It’s an unexpected honor to be included in this series along
with greats like Tom Hurka, Tim Scanlon, Sally Haslanger, Elizabeth Anderson, Nomy
Arpaly, and David Enoch, so let me start by thanking Dale and
the Daves for including me.  This summer,
between moving house and trying to enjoy some time with my daughter, who just
had her first birthday, I’m trying to spend some time reflecting on my work so
far, and how it fits together.  So I
thought that the best way to open this discussion would be to say something
general, and then open up the discussion to questions on any topic whatsoever.

Over the last ten years or so, I’ve written about a lot of
different topics, including reasons, rationality, supervenience, reduction,
expressivism, the conseqentialism/deontology divide, particularism, moral
semantics, the wrong kind of reason, pragmatic encroachment, truth, epistemic
expressions, and a little about the history of moral philosophy.  That might sound like a mess of unconnected topics,
but to me, they’re all tightly interrelated in interesting ways:

I’m a reductive realist about the normative, because I think
that this offers the best explanation of the necessary connections between the
normative and the non-normative, including supervenience.  I’m a proponent of the idea that reasons are
the key to understanding the normative, both because I think that this yields
the most promising kind of substantive normative theory and because I think
that reasons are the key to a plausible and defensible reductive normative
realism.  Key to understanding other
normative concepts in terms of reasons is the wrong kind of reason problem,
whose proper solution, I think, makes room for a better understanding of many
important issues in epistemology, including pragmatic encroachment on
knowledge.  I’m an increasingly
sympathetic critic of expressivism, one of the main competitor approaches to
reductive realism in metaethics, and have become particularly interested in its
most promising applications outside of metaethics, including to epistemic
expressions and truth.  And I’ve done a
lot of work using both metaphysical and linguistic arguments to examine the
argument structure of the relations expressed by normative words like reason,
good, and ought, because
only if we think clearly about what we are actually talking about are we going
to avoid making a variety of simple mistakes. 

I think that’s
about the best that I can sum up my views in a single paragraph.  I don’t claim to still believe quite
everything that I’ve said in print, however, so maybe I’ll get a chance to
clarify or revise in the discussion. 
Looking forward to hearing from you!

31 Replies to “Open for Questions…?

  1. Hi Mark
    I hope you are well and thanks for taking part in this discussion on your work. I really want to thank you for all your previous contributions on Pea Soup and also for all your books and papers. I have learned so much!
    There is only one question I would like to ask here. I’m sure you’ve addressed this somewhere but at the moment this isn’t quite clear for me from your work. I share your view that reasons are the basic normative concept. We should be able to give reductive accounts of all other normative properties in terms of reasons. In the case of value/goodness, this goes via something like the buck-passing account or fitting attitudes account. I was wondering, what kind of reduction are you thinking of in the cases of explicitly moral normative concepts such as right and wrong, kind and cruel and so on? How should we reduce these concepts/properties to reasons? I know that there are several strategies for doing this but it seems difficult to get the extensions right in the reductions.

  2. Hi, Jussi.
    Thanks for your great question. I have much less of a settled view about the moral concepts that you mention. I used to think that what is morally wrong is just what there is conclusive moral reason not to do, so as long as we can make a distinction between moral and other reasons, that would do the trick. I no longer think that’s true; I think a more careful account needs to be sensitive to things like the relationship between ‘wrong’ and each of ‘ought’ and ‘must’, the fact that ‘wrong’ but not ‘right’ seems to be gradable, and the existence of categories like the supererogatory and suberogatory. I’m not sure at the moment what I think of the ‘Millian’ idea of appealing to reasons of third parties to praise or blame, in addition to reasons of the agent to act, in order to explain what is wrong, although it’s one I’ve also flirted with. What I am pretty confident of, however, is that there is some close connection between what is wrong and the balance of some kind of reasons to act.

  3. More:
    I know even less of what to think about categories like kind and cruel. Many times I’m not sure which concepts I think are genuinely “thick” and which are just descriptive categories to which we attach an obvious importance. I’m not even sure that “kind” and “cruel” fall on the same side of this line. One way of thinking about this is to start not with gradable adjectives like “kind” and “cruel”, but with their comparatives, “kinder” and “crueler”. With “kind” I have the gut feeling that there is a non-normative “kinder than” scale, where we don’t have to answer any normative questions to know what is kinder than what, but we all probably agree that it’s better to be kinder than to be less kind. Whereas with “cruel”, I have the feeling that it’s a matter for normative inquiry what even counts as crueler than what. But this is outside of anything I’ve thought very much about. I do think that there are some normative concepts – “jerk” is one I’ve thought more about – that contain both rich descriptive content and rich normative content, but are not just a conjunction of two separate things. I think jerks are people who display an inappropriately callous disregard to others, or something very much in that neighborhood – where ‘inappropriate’ can be understood in terms of the balance of right-kind reasons.

