Rønnow-Rasmussen's book explores the distinction between two kinds of value: good, and good-for. The basic distinction is intuitive enough: there is a difference between something being good for someone, and its being good in some more impersonal sense. Still, it might seem that there must be something in common between these two kinds of goodness, such as that both are kinds of the same thing. Rønnow-Rasmussen provides a reductive theory of both kinds of goodness: in fact, a fitting attitude account of goodness, on which facts about value reduce to facts about the norms governing agents’ attitudes. On the proposed view, very roughly, something is good for someone if we have reason to favour it for their sake, whereas something is impersonally good just in case we have reason to favour it but not just for someone's sake. This theory enables us to see what good, and good-for, have in common. But at the same time, by having these two kinds of value demand different responses, Rønnow-Rasmussen argues that they conflict in an especially sharp way, so that we have a kind of choice about which to prioritise, and no obvious grounds on which to choose one over the other.
The book has many virtues. It is full of interesting puzzles and arguments about the distinction between good and good-for, as well as about the fitting attitude account of value, and will be of interest to anyone working on either topic. The book effectively continues his previous book, Personal Value (2011), and like that book, has an exploratory style which is endearingly humble, and which enables Rønnow-Rasmussen to explore the issues in a fair-minded way: the tentative conclusions are transparently driven by the arguments on offer rather than some desire to vindicate some predetermined conclusion. At the same time, at some points this style makes the content harder to grasp: it is sometimes difficult to tell whether some argument contributes something crucial or is instead an interesting digression or optional avenue for further exploration.
The book is split into two parts, with part I arguing for ‘value dualism’, the idea that good and good-for are two distinct kinds of value, with neither reducible to the other. Part II turns to argue for two corresponding fitting attitude accounts of value.
Part I has 5 chapters. Chapter 1 clarifies various distinctions between kinds of value, primarily between intrinsic and extrinsic value, and between final and non-final value. Chapter 2 turns to the crucial distinction between good and good-for, distinguishing three key positions: ‘Mooreanism’, which reduces good-for to good, ‘good-for monism’, which reduces good to good-for, and Rønnow-Rasmussen's preferred view, ‘value dualism’, which treats these two kinds of goodness as distinct and neither reducible to the other. Chapter 3 argues against Mooreanism – sometimes, things are good-for people in ways that are irreducible to any kind of goodness. And chapter 4 argues against good-for monism, arguing that some things are good in ways that are irreducible to facts about goodness-for. Chapter 5 is a mild digression, arguing that good-for encompasses more than just wellbeing: this chapter most obviously builds on Rønnow-Rasmussen's earlier Personal Value (2011).
The example-driven arguments across these chapters are largely compelling, but some room for doubt remains. For one, Rønnow-Rasmussen focuses on ‘formal’ as opposed to ‘substantive’ value theory, so that the challenges for these views consist partly in making certain (sometimes recherché) claims intelligible, even if they may be false. But then defenders of either of these views might just concede Rønnow-Rasmussen's arguments: so long as their preferred theory is true, does it matter whether opposition is intelligible? Or for another worry, Rønnow-Rasmussen sets aside versions of good-for monism which reduce goodness to anything other than good-for (51). But it is not clear why good-for monists could not pursue that strategy. A distinct concern about part I of the book is that possible motivations for adopting either monistic view are left unexplored, and given the tentativeness of the objections to those views, it's possible that such arguments might outweigh some of the alleged costs.
Part II of the book turns to the fitting attitude analysis. Chapter 6 introduces the view, suggesting that goodness should be understood in terms of reasons to favour, and good-for in terms of reasons to favour for someone's sake. Chapter 7 introduces a simple but forceful worry, which is that these analyses are inconsistent with value dualism: if you favour something for someone's sake, surely you thereby favour it. If so, it looks as though anything that is good for someone will, on the fitting attitude analysis, also be good. But that is inconsistent with value dualism. Chapter 8 introduces the solution: if the fitting attitude analysis of good-for says that something is good-for someone if there are reasons to favour it for their sake, we should modify the analysis of good to say that something is (finally) good if there are reasons to favour it and not for the sake of anything else. The suggestion here seems simple but plausible.
