| Welcome to what will hopefully be a very interesting discussion of Dale Dorsey‘s “Moral Distinctiveness and Moral Inquiry”. The paper is published in the most recent edition of Ethics and available to read through open access here. Kathryn M Lindeman has kindly contributed a critical précis, and it appears immediately below. Please join in the discussion! |
Dale Dorsey’s paper “Moral Distinctiveness and Moral Inquiry” is a clearly motivated and impressive contribution to the literature on foundational questions in Ethics. I am very glad Ethics has chosen to publish this paper and that PEA Soup has chosen to provide this space for us to discuss it.
Dorsey’s concern is with how it is possible to settle what distinguishes moral standards from other standards according to which we can evaluate actions (e.g. those of prudence, honor, etiquette, sportsmanship, etc). He is specifically concerned with the feasibility of what he terms the project (italics his) – the attempt to determine the unique content of the moral domain, prior to substantive first-order inquiry. His conclusion is that the project is not feasible, that contrary to the hopes of moral philosophers, we cannot know what is distinctive of morality prior to substantial first-order deliberation.
In what follows, I’ll do three things: (I) I’ll quickly explain the motivations others have for undertaking the project; (II) I’ll summarize the argument Dorsey uses to motivate skepticism about the project and his positive view of how we should understand the distinguishing feature of the moral domain independently of the project; and (III) then I will raise two worries: one about the argumentative assumptions Dorsey makes and one about where we are left if the negative argument against the project is successful.
(I) Dorsey begins by jointly motivating not only why we might want a way to distinguish moral from other standards applicable to action but also why we might want to do so before engaging in first-order moral inquiry. He covers a lot of ground, explaining how various moral theorists have motivated and engaged in the project. There are two main reasons Dorsey highlights.
First, determining what is distinctive about morality can set the terms of moral inquiry. Warnock, for example, thinks that in order for moral philosophers to know what they are talking about, they must have an account of what is distinctive of morality. Further, this is critical if moral philosophers are going to be sure that they are all talking about the same thing. Without this, disputants might just be talking past each other.
Second, Dorsey notes that the project aspires to provide distinctive features of morality that can serve a gatekeeping function. In deciding between moral theories, one important (and frequently used) tool available to us is comparing the features of those theories with the commitments we gain from having an answer to the project. Moral theorists frequently use claims about what is distinctive of morality in arguments against other first-order moral theories in ethics (or, in the case of nihilists, of morality in general). And we must have a sense of what morality is up to in order to conclude that some particular account of morality isn’t up to snuff. I’ll return to these points in (III).
(II). Despite these two appealing uses for determining what is distinctive of morality before substantial moral inquiry, Dorsey presents a striking argument that the project cannot succeed. He considers and dismisses five ways to account for morality’s distinctiveness: its content, how it is grounded, the relationship it has to reactive attitudes, the relationship it has to motivation, and its distinctive normativity. None, he argues, can be established prior to substantive first-order inquiry.
The whole of the argument against the project ends up turning on the first possible distinctive feature: its content. All of the other ways morality might turn out to be distinctive turn on the determinability of its content, prior to substantive first order theorizing.
Dorsey takes content to be saddled with what he calls the “coarse/fine tension.” Content, to be distinctive of morality, must be fine-grained enough to distinguish morality from non-moral domains; but to satisfy the project it must be coarse-grained enough not to rule out any plausible first-order moral theories. So, for example, utility-maximizing cannot be the property distinctive of moral content, because not all first-order theories would take this to be distinctive of morality, and so it is not coarse-grained enough. However, this property also cannot be something like ‘being other-oriented,’ because it seems to be shared by content that is not in the domain of ethics, like the dictates of etiquette.
Dorsey argues these two are in tension. No account that is fine-grained enough to uniquely distinguish morality will be coarse-grained enough not to rule out a first-order theory prior to first-order deliberation. Dorsey concludes, “In sum, morality clearly has a distinctive content. So much is trivial. But we should reject the claim that a content-based account of morality’s distinctiveness could satisfy the project” (Doresy, 760).
The main objection Dorsey raises for taking the grounds of the moral domain to be its distinctive trait relies on his argument about content. He notes that often grounding properties radically underdetermine the content of the resulting domain. He argues that we will need to check to make sure that the claimed grounding property actually uniquely and appropriately determines the content of the moral domain. But this means that we need an understanding of moral content to check whether the grounds of the moral domain are doing their job. Thus, because we cannot determine the content of morality prior to first-order deliberation, we will be unable to determine its grounds prior to this deliberation either.
The last three possible ways of determining what is distinctive of morality—its connection to reactive attitudes, motivation, and normativity—are addressed together. I’ll just give the concluding argument of this section, roughly as Dorsey presents it.
Dorsey argues:
- Morality’s normative authority must be explained by reference to a further distinctive property.
- If explained by its distinctive content or grounds, morality’s normative authority cannot satisfy the project. (What was just shown.)