  4. Hi Mark,
    Thanks for this opportunity to ask questions about your work.
    Related to Jussi’s question: I wonder what you think about whether there is a reason not to phi provided by the fact that phi-ing would be wrong. If we are trying to reduce normative properties to reasons, is there anything necessarily problematic about then also having those normative properties generate or constitute new reasons?
    A view like your initial one (from above) would seem to be fairly analogous to evaluative buck-passing, and so the fact that an act is wrong just would be for there to be conclusive reason not to do that thing (and so there would be no new reason generated — the buck would be passed). But if you have abandoned that view, do you have thoughts about whether there might be a reason not to phi constituted by the fact that phi-ing would be wrong?
    I’m particularly interested because I can desire to be a good person, and so if acting morally rightly in this instance promotes my being a good person, then the fact that this act is morally right seems to be a reason to do it on the hypotheticalist view you defend (so long as I haven’t butchered it). So if there are normative properties that must be analyzed in a more complicated way than simple buck-passing, as you suggest, then maybe even the reductionist can make sense of the reason, ‘because it would be wrong’?

  5. Thanks, Travis.
    Actually, I reject the thought that the fact that there is a reason, or is most reason, or is conclusive reason, to do A cannot itself be a reason to do A. So I think, for example, that the ordinary “buck-passing” account of value does not have this commitment, and I think the simplest view of ‘wrong’ would not have it, either. I believe there is an argument for this point in chapter 2 of Slaves of the Passions, but I set it out most clearly in a short little paper called “Buck-Passers’ Negative Thesis” in Philosophical Explorations. So if you reject the inference from “X is a claim about the existence or balance of reasons to do A” to “X cannot itself be a reason to do A”, as I do, then you can make sense of “because it would be wrong” in the way that you want to, even on the simplest view of “wrong”.

  6. Hi Mark, thanks for taking the time to answer questions!
    I would like to know why you think reductive realism is a better explanation for supervenience than expressivism (understood as a fundamentally explanatory project). I’ll suggest two reasons for thinking that it isn’t. The first is that it seems to me that attempts at reductive realism have failed very badly. For example, one of the problems I had with your (in many ways excellent) book ‘Slaves’ is that I was unable to understand how the account of the *weight* of reasons is reductive. Do you now think of either your own or some other specific account of the strength or weight of reasons to be both reductive and either correct or very promising? Or is the commitment to reductive realism more faute de mieux?
    As for expressivism, I think it should be committed not to explaining supervenience but rather our judgments of supervenience (this is the kind of ‘expressivism’ I am interested in at any rate). We could say that the expressivist is practicing explanatory (or pragmatic) naturalism rather than reductive naturalism at least in the sense that the expressivist isn’t trying to reduce reasons (or their weight) to natural facts, but rather trying to explain *judgments* of (strength of) reasons in naturalistic terms. The expressivist should also be looking to explain not supervenience but rather our judgments thereof.
    But assuming that we are natural creatures designed to operate in a natural environment, it shouldn’t be surprising that we make roughly the supervenience judgments we do, right? We ought to expect that we wouldn’t make different reason-judgments in two contexts where all the natural facts were identical (on the assumption that reason-judgments are in the action-guiding business, as the expressivist should think). That is, natural creatures operating in a natural environment, using their (naturally developed) concepts of (strength) of reasons, ought to not differ with respect to their normative judgments so long as the natural facts are identical. The disadvantages of doing so are obvious, and I can’t think of any advantages off the top.
    It seems to me that the attractions of reductive realism have much to do with the idea that the options are between metaphysical naturalism and metaphysical non-naturalism, and that the option of a fundamentally pragmatic naturalism (roughly, the kind of ‘subject naturalism’ or ‘global expressivism’ advocated by Huw Price) remains largely invisible. Given your sympathy to the deflationism about truth that largely motivates Price’s subject naturalism, I’m curious what you don’t like about this strategy. For my part, I think subject/pragmatic naturalism can take or leave deflationism, but I’ve already said too much as it is to explain why.
    Thanks again!