Chapter 9 explores the notion of a ‘sake’, which – it will be clear – is central in the FA analysis of both kinds of value. It argues, amongst various other things, that the concept of a sake is not an evaluative one.
Chapters 10 and 11 turn to questions about motivating reasons. Chapter 10 clarifies the notion of a motivating reason, and assesses ‘the guise of the normative’, understood as the claim that ‘what a person a favours is necessarily thought of, by a, as, or as providing, a reason for a to act’ (154). Rønnow-Rasmussen rejects this view, claiming that favourings only sometimes involve such normative thoughts. This is interesting, but we might yet wonder about more modest views. For example, Rønnow-Rasmussen leaves the notion of ‘favouring’ very open-ended, and if just any pro-attitude counts as ‘favouring’, the view seems to rule out the possibility any non-rational pro-attitudes, such as appetites, or likings. But defenders of views in the vicinity of the guise of the normative might try to exclude those from the remit of their theory (e.g. Gregory, Desire as Belief, 2021). Or for another worry, this view seems to say that every favouring includes a thought about a reason to act, but we might have thought that only some pro-attitudes such as desires and intentions are practical attitudes connected to reasons for action, and perhaps some other pro-attitudes might involve normative thoughts of other kinds. On this latter point, perhaps this is just an accidental slip on Rønnow-Rasmussen's part, since later the book considers the different idea that what a person favours necessarily appears to that person as giving them a reason to favour it (160), and perhaps this is really what Rønnow-Rasmussen has in mind.
That interpretation also fits with the fact that the first half of Chapter 11 seems to continue on the same topic, and presents further arguments that it is possible to favour things for no reason. Regardless, at this point you might reasonably wonder how chapter 10, and the first half of chapter 11, relates to the wider ambitions of the book. The idea in the second half of chapter 11 (174–80) is that the fitting attitude account should be modified so that – roughly – for something to be good in some respect R is for rightly placed agents to have reasons to favour it on account of R – i.e. to have R as their motivating reason for favouring it. This development of the FA account seems natural, and Rønnow-Rasmussen effectively sees off some possible challenges to it. At the same time, if this is the ultimate upshot of this discussion across these chapters, it might seem that discussion of the guise of the normative is something of a digression: so long as agents can sometimes favour things for reasons, that would suffice to permit this modification to the fitting attitude account. In this way, the discussion of the guise of the normative was interesting, but it was not obvious exactly how much it mattered for Rønnow-Rasmussen's wider ambitions.
Chapter 12 concludes by discussing the value gap that is the title of the book. Again, the central idea is that the divide between the two kinds of value is somehow alarming, telling us that we are subject to two competing sets of demands that are hard to reconcile. This idea is interesting, but the chapter is somewhat brief, and leaves various questions open. Here is one: we know that reasons can conflict: their ubiquity in contemporary ethical theorising is grounded precisely in the fact that ethical world seems rife with conflict that requires us to weigh competing considerations. So if the conflict between good and good-for just consists in competition amongst reasons that might well be a kind of conflict, but more could be said about why this particular conflict is more alarming than the many other conflicts between reasons that we are familiar with.
In short, I had some doubts about the central thesis of the book as defended in chapter 12, and various niggles about points of detail in various places. On the other hand, it is worth stressing that these doubts hardly undercut the interest of the book: though the central thesis makes the book sound more exciting, the two parts of the book might yet be true and interesting even if they don't have the dramatic implication Rønnow-Rasmussen claims they have in combination: it is interesting to understand the nature of good and good-for even if the gap between them is less heartbreaking than Rønnow-Rasmussen suggests. In short, despite these concerns the book remains highly interesting and would certainly reward repeated attention. I highly recommend it to those working on this topic.