- Morality’s normative authority cannot be explained by a special connection to reactive attitudes or motivation (because explanation goes in the other direction).
- Morality’s normative authority is best explained either by the ground or content of morality.
- Because morality’s normative authority explains any connection moral obligations may have to the reactive attitudes or to motivation, and because morality’s normative authority is not prior to substantive moral investigation, a purported special connection to the reactive attitudes or the motivations of rational agents is also not prior to substantive first-order inquiry. (Dorsey, 771)
By showing that determining the content of morality requires first-order deliberation and all other options he’s aware of require understanding of the distinctive content of morality, Dorsey takes himself to have justified substantial skepticism about the project. Rather than determine what is distinctive of morality prior to first-order inquiry, Dorsey concludes with a sketch of how we might determine what is distinctive of morality by understanding it as determined by the correct moral theory.
III. Something that will be fruitful to open for discussion is whether it is surprising that content, as Dorsey understands it, must be determined after first-order inquiry. This can look unsurprising since he seems to think that the content of morality is a matter of what actual reasons, requirements, and dictates are correct according to morality, rather than a matter of the questions that morality addresses. Consider his positive account:
“Instead [of appealing to the project], we should say that whether a considered judgment has moral content or not should be determined by its capacity to survive whatever proper epistemic procedure is appropriate for first-order moral inquiry—such as reflective equilibrium with our substantive considered judgments, including our considered judgments concerning the distinctiveness of the moral domain. If, for instance, I judge that I am morally required to treat the Queen of England with traditional deference, whether or not this judgment has moral content should be determined by whether this judgment is or is not coherent with the remainder of our purportedly moral judgments in reflective equilibrium… if it is, then my judgment has moral content. If it is not (which I suspect), then it does not” (Dorsey, 772).
Here whether a judgment has moral content seems to turn on whether it is morally correct. But if that is the sense of content we are after, then surely it cannot be determined prior to first-order deliberation, since it will be a matter of contention. It seems unlikely, however, that this is what gate-keepers want to appeal to in ruling out other first-order theories, because, among other things, this would make their arguments circular.
We might worry that if Dorsey is right, things are much worse for moral theorizing than he has let on, because the only use of the project is not gate-keeping. As Dorsey notes, another reason theorists undertake the project is to ensure that they are deliberating and disagreeing about the same thing as those they are deliberating and disagreeing with. One might then worry that the project is necessary to ensure not only that moral philosophers do know what they’re theorizing about, but also that they can be confident that disagreements between moral philosophers are disagreements at all.
Dorsey himself comes close to making this point when he acknowledges that “…for instance, Mill and Gibbard may use the term ‘morality’ simply to pick out the norms relevant to praise and blame; Foot may use the term to refer to norms relevant to well-being and cognate concepts; Mackie may use it to pick out norms that are generally other-regarding and so on. This would entail, of course, that these philosophers are simply talking past each other, and hence there may be no substantive issue when it comes to the distinguishing mark of morality. However, this interpretation of the dialectic should be treated as a last resort” (Dorsey 754-5). The worry, of course, is that this so-called “last resort” is exactly what Warnock, on Dorsey’s reading, seems to be claiming is the result of giving up on the project. Without some handle on what is distinctive about the domain of morality, it’s not clear what could settle the matter of whether these philosophers are talking past each other.
It seems that we’ll need to have some sense of what the domain of morality is in order to ensure we’re not talking past our interlocutors. We might also want to use these fixed points as gate-keepers, as ways to rule out other first-order theories, but more fundamentally, we also want them to fix the grounds of the debate itself. Dorsey mainly focuses on the consequences that the failure of the project will have on first-order theorizing, and the inability of moral theorists to reject first-order theories on the grounds that they are incompatible with what is distinctive of morality, but if he is right, it is unclear that we can even be sure that theorists developing competing first-order moral views are even developing views of the same thing.
This seems to have repercussions not only for moral theorists doing first-order work, but also for the very project that Dorsey is engaged in. (Those familiar with Katia Vavova’s “Debunking Evolutionary Debunking” (Oxford Studies in Metaethics 9:76-101 (2014)) will note the similarity between the worry to follow and the debunking strategy she turns against the debunker there.) Dorsey early on states his main thesis as “Though morality is surely distinct from other domains, morality’s distinctiveness cannot be settled prior to substantive inquiry into the content of moral reasons, requirements, and concerns” (Dorsey, 748). One might worry, however, that if Warnock is right, Dorsey is not entitled to the first clause of this claim and many others like it. How can we know that morality is surely distinct from other domains unless we know that what is distinctive of morality is not distinctive of other domains? Dorsey assumes that we have a pre-theoretical understanding of distinct normative disciplines, but it seems particularly this sort of pre-theoretical understanding that his argument against the project brings under suspicion.