  7. Hi, Eric.
    That’s a lot of questions. Let me start at the beginning. I prefer reductive realism to expressivism because I am more optimistic about the overall package. Reductive realism offers a surprising view about what normative discourse is about, within a package that takes advantage of what our best independent theories of language and the mind have to offer, and in a way that make moral epistemology easier. Expressivism requires a new and not wholly worked-out picture of thought and language, and leaves some real puzzles about what moral epistemology is doing and how closely we can respect the phenomena.
    As far as explaining supervenience goes, I don’t think that anyone has fully spelled out how this is supposed to go for an expressivist, although I think Gibbard’s approach in Thinking How to Live is the most promising and I think I know how I would spell it out myself. If I were an expressivist, I wouldn’t substitute an explanation of our judgements of supervenience for an explanation of supervenience, as you suggest, I would explain why supervenience itself has the same status as other analytic truths. I discuss how this sort of strategy works in general in chapter eight of Noncognitivism in Ethics. This is consistent with the idea that we explain normative language by understanding what it is to have normative thoughts, rather than in terms of what it is for something to be true.
    I think it’s a strange charge to send at me in particular that reductive realism is attractive only because the option of expressivism is invisible. But maybe you mean that global expressivism like Huw Price’s is invisible, rather than that topic-specific expressivism is. But I’m not sure what that means. I think any workable expressivism is going to be global in the sense that it applies to all language, so in that sense global expressivism is no surprise. But if you mean a kind of global irrealism, it’s true that I haven’t worried a lot about that. For one, I still can’t get my head around that idea, or see how to make sense of it in independent terms. I also don’t think it will help you explain supervenience – indeed, the Gibbard-style explanation of supervenience that I favor requires a notion of a possible world that is individuated in terms that are not friendly to global irrealism. And I don’t see the path that you suggest from expressivism about truth to global irrealism. I don’t really understand how that path is supposed to go.

  8. Thanks Mark. Sorry, I didn’t mean to send a charge at you, but just note that it seems that metaphysical (whether reductive or non-reductive) naturalism seems often motivated in large part by the idea that the alternative is metaphysical non-naturalism, or versions of ‘expressivism’ (it might be better to use a different term) that are committed to non-cognitivism or some other of the expressivists’ traditional negative theses (including irrealism).
    Of course the traditional versions of expressivism are not invisible to you, but I meant the specific sort that has no truck with these negative theses, but is simply committed to providing use-explanations of problematic philosophical concepts (perhaps being quietist about metaphysics a la Price–though that is not my approach).
    Price’s path is from linguistic priority and semantic minimalism to global expressivism (understood as providing use-explanations), not irrealism. I really didn’t know whether you had considered Price’s ‘global expressivism’ and if so, what you found unattractive about it. In reviewing the literature and talking to philosophers, this view still seems nearly invisible to me among metaethicists.
    It also seems to me that global expressivism would be a surprise to almost all the expressivists out there. For example, Gibbard and Blackburn don’t think of scientific discourse as warranting an expressivist treatment. It seems to me that expressivism is still very widely considered a form of non-cognitivism (or nonfactualism, or irrealism), where this is thought to apply only to selected domains. If there are expressivists out there who think that scientific discourse warrants an expressivist treatment, please let me know. I know that Blackburn has explicitly rejected this kind of global expressivism, though Price has used Blackburn’s own work to argue in favor of it.
    For what it’s worth, I’m working on a book manuscript, part of which is to argue for a pragmatically naturalistic approach to metanormativity. Pragmatic naturalism in my sense has two primary goals. The first is to provide use-explanations of problematic philosophical concepts. The second goal is, where appropriate, to offer (sometimes constructive, sometimes destructive) critiques of target concepts in light of our understanding of their functional significance. This aspect of pragmatic naturalism marks a departure from both Price and expressivists, who (to my knowledge) make no attempt to provide any such criticism. If you know of any other philosophers adopting such an approach, it would be very helpful if you could point me toward them. I do not know of any.

  9. Maybe this is a question of terminology Mark, but you said above:
    “I’m a reductive realist about the normative, because I think that this offers the best explanation of the necessary connections between the normative and the non-normative, including supervenience.”
    Do you think normative facts are strictly identical to non-normative facts? If so, what are the necessary connections you’re alluding to? It’s hard to see what interesting necessary connections there are between x and y, when x = y. On the other hand, if the normative and non-normative are not identical, in what sense is your view a reductive account? Normative facts hold at least partly because non-normative fact hold and not the other way (itself a fact not captured by supervenience theses). This fact would seem to suffice for the irreducibility of the normative.

