This is, in every sense, a big book. It weighs in, by my rough calculation, at over 200,000 words. It is hugely ambitious, covering a very large range of topics with great care. It is also novel in its approach to both metaethics and ethical theory. Because of its wide range and intricacy I will, of necessity, leave out much that is important. I am also conscious that, in not noting all the qualifications, I may have unwittingly misrepresented Mendola's view at various points.

Unsurprisingly, given his physicalism, Mendola holds that there are no non-natural properties, and so no ‘fully normative’ properties (p. 94). He argues, however, that physicalism is consistent with the existence of quasi-normative properties, so that physicalism can sustain an ethical theory that offers in many cases clear verdicts with sufficient normative force to make morality matter. Mendola dubs the novel and complex ethical theory that emerges, Material Morality (MM). MM is primarily a form of direct act-consequentalism that, perhaps surprisingly, delivers moral verdicts that chime reasonably well with the intuitions that often drive various non-consequentialist views, such as pluralist intuitionism and virtue ethics. I begin by looking at what else that is relevant to ethical theorizing is ruled out by physicalism, before turning to Mendola's positive theory.

First, Mendola argues that physicalism entails that many claims are indeterminate; while some degree of indeterminacy can be tolerated, what he dubs ‘virulent’ indeterminacy cannot. Why is this a problem for ethics? It is pointless to evaluate actions or institutions, unless there are available alternatives, but the well-known indeterminacy of some subjunctive conditionals means that, for at least some options, there is no fact of the matter as to whether an agent has that option. This claim seems undeniable, and many whose outlooks are far from physicalism agree with it. However, Mendola claims, the indeterminacy here is moderate, but not crippling to the appraisal of options.

Second, we ought not to conceive of agents as having lengthy life spans but rather as atomic agents, who should be thought of as lasting just long enough to have coherent basic preferences – about two seconds, apparently. One consequence of this is that virtually all the actions that anyone performs are to be seen as group acts. Not only do I cooperate with others in mutually desired projects, but the formation and carrying out of plans about my own life ‘requires temporally-extended cooperation among different periods’ of my life (p. 81). Indeed, at any moment, you are participating in a large number of group acts, and this (as we shall see) is key to defending the kind of consequentialism Mendola advocates. I suggested that Mendola's view of a life as constituted by a series of atomic agents may be a consequence of his physicalism, but I am not sure about this. On the one hand, he sometimes advocates this view merely as an analytical tool, but elsewhere he appears to think that this conception is backed up by the metaphysics (or lack of it) of identity. Given that we do not have what he calls ‘literal identity’, which, like Hume, he conceives of as endurance without change, some claims about us that depend on our having that kind of identity are unsupported. Here are two. First, just as there cannot really be compensation across lives, there cannot be compensation within a life ‘in any normatively deep way’ (p. 241). Second, ‘the future satisfaction of [A’s] current desires for his own future counts on my proposal towards his current well-being, his well-being from the perspective of the moment’ (my emphasis, p. 243). Mendola's proposals here are deeply radical. Some philosophers accept that a person's well-being may be affected by what happens pretty shortly after she has ceased to exist, but on Mendola's view this is the norm. By the time an atomic agent's desires are satisfied, that agent has ceased to exist. Indeed, one may wonder in what sense such fleeting creatures are agents at all.

With all this in place, we can turn to Mendola's positive theory about value and ethics. He rejects non-desire-based theories of well-being, such as the objective list theory, because there is nothing in the physical world that could ground such theories. Sophisticated variants of desire satisfaction theories are virulently indeterminate, since there is no fact of the matter about what our ideal self would do, or would advise our non-ideal self to do. The only account of well-being that makes sense in a physical world is a simple desire-based theory. Your good at some time is fixed by your all-in desire at that time, which is itself fixed by facts about what you actually prefer at that time among hypothetical (but not fanciful) options. Call these basic preferences. Assuming that they are reasonably coherent, we can build up an account of intra-personal and inter-personal well-being by appealing to a consensus of preference, both across time and across persons. People's preferences for things such as health over sickness, nourishment over starvation, are fairly stable.

There are well-known difficulties with desire-based accounts of well-being, but Mendola's seems to have extra ones of its own. Here are three. First, whatever consensus we come up with is going to be coarse- rather than fine-grained. The theory is not going to deliver even moderately specific answers to questions about any individual's well-being. Second, what you prefer is determined, not by your reflective rankings, but by what you would actually try to do, were you faced with various options. So, on this proposal, there is ‘no relevant difference between unconscious and conscious desires, between those that one articulates as reasons and does not, between those that rest on emotion and on cool assessment’, for all of these can be motivating (p. 187). Mendola argues that this conception is not as metaphysically indeterminate as properly informed desire accounts. However that may be, it seems to add a new indeterminacy: is there a fact of the matter about what someone would actually do in the case of options that they do not face but might? And even if there is, surely such facts are extremely difficult to discover; they seem close to being epistemically inaccessible.