So, it seems we cannot assume prior to first-order deliberation that certain things are not part of the domain of morality. But it seems that throughout the argument against the project, Dorsey seems to assume exactly this. To give just one example, in his discussion of the failure of appeals to grounding to find what is distinctive of morality, Dorsey writes:
“According to Southwood’s approach, the moral domain is not grounded in social practices. But many different normative concerns are not so grounded: a concern for beauty, a concern for my own welfare, and so on. Southwood’s approach thus cannot distinguish morality from, e.g., prudence and aesthetics (to begin with)” (Dorsey, 762).
But this move, claiming that the chosen grounding property doesn’t uniquely ground morality, presupposes we know quite a lot about morality, at least, in the negative. But if we don’t know what distinguishes morality prior to first-order inquiry, why do we think we can know what distinguishes prudence and aesthetics? Why can we be confident, as Dorsey is, that prudence and aesthetics are those domains that have these concerns and that morality isn’t fundamentally concerned with beauty?
This worry also runs through Dorsey’s negative argument. It is central to the “coarse/fine tension” about the content of morality. This tension presupposes that there is content we want to rule out. But how are we entitled to this on Dorsey’s account? If we are not able to determine the content of morality prior to first-order deliberation, it seems that we might not be entitled to this sort of argumentative move. Of course, there are responses Dorsey could make here, but it is puzzling how we can be entitled to know what isn’t part of morality before moral inquiry and yet not be entitled to know what is part of the content of morality before this inquiry.
Dorsey’s positive suggestion, too, seems to presuppose some already running first-order moral commitments. Of course, specific judgments about moral requiredness can be assessed once we have a moral system up and running. But this is exactly the issue: how can we get the whole thing off the ground? How do we know whether considerations of etiquette and morality are really distinct or whether the love of beauty is a moral consideration? What makes negative facts about morality’s content free game while positive facts about it must be won with hard earned toil in first-order normative deliberation? If I begin with a moral view in which considerations of beauty are in my account, then it will likely not distinguish the domains of morality and aesthetics. Dorsey seems committed to distinguishing these, but it is not clear how his account will ensure that things work this way.
One final surprising result is that Dorsey’s argument seems to also undermine common strategies for arguing for moral nihilism. Not only do proponents of specific positive first-order moral theories deploy claims about what is distinctive of morality to argue against opponents, some theorists take accounts of what is distinctive of morality to show that nothing of the sort could exist. Moral nihilists and error theorists take there to be something distinctive of morality, and take this to be good reason to think that, because no one thing could have those distinctive traits, we should accept moral nihilism.
If Dorsey is right, there is no way that there could be such things that the moral nihilists could reasonably take to be distinctive of morality before they had a positive picture of what morality was, that is, prior to first-order deliberation. But it is just this kind of deliberation that the nihilist is committed to thinking is not going to yield sensible results, I’d have thought. In any case, it is possible that there is an argument against the very coherence of nihilism in this account, which I suppose could end up being good news for those currently caught in its grip.
Hi Dale,
Thanks for a very interesting paper. I have have several questions, but let me start with my main one.
Many people, including myself, argue as follows.
– Moral theory T (e.g., utilitarianism) has property P (e.g., the property of requiring agents to do what they lack decisive reason to do, all things considered). [Assumption]
– If T were the correct moral theory, it would not have property P. [Intuition]
– Therefore, T is not the correct moral theory. [From 1 and 2]
Your problem with such an argument seems to be that you take those making it to be treating 2 as having a “gatekeeping function” where our confidence in 2 is “quite high” such that we believe that, prior to any substantive inquiry into morality, we can determine which first-order theories are incorrect. Yet you allow that judgments such as 2 “can certainly be participants” in, say, a wide reflective equilibrium process by which we come to determine which is the correct moral theory. So your objection is just to giving such high confidence in a judgment like 2 that we could dismiss a first-order moral theory without evaluating how it fares in the wide reflective equilibrium process. Am I right?
But why think that people like myself, Darwall, Hooker, Mackie, etc. are treating 2 with such high confidence? All I think is that the above argument is a strike against utilitarianism. But I admit that we can’t just look at any one argument and conclude that a theory is correct or incorrect. We have to look at all the arguments. And some of those arguments will be about how utilitarianism does or doesn’t cohere well with our first-order moral judgments. Some of those arguments will, like the above, be about how utilitarianism does or doesn’t cohere well with our judgments about the nature of morality itself. Still others will be about how utilitarianism does or doesn’t cohere well with our judgments about epistemology, metaphysics, moral semantics, etc. Which is just to say that we must employ the method of wide reflective equilibrium. Yet you seem to suggest that merely by making an argument like the one above, we’re guilty of engaging in The Project. Yet your arguments seem to support only that those who make that argument while treating 2 as having a certain nonnegotiable or gatekeeping epistemic status are guilty of The Project. But I don’t think me, Darwall, Hooker, or Mackie would ever claim that because we’ve given an argument like the one above we can ignore all arguments about, say, how T coheres well with many of our first-order moral judgments. And if we don’t think that we can ignore such arguments, then how are we engaged in what you call The Project? I certainly think that all intuitions and/or judgments are negotiable.