  10. Eric, Gibbard does give scientific discourse an expressivist treatment. It’s just not plan-laden. I don’t know how to say in what sense his view is not expressivist about all discourse without using some negative thesis. And I don’t see how questions about metaphysics or epistemology go away by not saying something about metaphysics – it’s just less clear what it would take to answer them.

  11. Hi, Christian.
    It can’t be a matter of terminology whether normative properties have to be identical to non-normative properties for reduction to be true. That just defines makes reduction false by definition, because no property can be both normative and not normative at the same time. In my work I’ve defended the view that reduction is property analysis. Supervenience is not identity, and supervenience involves a necessary relationship.

  12. Hi, Mark –
    Thanks for doing this. I have a question about how some of the views you’ve defended fit together. (I think we may have talked about this before, but I confess that if so I can’t remember what you said. Apologies for the repetition!)
    On the account of the right kind of reasons you give in ‘Value and the Right Kind of Reason’, the right kind of reasons with respect to some activity are, very roughly, those shared by everyone engaged in that activity, along with reasons which are derivative from such reasons. Activities are understood very broadly so as to include things like being a believer or intender. So, for example, the right kind of reasons to intend are those shared by everyone engaged in the activity of intending.
    An initial worry is that this seems to rule out agent-relative right kind reasons. For example, the fact that there will be dancing at the party is a reason for Ronnie to go to the party and (so) a right kind reason for Ronnie to intend to go to the party. But the fact that there will be dancing at the party is not a reason for all intenders to intend to go to the party – it’s only a reason for those who like to dance. So your account seems not to count it as a right kind reason.
    One way out of this would be to say that Ronnie’s reason to go to the party derives from a reason to promote his desires which is shared by all intenders. However, this option seems to be ruled out by your arguments against the ‘standard model’ of reasons explanations in Slaves. There you argue that reasons to promote one’s desires should not be understood as deriving from a general reason, shared by all agents, to promote one’s desires, since there’s no good candidate to be this reason.
    So I’d be interested to hear what you thought of this. Of course, another option would be to appeal to a different activity – for instance, you could say that Ronnie’s reason is right kind with respect to the activity of being an intender who likes dancing. Here I’d worry that this claim seems too weak. Without knowing exactly how to argue for this, I would have thought that Ronnie’s reason is right kind in just the same way that, e.g. his reasons to intend to keep his promises are. But as I say, I’d interested to hear what you think.

  13. Hi, Jonathan –
    Thanks; that’s a great and forceful question. The ‘shared by everyone’ versus ‘derived from one that is shared by everyone’ distinction was the move that I made between chapter seven of Slaves and ‘Value and the Right Kind of Reason’ in order to try to accommodate agent-relative right-kind reasons, so you’re right about the move I would have made. I actually talked myself out of the view in ‘Value and the Right Kind of Reason’ for different reasons – I think because right-kind reasons for intentions and beliefs do not transmit to right-kind reasons for means to those intentions and beliefs in the way that the ‘derivative’ model would seem to predict.
    As a result, I no longer think that we can get the right-kind/wrong-kind distinction merely out of who shares the reason. However, I do still think that right-kind reasons are the ones that arise out of some activity, and that my original account can be thought of as a way of making sense of the ‘arise out of’ relation. I retreated to this gloss in ‘The Ubiquity of State-Given Reasons’.
    Your objection, however, is different. You note that if right-kind reasons for intention are often intuitively object-given, and reasons for action can be agent-relative, then the corresponding right-kind reasons for intention will be both agent-relative and not derivative from some reason shared by all intenders. This assumes that the reason that RKRs for intention are derivative from is the reason for action, of course – if there is a general reason to intend what you have reason to do that is shared by all intenders, then the fact that Ronnie has a reason to go to the party could be the feature of his circumstances that gives him a derivative reason to intend to go to the party.
    So that’s the move that I would feel forced to make in order to fit this to the view from ‘Value and the Right Kind of Reason’. It’s not directly subject to the objection I leveled in Slaves to the Standard Model treatment of Ronnie’s reason, but it might be subject to analogous objections. So I think I hope that there will be a good way of making sense of the ‘arise out of’ relation that doesn’t share this commitment. I don’t have that solution, at this point, but I feel pretty confident that I’m closer to at least characterizing the right ballpark for an account of the right-kind/wrong-kind distinction.

