognition of individual possession of objects is necessary. This argument neatly applies the general assessment of the importance of the universalization test to the specific question of rightful possession to show how this sort of construction would work.

One underlying theme becomes more explicit in the final chapter. Westphal holds that moral philosophy ought to return to the position in which issues of justice are one of two coordinate species, alongside ethics, under the more general genus of moral philosophy, with justice as the requirement for social coordination serving as a basis for the possibility of individual ethical behaviour. He makes the strong case that political philosophy is not a subset of ethical behaviour, as if an individual first discerns and acts upon individual duties and only then adds as a further consideration a need for institutions for social coordination derivative upon the individual point of view. Rather the need for political and social institutions is itself independently derived from the broader moral considerations and is as essential for the realization of morality as is individual ethics. For Kantian scholarship in particular, Westphal is correctly urging the rejection of views that claim that the Doctrine of Right is not itself grounded by the Groundwork, or worse that it does not even belong in the Metaphysics of Morals, in favour of a view that puts considerations of external freedom at the core of the realization of morality as a whole.

Overall Westphal’s analysis highlights some very important points about Kant’s moral theory – the central role of rational agency in grounding other parts of ethical theory, the plausibility of universalizability as a test, the dependence of value and dignity on reason’s activity, the vital role of establishing external freedom for morality overall, the limitations of Kant’s defence of property largely to rightful use and the combination of a priori principle with empirical and anthropological considerations in giving content to morality. His attempt to read these as complete without consideration of the issues raised by the moral realism/anti-realism debate, however, falls short.