The first thing I’d like to say is that I’m genuinely honored to have my work discussed here—so thank you to all the folks who were involved in the deliberations on that. I’d also like to give Kathryn Lindeman a very hearty thank you for an excellent precis which proceeds from a thorough and very generous read of my paper. I don’t have much to comment on when it comes to her recap because I think she captures my intentions pretty well. Her critical commentary has also helped me think through my own views about the upshots of this argument and even what I’ve taken myself to show. So I’m quite grateful.
I do have a few (minor, sketchy) things to say about her substantive critique of the paper in section III, but ultimately I found myself in genuine agreement with a lot of what she suggests are the upshots of my argument. It’s just that I think that (in lots of cases) that’s the position in which we find ourselves. (In particular, I thought her very interesting suggestion about nihilism was congenial to me.)
One point of contention, tho. Kathryn claims that one of the presumptions I’ve adopted throughout the paper—that morality is, in fact, distinct—is in tension with my ultimate conclusion, viz., that whether or not morality is distinct is to be settled only after substantive first-order inquiry into morality. Kathryn asks: “How can we know that morality is surely distinct from other domains unless we know that what is distinctive of morality is not distinctive of other domains?” This infects my arguments throughout, insofar as I try to show that some proposals for morality’s distinctiveness cannot capture its, well, distinctiveness (such as the “grounding” proposal, among others).
This is a very interesting point. There are two things I might say in response, depending on my mood (?). The first is more conservative, the second somewhat more radical. The conservative thing: I don’t argue that there is nothing one can know about morality or other domains prior to substantive inquiry. For all I have argued here there could be a number of prior-to-first-order-inquiry things we might be able to know about morality and a bunch of other domains. One such thing might just be that morality is distinctive in some way or other. It just so happens, or so I argue, that we should be skeptical that what distinguishes the moral domain in particular is similarly prior. (In the case of the grounding of morality, it could be that I agree (prior to the first-order) that morality is grounded in some particular way—it’s just that to know that this grounding is distinctive requires first-order inquiry, which is not compatible with the project. By the way, quick note: I’m a little bit sheepish about calling the grounding view “Southwood’s approach”, because I take him to be up to something a little different in his paper; his paper is a useful expression, though, of what such a grounding approach mind end up looking like.)
So this is what I take myself to have argued for, but ultimately (as sketched toward the end of the article) I’d like to plump for something a little more radical. I do have confidence that morality is distinct from, e.g., prudence, but ultimately (or so I think) that confidence must be vindicated by substantive first-order inquiry. Does this entail that my views are in tension with the argumentative moves I made? I hope not! The key is that I’m trying to argue against the project—but a test for any acceptable project-satisfying distinguisher must be that it passes the substantive presumptions that guide the project, including that morality is distinct; that there is genuine risk that some of our judgments are non-moral in content, and so on. But, or so I argue, the various properties that are identified as distinguishers don’t really succeed on the project’s own terms. I suppose someone might say: “well, look. I’ve given you a distinguishing property of morality. It turns out that this means that there is no distinction between morality, prudence, and aesthetics.” To complete the project, it had better be that we have sufficient confidence in this distinguishing property to accept that there is no such distinction prior to substantive inquiry. I suppose this is a position in logical space—but would we have confidence enough to accept such a claim without first-order investigation? My suspicion is no, but perhaps I’m wrong here. And that’s compatible with saying that to know for sure (rather than just having confidence) that there is a distinction also requires us to investigate.
Now Kathryn does (helpfully) press me on what this might mean for substantive moral investigation. She’s less sanguine than I am that we can keep on keepin’ on, as it were. But I think there are a couple of things we might be able to do even if I’m right that the distinctiveness of morality is an a posteriori matter. First, we might accept a number of potential organizing principles of the moral domain as terms that “fix the debate”, but accept them provisionally—as subject to revision if and when the first-order results work out. For instance, let’s say we accept two general organizing thoughts: immorality is typically blameworthy, and that the moral domain takes everyone’s interests seriously. Now, we might find that when we try to figure out what it means to take the interests of everyone seriously we end up with something like an impartial moral view (just play along). But then we recognize that that would be very demanding on folks. But our inquiry into the nature of blameworthiness might reveal that we’re reluctant to blame individuals for failing norms that are very demanding, and so on, and so we must then reconsider one of the primary organizing principles based on our considered judgments and how the results work out in the end.
I’ve gone on too long! Does any of that make sense?
Hi
interesting paper. I just had two questions:
1. This is a methodological question. The Project has something to do with explaining what is distinctive about moral standards. The way the paper formulates the different views on this issue is in terms of single unique necessary and sufficient conditions for something to count as a moral standard. But, as we all know, attempts to give such conditions have a bad track-record. So, for example, could it not be that the notion of moral standards is a cluster concept? The idea would be that all the accounts tell us something informative about the core features of moral standards that are usually found together. But, the idea would not be that all of these have to be found together always just as long as sufficiently many features are. In this way, there could be a sensible project of investigating and describing these features together even if individually none of the features work. Perhaps the features together also explain the normative authority of the moral standards.