  14. Thanks Mark. Is that in Thinking How to Live? Sorry, I think the topics I’ve raised are too difficult to discuss here, so I will ask shorter, hopefully more tractable questions more directly connected to your own work.
    In Slaves you say that any Humean Theory of Reasons must say that all desires generate reasons (88). I’m developing a version that doesn’t say this. All reasons are explained by (and relative to) desires, but it is not the case that all desires generate reasons (I view this metanormative claim as itself normative, but it can be interpreted in the traditional metaphysical way as well).
    I think this is both more independently plausible and consistent with the core of Humeanism about reasons, which I take to be that 1) desires are never intrinsically irrational and 2) all reasons for action are in relation to some desire(s).
    Do you still think that any HTR must hold that all desires generate reasons and if so, why?

  15. Hi, Eric.
    I look forward to discussing the general issues about global expressivism with you some other time.
    I think you’re being misled on page 88 of Slaves by the fact that throughout the book up to chapter 8 I use “desire” in a stipulative sense to refer to whatever psychological state it is that the Humean takes to explain reasons. I do this because the issues up to this point in the book cross-cut views about what kind of psychological state this is. This could, in principle, depart quite considerably from, or be quite a substantial restriction on, desire in the ordinary sense. However, I go on in chapter 8 to defend a view that I think fits pretty closely with ordinary use of “desire”.
    Does that help?

  16. Thanks Mark for taking the time to read and respond to “fan mail” (I am a fan!)
    In Slaves of the Passions you defend a Humean analysis of reasons that goes (very, very roughly) like this:
    (1) a state of affairs R is a reason for Aing iff R helps to explain why Aing promotes one’s desires
    And in chapter 8, you say something like the following at least initially seems plausible as part of the analysis of desire:
    (2) a mental state D is a desire iff, among many other things, D disposes S to take certain states of affairs (namely, those that help to explain why Aing promotes D) as a reason for Aing
    It appears that (1) and (2) are circular, and indeed that’s what you claim on page 155. But the issue of circularity isn’t probed in any detail, so we don’t hear detailed arguments about the nature of the circularity and whether it is vicious or benign. I want to press you on this issue.
    I don’t think there is really any vicious circularity going on. Things only appear a little circular because DESIRE and REASON are *interdependent* notions. But interdependence and vicious circularity are two different things.
    Consider the notion of the LAW and an OUTLAW. The law has many functions, one of which is to specify punishments for outlaws. An outlaw is a person whose actions stand in the violate relation to the rules specified in the law. Each notion uses the other in its analysis, but there is no vicious circularity going on here. Rather, law and outlaw exhibit healthy interdependence. (I could offer a detailed argument for this, but I am hoping it is intuitively clear that LAW and OUTLAW are not viciously circular notions.)
    DESIRE and REASON stand in a broadly analogous relation. From a design perspective, it makes sense that among the sequelae of desires (dispositions to act, attend, and feel certain ways) there would be one further disposition: the disposition to recognize those states of affairs relevant to promoting the satisfaction of the desire as being in some way special. These special states of affairs are of course what are picked out by the term “reason” (so says the Humean analysis). So reason appears in the analysis desire and desire appears in the analysis of reason. But I don’t see any vicious circularity here. Rather like LAW and OUTLAW, DESIRE and REASON appear to be only healthily interdependent.
    I am curious to hear your take on the plausibility of this line of argument.

  17. Hi, Chandra –
    Thanks! Officially, the view in Slaves is that we can replace the last part of your (2) with a phenomenological characterization of certain things “striking you as salient” and “prompting action in a non-alienating way”, in the analysis of “desire”. Then (2) turns out to be true after all, but not an analysis – which is why I claim it’s not circular.
    Your way of thinking about things is different. I think that I would have to revise the view about analyses that I advocate in chapter 4 in order to make room for it, but that’s not an argument against it! The suggestion that “desire” and “reason” might be linked in such a way is one that I’d like to think about more – unlike most such examples, this one looks like it “bridges” the normative/non-normative gap in an interesting way that could preserve many of the intuitively “naturalist” motivations of the project.

  18. Thanks Mark. I look forward to it too!
    Thanks for reminding me that you are using desire that way to that point in the book (for the view to be Humean, I think we should understand ‘desire’ as in some way(s) inherently connected to motivation, but we don’t need to say more yet). Still, what I’m wondering is why it has to be the case that all desires give rise to reasons, even if all reasons are relative to desires.
    I think it’s key to defending (the best possible) Humeanism about reasons that we not go from thinking ‘if reason, then desire’ to ‘if desire, then reason’. It seems to me that the impetus to think ‘if desire, then reason’ stems in large part from the idea that if one denies this, then one will feel required to provide a (nonnormative) criterion separating the desires that do from the desires that don’t provide reasons, and this task looks daunting.
    I think we should reject this requirement, but instead of saying why, I’d like to know if and why you think that Humeans, who are committed to explaining reasons in terms of desires, must (therefore?) think that all desires must give rise to reasons.