Finally, for Mendola, our well-being is increased if our desires are satisfied. What I desire may be a future state of mine, or it may be the future state of some person(s) other than me. Ordinarily, this can make a great deal of difference to preference rankings. For example, I may prefer to sacrifice short-term desire satisfaction for a greater later satisfaction, where my nearer loss will be compensated for by my farther gain. I may not have the same preference ranking when it is my nearer loss that will furnish your later gain. However, as we have seen, Mendola holds that there can be no compensation, even within lives. Whenever a desire is satisfied as a result of a prior choice, the person whose desire is satisfied is not, strictly, the same as the person who made the choice. The claim that we are all atomic agents, while it provides some kind of metaphysical backing for the kind of impartiality between preferences favoured by consequentialists, is so antithetical to so many of our beliefs and practices that it requires more attention and support than Mendola gives it.

There are some preferences, Mendola holds, that are shared by nearly all normal human agents, such that we can reasonably say that there is a consensus. We accept certain considerations as reasons (though I think Mendola holds that, strictly, there are no reasons). From this contingent agreement, Mendola proceeds to construct his ethical theory by appealing to the following principles.

Of these claims, perhaps the most controversial is VLD, but since objections to it have been very ably put by Becka Williams in her review in the Notre Dame Philosophical Reviews (<https://www.ndpr.nd.edu/news/54560-human-interests-or-ethics-for-physicalists/>), I shall not rehearse these here.

These proposals about how to evaluate group acts are, I think, independent of physicalism, and so of general interest. Their importance for Mendola is that, on his account, group acts are much more common than we suppose, and it is to group acts that the consequentialist criterion is to be directly applied. I have already expressed doubts about the doctrine of atomic agents, but even in standard cases where a number of agents engage in some practice, what is it that makes a kind of act in which a group has an interest, a group act? Mendola's definition is plausible and, at first glance, fairly restrictive. A group act occurs when:

Given that acceptance, any member who stops coordination before the goal is achieved can expect to be the subject of criticism.

Just how common are group acts in morality, on Mendola's view? It seems that they are all pervasive. He claims that the practices that forbid lying and murder are group acts. Indeed, ‘it is even plausible to conceive of commonsense morality as a single widespread group act’ (p. 337), in which case all of morality falls under VLD. But this extension threatens to weaken the notion of a group act to the point of evanescence: a result, I think, of Mendola failing to distinguish clearly between coordination and cooperation. Here it is useful to invoke Hume, who explores an analogy to explicate his distinction between natural and artificial virtues. In both cases there is a shared aim, and criticism of defectors, but only in the latter is there a need for coordination. Those acts that fall under natural virtue can be likened to building a wall, while those that involve artificial virtue are akin to building a vault. In wall-building, the laying of each individual brick contributes to raising the wall, regardless of what others do. Thus each individual act makes things better, although perfection can be achieved only if all contribute. Here we need cooperation fully to achieve our goal, but not coordination. Contrast this with building a vaulted ceiling where, unless the bricks are all in place, the ceiling will collapse. Here we need coordination to achieve our goal at all, let alone perfectly, and sufficient defections threaten the whole enterprise. Without sufficient coordination, the practice will cease, or never get off the ground, so that it will be pointless to obey the rules of the practice. (Hume suggests that rules about property are like this.) So only activities analogous to building a vault require coordination, and hence are plausibly governed by VLD. The practice of abstaining from lying is, I submit, more like the building of a wall. Each act of veracity individually contributes to the goal, and there is reason to stick to the rule even when many others do not.

Behind Mendola's moral theory lies his physicalism, which he presupposes. In assessing the success of his project as a whole, I feel rather like the proverbial peasant who, on being asked directions, replied that he would not start from here. The most important thing that I take away from his book is a recognition of the radical nature of that world picture for ethics. Often, the debate between normative or ethical non-natural realists and their naturalist opponents centres on the existence of ‘spooky’ normative properties. However, at least as far as physicalism is concerned, the debate goes both wider and deeper. Physicalism threatens to undermine our sense of identity, of agency, and of reasons. Without these underpinnings it is hard to see how there can be ethics at all.