2. The positive account seems a non-starter to me but I worry that I have misunderstood something. I’m thinking of this passage at the end:
“Instead, we should say that whether a considered judgment has moral content or not should be determined by its capacity to survive whatever proper epistemic procedure
is appropriate for first-order moral inquiry—such as reflective equilibrium with our substantive considered judgments, including our considered judgments concerning the distinctiveness of the moral domain. If, for instance, I judge that I am morally required to treat the Queen of
England with traditional deference, whether or not this judgment has moral content should be determined by whether this judgment is or is not coherent with the remainder of our purportedly moral judgments in reflective equilibrium if, in fact, this epistemological method is appropriate.
If it is, then my judgment has moral content. If it is not which I suspect, then it does not.”
But surely it is possible for myself and others to make judgments that have moral content and are false. In Hare’s missionary and cannibals case, I can well judge that the cannibals’ views about scalping have moral content even if I cannot bring them to reflective equilibrium given my moral views. And, I do not see why I could not make moral judgments myself that fail to cohere with the rest of my own judgments.
But, maybe I am missing something. Maybe the idea is that I should check a particular judgment only against my views of which judgments count as moral. This, however, doesn’t sound to me like moral inquiry given that my views about which judgments are moral can be those that fit what the defenders of the project say. It also fits well my first comment about moral standards and judgments being cluster/family-resemblance notions.
Hi Doug –
If you are thinking about your argument against utilitarianism in that way, then I really have no problem with it and the argument I offer in this paper doesn’t apply to you. (Incidentally, I took Hooker, Mackie, Foot, et. al. just to be offering views of morality’s special character, rather than making arguments of the sort you note.) But whether or not such an argument commits to the project is something that should be decided case-by-case, I should think. And what you suggest about your own argument would surprise me given some of the stuff you say in your book. For instance, you eschew discussion of the sense in which utilitarianism is inappropriately demanding from within the moral POV for discussion of the sense in which utilitarianism is unreasonably demanding. But one would have thought that if you were open to morality’s special force being revised in reflective equilibrium, the first sense would have been importantly relevant. Furthermore, you defend morality’s special force independently of substantive moral inquiry (by reference to a conceptual claim about morality). (By the way, nothing in your argument suggests that you need to be committed to morality being distinctively normatively significant, rather than just normatively significant, but I take it you would assent to the former claim.) Something similar is going on in Darwall and Hurley, at least as I read them.
But if I’ve misread you, I apologize!
Hi Jussi –
Re: cluster concept. I think this is possible, and indeed an attractive option. I’d resist the possibility that it could be vindicated prior to substantive inquiry, tho. (You might disagree here, if so I’d like to hear more.) I think the best way to make such a theory work (if I’m reading you right) would be similar to the last paragraph in my response to Kathryn.
Re: positive proposal. I think you rightly captured what was going on in that paragraph, but I think I worded it poorly. I guess my thought, generally speaking, is that there being a distinct line of investigation, i.e., does x judgment have moral content, is something we should give up instead for investigations like, i.e., is x true of morality? If my positive proposal is right, there really isn’t much distinction in the traditional sense between a judgment having moral content and a judgment not having moral content (in the sense of being admissible or inadmissible in genuine moral inquiry). Rather, the best sense we can make of it is whether this judgment survives moral reflective equilibrium. So of course you can make false claims about morality on my proposal, it’s just that there’s no distinction in false claims between those that are admissible and those that are inadmissible. The point of that paragraph was to indicate that even if there is no such distinction, this doesn’t mean our inquiry is somehow tainted or less genuinely moral.
Hey Dale!
Thanks for the engagement with my points.
I want to ask you to say a little more about the conservative/radical responses.
The conservative response is that you don’t mean to rule out all information about morality, just knowledge of what is distinctive of it. That’s what the project is committed to being prior to first-order theorizing. It’s fine to have commitments to the distinction between morality and etiquette there. But I worry because on some first-order moral theories (though none in vogue now) there are is no such distinction. Shouldn’t you be just as committed to the application of the “fine/coarse tension” here? Maybe not, and I’d be interested to hear why.
The radical view, however, was the one that I thought would get your argument into real trouble. But you’re right that the dialectic is really complicated here. If those invested in the project are committed to there being a distinction between morality and etiquette (say) prior to first-order theorizing, then it seems that their inability to distinguish them prior to distinguishing the content of morality is a problem. I worry, though, that your argument showing that they cannot do so turns on your own endorsement of claims your conclusion forces us to reject.
It might be that this isn’t the case, and that you are only showing that the endorsed positions of your opponents are untenable. It’s a very tricky distinction to make.
Consider these two arguments:
1a. On view X, squircles have four right angles.