  19. Hi, Eric.
    I don’t really care how we use the word “Humean”, but the problems in the first seven chapters of Slaves are neutral as to whether the psychological state that explains reasons is motivational in character, so in the book I intend “desire” to be even more neutral than that.
    I also don’t think the task of providing a distinction between the desires that provide reasons and those that do not is “too daunting”. In Slaves I explored the most expansive possible reading of what sort of psychological state gives rise to reasons partly for methodological reasons – I don’t think there is a way of restricting which states give rise to reasons that changes the issues facing Humean views in kind, as opposed to degree, so I explored a view that I thought made those issues the most vivid.

  20. Ok, thanks, that’s helpful. By the way, I should also indicate that I’m a fan :).
    I still don’t see why a humean theorist as such should be committed to the idea that all desires (even in the very neutral sense) must give rise to reasons, just because reasons are explained in terms of desires. But maybe you don’t think this after all.

  21. Hi, Eric.
    Let’s follow my usage in Slaves and use ‘desire’ to mean that psychological state, whatever it is, that is sufficient to explain the existence of a reason, given uncontroversial background facts. Now it follows that whenever there is a desire, there is a corresponding reason. I don’t see what the problem with this can be, unless you think that there is no psychological condition whatsoever that is sufficient to explain the existence of a reason. Is that what you mean?

  22. Thanks, Mark. I share the suspicion that appealing to a general reason to intend what you have reason to do will raise worries analogous to those you raise about the general reason to promote your desires. It may also be worth noting that similar worries come up for other attitudes – e.g. Ronnie might have agent-relative right kind reason to wish that there were more places to go dancing nearby. An agent-neutral reason to intend what you have reason to do wouldn’t help here. Of course, there might be an agent-neutral reason to wish for opportunities to promote your desires. But I imagine that decent candidates to be that reason will be also be (or be closely related to) decent candidates to be the reason to promote your desires.
    I was thinking that the worry about agent-relative right-kind reasons looks like a more general worry than the one about reason transmission. At least in principle the latter could be avoided by being more specific about the kinds of derivation which generate right kind reasons. The former seems more directly to target the idea that right kind reasons arise out of a shared activity because – at least if the worries about deriving agent-relative reasons of this sort from agent-neutral ones are right – they don’t seem to arise simply from the activity of intending (or wishing, in the other example).

  23. Hi, Mark.
    On hybridism – According to hybrid theories, accepting a moral sentence involves more than just having an ordinary belief. As you say (2009), in claiming this, hybridists could potentially facilitate a non-coincidental explanations of moral motivation whilst also having an advantage over pure expressivist views in explaining moral inferences. But as you argue, they simply cannot have it both ways – getting the advantages of cognitivism in explaining logic and inferences and the advantages of expressivism in explaining moral motivation.
    According to Horgan and Timmons, accepting a moral sentence involves more than having an ordinary belief – it involves having a moral belief (ought-commitment), it involves coming down motivationally on what ‘feels’ like an objective matter of fact. From H&T’s point of view, understanding moral beliefs as being a ‘genuine kind of belief’ despite the fact that their overall content is not descriptive, allows them to try to improve expressivism with regards to logic and inferences (by appealing to logical complexity of moral beliefs, their constitutive inferential role…), whilst also explaining moral motivation by utilizing the expressivist ‘elegant’ resources. And I say, a similar kind of criticism applies: they cannot have it both ways. Do you very much disagree?
    We know that it would be implausible to apply your cost-benefit analysis of hybrid views to H&T’s cognitivist expressivism: H&T do not hold a hybrid view since they do not believe that moral sentences express two states, one of which fits the traditional Humean conception of belief and one which fits in the traditional Humean conception of desire. Also, they themselves do not distinguish (much) their views from Blackburn’s. And thank you, (Bart Streumer also) for deterring me from joining those who thought that you should have perhaps included H&T in your survey paper on hybridism. Nevertheless, in trying to make sense of H&T’s anti-Humean conception of practical reason in particular, their view that moral judgements dispose us directly toward appropriate action, independently of our pre-existing desires, their ‘morphological rationalism’, etc, I still wonder to what extent they can really use some of the expressivist resources. But regardless of how we categorise their meta-semantic project I think that their hybridising efforts, where the label ‘hybrid’ is used more loosely, are subject to the similar kind of tension/worry that you address in your paper.
    The problem, as I see it, is that no matter what the scope of improvement for the expressivists may be, and particularly in terms of their ability to answer the Frege-Geach question, explaining moral motivation is never going to be easy for them. Now, we know that expressivism, whether pure, ecumenical, of H&T version, is motivated by the idea that judgement motivational internalism is necessary and a conceptual truth. But if this assumption is false, then any friendly suggestion to expressivists and which they pursue will be ill-grounded as far as the motivational story goes. If you still don’t think, as I don’t, that judgement motivational internalism is a conceptual truth then you don’t think that a promising way to go about hybridism entails a promise of a plausible explanation of moral motivation. Or do you?
    I look forward to seeing your forthcoming paper on hybridism, and I wonder if you have made any significant revisions with regards to your original views and objections.

