Joshua Gert has been carving out interesting corners of logical space for a long time. This book is no different. The middle portion of this book ought to become central to the debate about response-dependent accounts of normative properties. These chapters also make intriguing contributions to the debate about response-dependent accounts of colour. I recommend the book to all who are interested in the metaphysics of the normative. It is required reading for those who are interested in response-dependent accounts.
Gert's stated aim is to ‘demystify’ the normative. The strategy is to defend a response-dependent account of some basic normative concepts, namely irrationality and harm. Gert explicitly refuses to tell us what the normative is. Instead, he invites the reader to have as broad a view of the normative as she likes. He implicitly refuses to tell us what is mystifying about the normative. I’ll return to this at the end.
The focal piece of the beginning of the book is Gert's methodological view, which he dubs Linguistic Naturalism (LN). An important part of the method is a disavowal of traditional analysis. Gert is not in the business of offering necessary and sufficient conditions. Another important tenet of LN is that ‘stable parts of our language . . . are in perfect working order as they are’ (p. 7). This commitment means, roughly, that error theory is ruled out for ‘tried-and-true’ words and the concepts deployed when using them. The tried-and-true concepts are coherent concepts that have been indispensable for a long time. A final important tenet of LN is ‘a willingness to take very many domains as sui generis without worrying about it’ (p. 9). We are told that this is an anti-reductionist stance. Not only should we not hope for necessary and sufficient conditions, we should also not hope for interesting connections between different domains of facts.
Metaethicists should not, in my view, read this book in order to be convinced that LN is true. Nor, as I’ll argue shortly, should metaethicists read this book in order to find out why the normative isn't mystifying. They should read this book for a more local reason. Metaethicists should read this book because it offers a novel response-dependent view of irrationality and harm.
There are two features of Gert's discussion of response-dependence worth highlighting here. First, he takes an interesting stance on the classic problem raised by McDowell for response-dependent accounts of normative properties that rely on analogies with colour. McDowell famously argues that, in contrast to the case of colour, response-dependent accounts of normative properties must include normative concepts as analysans or explanans. Normative properties, the thought goes, correlate with the responses that it's fitting to have, not with responses we happen to have.
Gert argues that we must appeal to the normative even in the case of colour. Colour properties, like normative ones, correlate with the responses that it would be fitting to have in the relevant circumstances. The type of normativity here is an anaemic kind. It's the kind, Gert says, that ‘comes for free with the idea of representation’ (p. 77). In short, fittingness gets into the analysis just because the properties in question are objective – one can be wrong about whether they are instantiated. Gert argues that despite the anaemia of this normativity, it's crucial to the success of response-dependent accounts of colour that they build in fittingness.
Further, Gert argues that the type of normativity involved in response-dependent accounts of (basic) normative terms is the same type of normativity involved in the best response-dependent account of colour. This is because, just as in the case of colour, basic normative properties are objective. This objectivity stems from the fact that in both the case of colour and the case of certain basic normative concepts there is a large enough convergence in response (claims Gert). That is sufficient for objectivity.
The last point requires some care. Gert thinks that the objectivity of colours and of the basic normative properties stems from the fact that there is a large enough convergence in response. This is why some responses are fitting and some are not. Thus, the type of fittingness involved in Gert's response-dependence account of colour and his response-dependent account of basic normative properties is the same. This is how he responds to McDowell's challenge.
One might wonder why we should go this way. Why think that those with differing colour responses see things incorrectly? Why not simply think that those with differing colour responses see things differently? In response to this, Gert writes:
[T]here is no ‘why should’ about it; it is simply the way we and our languages do it. The practical exigencies of living in a world like ours, and the nature of our linguistic capacities and the ways they facilitate the application of pressure toward uniformity of attitude where that is useful, have yielded words for objective color properties. (p. 107)
Given Gert's acceptance of LN, our language functioning in this objective way suffices for objective colour and normative properties.
The second feature of Gert's account that is worth highlighting is his particular view about irrationality. Previous response-dependent accounts of normative properties all hold that the relevant response is the formation of particular conative states. Gert deviates from this tradition. Instead, he thinks that the relevant response for the type of irrationality he is interested in is a certain kind of befuddlement – what he calls ‘puzzlement’. When one has this response in reaction to a bit of behaviour, one cannot see the basic rationale for the behaviour qua action. When one is puzzled, one, at least initially, has a hard time interpreting the behaviour as goal directed.
Gert holds that the concept irrationality is response dependent in so far as it is true that an action is irrational just in case it is fitting to be puzzled by that action in the relevant circumstances. Importantly, he doesn't think this is an analysis of irrationality. But he does seem to think that this is as much as can be said to illuminate irrationality. Indeed, he holds that what is distinct about the response-dependent concepts is that the most illuminating explanation of their origins is in terms of a convergence of responses in the relevant circumstances. There will be this type of fitting convergence for other non-response-dependent concepts as well. However, for the non-response-dependent concepts, there will be something more to say about the nature of the stuff converged upon (see his discussion of water on p. 77).
He analyses rational in terms of irrational – an act is rational just in case it is not irrational. He then suggests we can analyse normative reasons in terms of these concepts. According to his suggestion (which he doesn't intend to defend fully), something is a reason for some reaction just in case it can turn that reaction from irrational to not irrational.
Gert's conception of irrationality is very weak in the sense that not that many actions are ruled out. This is by design. He makes it explicit that he thinks this is the only hope for the type of convergence that is needed to underwrite objectivity. Thus, many actions that many will consider irrational will come out rational on Gert's account. All that it takes to be rational is for the act to make sense or have a rationale. Many consider many actions that make sense irrational – the actions of those who disavow vaccinations, for example. Further, many will seemingly have normative reasons to do some heinous things if Gert's account is true, for all it takes for something to be a reason is for that thing to make it the case that some behaviour makes sense in some context. Many despicable things can go from not making sense to making sense in the relevant way.
With this much in hand, let's return to what is mystifying about the normative. Although it goes unstated, it is clear that Gert takes the normative to be mystifying in the same way that colours are. In particular, it is mystifying how these properties can both be objective and intimately tied to human responses. Gert – along with other response-dependence theorists – tries to demystify both types of properties via response-dependence. This has a strong claim for being the motivation for adopting a response-dependence account. That said, we shouldn't lose the response-dependence forest for Gert's trees. There is a good reason that traditional response-dependence accounts appeal to conative responses. It is because what has seemed mystifying is that normative properties can be both objective and motivating. By giving up this feature of the view, Gert might win some battles, but at the cost of losing the war. Once one strips away the dense (and often rewarding) discussions, it's not clear how Gert's view engages with what has truly mystified metaethicists.
For this reason, I don't think Gert has succeeded in his main aim of demystifying the normative. Even if you start out with the basic intuition that motivates traditional response-dependence accounts, you won't in the end find a view in Gert that is responsive to the purported features of the normative properties that are supposed to be mystifying. This objection notwithstanding, Gert's book is well worth reading. It contains many insights and is sure to advance the debate about response-dependence.