2a. On view X, squircles are lines composed of points equidistant from a center point.
3a. Conclusion: View X is committed to a contradiction.
and
1b. Squircles have four right angles.
2b. Squircles are lines composed of points equidistant from a center point.
(Implied: We can’t know things about impossible things; Squircles are impossible.)
3b. Conclusion: We can’t know things about squircles.
I’m worried that your argument looks more like 1b-3b than 1a-3a. That is, I’m worried that your argument requires endorsing things that, on your conclusion, you should deny.
Of course, this is super complicated, so maybe I’m misunderstanding the dialectic. Or maybe some of this turns on what counts as content or what counts as distinctive in the sense you’re interested in?
Your last point is also interesting. I’m still worried that what you allow yourself to commit to in your first-order theorizing will determine what is distinctive of morality. But lots of first-order theorists aren’t going to agree on starting points (especially if you think that nihilism has a specific first-order moral theory!). So, does your view lead to a sort of content relativism, such that what turns out to be distinctive of morality depends on first-order commitments you start with? I’m having a hard time seeing how you could get people to agree on starting points, even if they are revisable, unless you have some guarantee that regardless of where you start, you’ll end up in the same place. Do you think that’s right? Does this bring us to a sort of constructivist relativism about the moral?
(Hurrah to philosophy blogging!)
Hi Dale,
I do “defend morality’s special force independently of substantive moral inquiry (by reference to a conceptual claim about morality).” But that, of course, doesn’t mean that I treat my claim about morality’s special force as being non-negotiable. Nor does it mean that even if I were to find via first-order moral inquiry that no first-order moral theory that had the special force that I take morality to have could do nearly as good of a job of systematizing our first-order moral intuitions as some theory (such as utilitarianism) that didn’t have the special force that I take morality to have, then I still wouldn’t be willing to revise (or reject) my view about morality’s having a special kind of force. As I say in my book’s last paragraph, my aim was not to convince readers that commonsense consequentialism is true. For to do that, I would have to compare commonsense consequentialism to all the various other moral theories on offer, which I haven’t done. Rather my aim was merely to show that commonsense consequentialism coheres well with not only our first-order moral intuitions but also our intuitions about morality and its relationship to rationality.
So it’s unclear to me what would make me or any of the other people you mention as an advocate of The Project. We don’t have to hold that claims such as the claim that moral requirements must be such that agents have decisive reason, all things considered, to abide by them are non-negotiable. But rather it is sufficient to have too high of a confidence in such claims. And I gather that it is your view that one has too high of a confidence in such a claim if no results from first-order moral theorizing would result in your changing your mind about those claims. But, of course, no one who endorses the method of wide-reflective equilibrium would be that confident. And I wonder if any philosopher is so confident in their claims about the nature of what distinguishes morality from non-morality is unwilling to revise these claims in light of enough evidence to the contrary from first-order moral inquiry. So I’m wondering what makes one an advocate of The Project. What makes one confidence in a claim about what distinguishes morality from non-morality too high? Now, of course, you admit that it’s a tricky question to say what is too high. But you must be able to say something about what is too high in order to be able to claim that in the case of me and the others our confidence is too high.
Doug and Dale,
Given what you’ve both said, could it be that the only people truly felled by The Project are moral nihilists?
Dale seems happier with the refutation of nihilism (in this way) than I am, but might that be a point where we could get traction?
Hi Kathryn,
Sorry for the late response (and also to Doug, whose comment I hope to get to soon); I’m a touch under the weather and have had to retreat to warmer/fuzzier environs.
My suspicion is that many partisans of the project are interested in avoiding judgments that are not particularly moral in character. This seems to me to indicate a presumption that there is at least some distinction between morality, proper morality, and other paradigmatic domains (such as prudence and etiquette). And so my argument (when I’m feeling radical) is attempts to show that the various distinguishing characteristics cannot succeed as such distinguishers.
Now let’s say I’m totally wrong here. Let’s say that many of the claims I make are not legit without first-order inquiry (such as morality and etiquette are distinctive). But I don’t really need the knowledge claim. The real question is whether any distinguishing mark can satisfy the project. Someone might offer, of course, a distinguisher that doesn’t render a distinction between morality and etiquette. But even if we don’t know that there is such a distinction, can we be as confident as we need to be in that distinguisher for it to satisfy the project, especially given that (without first-order investigation) it delivers the verdict that there is no distinction between morality and etiquette? I claim not!
About your last paragraph—I guess the answer is that I don’t really know. One possibility is that it may be that after our own investigations it turns out we really were talking about different domains after all. But I think that would be an interesting result, and one worth finding out. But I would be surprised if we couldn’t get at least some agreement on broad starting points: morality matters, immorality is typically blameworthy, morality cares about everyone’s interests not just the interests of a few, and so on and so forth. But then the real arguments would be how to understand each of these starting points, and which conflict with the others, and in the case of such conflict which one to reject. Different people will have different answers to this, and I’m not really sure what happens then. (Sometimes, after all, we just will hit intuitive rock-bottom; but I don’t THINK that entails content relativism.)