  24. Hi Mark, thanks for putting up with what seems a difficult question for me to adequately express. What I’m wondering is why we have to start the way you suggest. Why, that is, should we start by requiring that there are any psychological states that are, given (nonnormative) background conditions, sufficient to explain the existence of a reason? I understand that there’s a very widespread assumption to this effect, and that given this assumption, your methodological strategy seems a good one. I began my own defense of Humeanism implicitly assuming it.
    But it seems to me that Humeanism is best thought of as committed to the idea that for every reason, there is a desire (expansive reading). It doesn’t thereby have to be committed to the idea that for every desire, there is a reason. Or does it? If so, I’d like to know why.
    If it doesn’t, then I think it can better deal with the Too Many Reasons Objection, which is driven by this assumption. If it does, then perhaps your responses to that objection in Slaves are about as good as we’re going to get. And that might be good enough, but it seems to me those responses, which allowed that we have reason to eat our cars, undermined your guiding claim that Bradley didn’t have reason to go to the dance. But unless Bradley is very much stranger than we were given reason to suppose, then he does have reason to go to the dance if I have reason to eat my car, right? I probably hate eating my car more than Bradley hates to dance. But if Bradley has reason to go to the dance, then many of the arguments in that book are either not going to go through or are going to get more complicated, right?
    Thanks again for your patience. This will be my last effort on this question in any case. If I have time in the next few days (leaving town), I’d like to ask you a different one.

  25. Thanks, Jonathan –
    I see the point that the reasons don’t arise wholly out of the activity of intending. Now I have a new constraint to worry about…

  26. Hi, Amna –
    Good to hear from you. My worry about Horgan and Timmons’s Cognitivist Expressivism is not so much that they can’t have it both ways, as that it depends on an extraordinarily extravagant set of assumptions. They postulate an infinite list of kinds of belief, and need an infinite list of assumptions about how those kinds of belief are related to one another. I found that objectionable enough that I didn’t worry too much about what else might be problematic about their view.
    I don’t see any problem for them in accounting for how ‘ought’ beliefs motivate, because they built the motivational role for ought-beliefs into their account of what ought-beliefs are. So on the assumption that the goal is to vindicate a kind of judgment internalism, I see no problem. Perhaps the problem is with vindicating why this is a conceptual truth, but I’d need to see that thought spelled out better before I responded to it.
    You, however, are not sympathetic to judgment internalism, at least as a conceptual truth. Some hybridists, of course, don’t want to vindicate judgment internalism at all, but only the appearance of judgment internalism. David Copp belongs in this category, so he finds it very strange to read discussions of hybrid views, like mine, which assume that vindicating judgment internalism is the goal. Others, like Ryan Hay, are interested in vindicating only a very weak version of judgment internalism.

  27. Eric –
    I wrote Slaves of the Passions with two goals. One was to articulate a concrete and defensible reductive normative view. And one was to offer a distinctive way of thinking about the attraction of the Humean Theory of Reasons without getting caught up in a lot of wrongheaded ideas about its commitments.
    I began the book by characterizing the Humean Theory – for my purposes – as the idea that every reason is explained by a psychological state in the same way that Ronnie’s is. So that’s the view that I’m talking about in the book. You’re right that there is an even weaker thesis, that every reason requires a corresponding desire, and you’re right that the Too Many Reasons objection doesn’t apply to it, but I don’t know of an argument that I find convincing to believe that is true that doesn’t turn on claims about explanation or reduction.
    Last thing: I don’t think it follows from the fact that you hate eating your care more than Bradley hates dancing that there is a reason for Bradley to go to the party, much less – which is the relevant point – that the fact that there will be dancing at the party is a reason for Bradley to go there. Tristram McPherson has made some nice points about how the dialectic of the book gets more complicated if the fact that there will be dancing at the party is in fact at least some reason for Bradley to go there.