Hi Doug,
I feel like any attempt at Portmore exegesis that runs afoul of, you know, Portmore, is destined to be futile, so I’ll take you at your word about what you’re committing yourself to in Chapter 2. (Again, I don’t have any per se beef with arguments in the schema you outline; only if those arguments claim that premise (2) is the result of morality’s distinctiveness which I took yours to be but again maybe I’m wrong.)
You’re wondering what the level of confidence is that someone must have in a distinguishing property that’s compatible with the project: my suggestion (briefly sketched in the paper) is that it should be that it is at least plausible to say that this property (“p”) ought to play the role of gatekeeper—that it is plausible to say that any contrary-to-p moral x (judgment, etc.) should not be admissible in moral inquiry. Of course, this would leave it open that you treat the property as revisable once all the facts are in. But it’s still a pretty high standard, and one that the potential distinguishers, or so I argue, fail.
Of course, this is a high standard. And let’s say that I’m wrong as a sociological matter—no one really engages in the project. Nevertheless, that the project fails remains important for understanding proper moral inquiry. If the project were to fail, then arguments that rule out particular first-order claims (“r”) on the grounds that they are incompatible with morality’s distinguishing property (“p”) are illegitimate, insofar as to establish such a distinguishing property requires a prior substantive investigation of “r” versus “p”. But I think there are a lot of such arguments, even if the authors don’t take themselves to be committed to the project qua project.
Oh, and for reasons in the last paragraph above, Kathryn, I would resist your suggestion that the only one felled by (I think your mean) the failure of the project is nihilism, although I think that’s an intriguing result that I’m happy to trumpet if it’s right.
Hi Dale
thanks for the response. I’m not sure I can agree or make sense of this:
“If my positive proposal is right, there really isn’t much distinction in the traditional sense between a judgment having moral content and a judgment not having moral content (in the sense of being admissible or inadmissible in genuine moral inquiry). Rather, the best sense we can make of it is whether this judgment survives moral reflective equilibrium. So of course you can make false claims about morality on my proposal, it’s just that there’s no distinction in false claims between those that are admissible and those that are inadmissible. The point of that paragraph was to indicate that even if there is no such distinction, this doesn’t mean our inquiry is somehow tainted or less genuinely moral.”
I think of this in terms of radical translations. Imagine that we came across a large number of different foreign communities and ways of practical life – kind of like the missionaries came across the cannibals in Hare’s story. I take it that one interesting question we have is which one of these practices have a part of language that is moral – that would be translated to our moral language and which of these communities would have moral practices (so that we would ascribe to them moral judgments, moral emotions, moral standards and the like).
When we take part in the radical translation project, sure we have to rely on some standards of which beliefs have moral content. As missionaries, we have to have some way of deciding whether the cannibals have moral beliefs about scalping or not. So, I cannot understand how this distinction would not be legitimate and important – after all it is written into our translation practices. Also, the reflective equilibrium doesn’t seem to come to the picture at all – what moral judgments I take to be true just doesn’t seem to help me to decide whether the cannibals are making moral judgments or not. As Hare thought, this seems more like a matter of what role their judgments play in their lives.
Hi Dale,
Thanks for your thoughtful replies to my previous questions. At this point, I want to move on to some questions that I have concerning your view that morality’s distinctive normative authority requires explanation.
(Q1) You are concerned with people, like myself, who reject certain first-order moral theories (e.g., utilitarianism) on the basis of the claim that morality has a certain type of normative authority such that, if S is morally required to φ, then S must have decisive reason, all things considered, to φ. I argue that, on the basis of this claim and the fact that utilitarianism implies that S is morally required to φ even in instances where S doesn’t have decisive reason, all things considered, to φ, we can conclude that utilitarianism is incorrect. But note that this sort of argument doesn’t involve holding that this sort of normative authority is distinctive of morality. It could be that other normative domains have this sort of normative authority as well. So, as you see things, is there a problem with my sort of argument only if I’m claiming that this sort of normative authority is distinctive of morality? And why would the success of argument depend on not just whether normative authority premise is true but also on whether it is thought to be something distinctive about morality?
(Q2) You write: “if we accept NMA and MC, those who would distinguish morality and moral judgments by comparative normative significance must hold that in such competition, morality wins.” What does it mean precisely for morality to “win”? On my view, if non-moral reasons decisively oppose what the balance of moral reasons support my doing (say, φ-ing), then morality cannot require me to φ. I would have thought that this is an instance of morality’s losing, but you seem to be saying that it is an instance of morality’s winning. Could you explain, then, what you mean by “winning”.