  28. Hi Mark,
    I had a few questions about your response.
    “It can’t be a matter of terminology whether normative properties have to be identical to non-normative properties for reduction to be true. That just defines makes reduction false by definition, because no property can be both normative and not normative at the same time.”
    This isn’t obvious to me. I think something can be short and tall (not short) at the same time; something can be good and bad (not good) at the same time; and maybe something can exist and not exist at the same time. In the first case, something can be short for a basketball player and tall for a child. In the second case, something can be good for you and bad for the the world. In the third case, something can (perhaps) tenselessly exist and not exist now. There are, seemingly, many cases like these.
    Many seemingly intrinsic properties turn out to be relational on analysis. One property can be picked out in different ways, and perhaps one way is a normative way, and the other is a non-normative way; and whether a property is normative depends on how it gets picked out. In this sense, something can be normative and non-normative. A thing’s being normative is a relational property of it because this property relates the thing in question to the way it is picked out. Maybe you think being normative and being non-normative are intrinsic properties? If so, why think that? It’s just not at all obvious why we should take that for granted. Take *being in pain* as an example. It seems to count as normative and non-normative in a fairly straightforward way.
    “In my work I’ve defended the view that reduction is property analysis. Supervenience is not identity, and supervenience involves a necessary relationship.”
    The property being a bachelor can be analyzed as the property being an unmarried adult male. Concepts are a kind of property. Are you saying that we can give a similar analysis for evaluative properties (contra Moore) in terms of non-evaluative properties? That is, an account which states that ‘x is good iff _______’ where what fills in the blank contains no evaluative terms and analyzes the good. I think this would be a really big deal.
    Have you proposed such an analysis, and if so, where?

  29. Hi, Christian.
    I’m not prepared to countenance true contradictions, and it doesn’t sound from your comment like you are, either. You don’t really think that something can be both tall and not tall at the same time, except by equivocating on what it takes to count as ‘tall’. You think that’s possible because ‘tall’ is context-dependent. I don’t think that ‘normative’ is context-dependent, and if it were, the claim that something can be ‘normative’ and not ‘normative’ at the same time, by filling in different contextual parameters, would not be a very interesting thesis. The idea that normative properties have to be identical to properties that are not normative (on a single, uniform reading of ‘normative’ for a reduction of the normative to the non-normative (on a single, uniform reading of ‘normative’) to be true is I think one of the least cogent ideas of the last fifty years of metaethics. With a few exceptions, such as Alex Miller’s textbook, most people who think that reduction is property identity don’t even consider whether the normative can reduce to the non-normative; they only consider whether it can reduce to the ‘natural’ or the ‘descriptive’ or something else that is not by definition incompatible with being normative.
    I have indeed proposed a strategy for analyzing normative properties in terms of non-normative properties. You can find it in my book, Slaves of the Passions.

  30. Hi, Mark.
    Thank you for your response! There is nothing strange about you discussing hybrid views by assuming that vindicating judgement internalism is their goal. You are saying something which is true. There may be something strange about a hybrid view (unless somehow incredibly cognitivist-leaning) that vindicates the appearance of judgement internalism, whilst not accepting the truth of judgement internalism – weak or otherwise. Hybridism, after all, descends from pure non-cognitivism, and vindicating judgement internalism (not just the appearance of it) is one of the major forces driving us towards non-cognitivism. We all share internalist intuitions, but as externalists – who must be cognitivists – point out, this intuition should not be given the fundamental role in metaethics.
    A major attraction of your paper to which I referred earlier, is that it highlights one of the problems for hybrid theories –the problem with vindicating judgement internalism. Horgan and Timmons’ cognitivist expressivism, which also descends from traditional non-cognitivism, is also problematic in this respect. I recognise that, in the later case, the details and methods are significantly different. An answer I was trying to tease out from you earlier – and I admit not very successfully – concerns the question as to where do you go from here. Given that you are, amongst other things, ‘an increasingly sympathetic critic of expressivism’ , and I assume this includes hybrid views, are the solutions you propose to the problems facing the hybrid theories you are interested in (i.e. M. Ridge), supposed to also lend more support to the internalist explanation of moral motivation? That would be ok – something for me to worry about, not you. But I guess I wonder about the way in which your suggestions can be consistent with your view that judgement internalism is not a conceptual truth. Do you think that some form of judgement internalism is plausible such that can be supported by a hybrid position (i.e. R. Hay)?

Comments are closed.