(Q3) So I think that it is a conceptual truth about morality that, (CT) for all subjects S and acts φ, if S is morally required to φ, then S has decisive reason, all things considered, to φ. And you write: “if S is morally required to φ, then S must have decisive reason, all things considered, to φ.” And you write in response to such a view: “But this proposal seems hard to square with the truth of MC and NMA. After all, if we accept these theses, the comparative normative significance of morality is not something that could be ascertained simply by coming to a conceptual understanding of the nature of a moral requirement. Instead, the comparative normative significance of morality is a fact ascertained only by a first-order inquiry into the nature of normativity, of practical rationality, and so forth.” I’m not following. First, the conceptual truth makes no claim about the *comparative* normative significance of morality to other normative domains. Second, I’ve shown in my book that this putative conceptual truth about morality is compatible with the view that non-moral reasons often decisively oppose moral reasons. So although I agree that we must do “first-order inquiry into the nature of normativity, of practical rationality, and so forth” in order to determine whether and when non-moral reason decisively oppose moral reasons in determining when and whether we ever have decisive reason to act as the balance of moral reasons favors our acting, I don’t see how this means that we can’t establish CT via conceptual understanding of how the concepts “morally required to φ,” “has decisive reason, all things considered to φ,” and “would be blameworthy for φ-ing” are related.
Hi Jussi,
I think I see better now what your concern is, and it’s very helpful. I worry that my “positive view” answers a slightly different question than the one you’re asking. I guess when I’m using the notion of “having moral content”, I mean to indicate those sorts of judgments or other things that are admissible in first-order moral inquiry. (I was taking your question to be whether I should allow the judgment that more scalps are better into my substantive moral inquiry, and I think the answer here is yes, but that it would immediately be bounced by contrary considered judgment.) But I take that to be a different question than the question of what sorts of judgments we translate as “moral” in cases of radical translation. But I would still want to resist a blanket set of standards in the latter case that arises prior to substantive inquiry. We might, for instance, do our first-order investigation (or set our “starting points”) and find out that morality seems to have a connection to the interests of others, or a connection to normativity, or praise/blame, and so on. This would allow a broad “neighborhood”, as it were, for interpretation (even if we would regard the specific judgments in that case as false). But finding out where the neighborhoods are (if you’ll permit me that somewhat bizarre metaphor) would still be a product of reflective equilibrium on our parts. And any time we would translate some judgment of others as, e.g., not a moral judgment but instead a judgment of, e.g., etiquette or military norms or something, I would still want to take this as subject to revision given further reflective equilibrium. I don’t think that would ruin the prospects for radical translation.
Hi Doug –
These are great questions, and I suspected and hoped that you’d ask them! I have to run to teach, but I’ll try to respond shortly after.
Dear Dale
thanks – that clarifies things. One question: I agree that in the radical translation process we are aiming at a reflective equilibrium (aren’t we always trying to find coherent views?). The question though is what is the needed input data to this reflective equilibrium. If Hare is right, the input can be pretty sparse. All we need our translations to cohere with are our descriptive beliefs about the practical role of moral judgments in motivation and reactive attitudes. I take it that the defenders of the Project and the suggestions you criticise would be fine with this.
You seem to suggest that the input to this reflective equilibrium must be much broader to include even first-order substantive normative and moral judgments about what is right and wrong, good and bad and the like. Thus, before we have views on these issues, we cannot do the radical translation. The question is: why think this? Why do the first-order substantial normative and moral views matter given that which things the translated community takes to be “right” and “wrong” in their language does not seem to influence which communities we translated in our moral language?
So, I guess the question is: when we consider what distinguishes moral language, judgments and practices in the radical translation, why must be rely on a reflective equilibrium with our first-order substantial moral and normative views? What goes wrong with Hare’s argument?
Hi Doug,
I’m not sure I can do supreme justice to your (quite fair) questions, but here are my first stabs.
Q1. You’re right that in general I’m skeptical of arguments of that type. But here I’m really just concerned with arguments that focus on what’s distinctive of morality, which I take to be an independent project. (My new book and some other papers take on the broader argument scheme.) But you’re right that the arguments are very similar. Basically my argument here goes: to say that morality is distinctively normatively authoritative must be vindicated by claims about the distinctive content of morality, but the distinctive content of morality cannot be settled without first-order inquiry. And hence the distinctive normative authority can’t be settled without first-order inquiry, and hence to use authority-qua-distinctive to rule out certain first-order views prior to substantive inquiry is question-begging. The argument against the more generals schema would work like this: because any normative authority of morality must be explained by content-claims, and because part of what will determine the authority of morality is the truth of the content-claims at issue, to insist on the normative authority of morality in the first premise of such arguments is question-begging. (By the way, given MC I think most moral rationalists ought to be committed to a distinctiveness claim, just on pain of implausiblity, but that’s a different argument.)
Q2. Sometimes you can win by losing! I was focusing on specific instances of MC—competition between moral requirements and requirements of other domains (such as prudence). Here it must be that if moral rationalism is true, the moral requirement must maintain normative authority—“win”. Now it may be that, especially on your view, the moral reason often loses out in the determination of what morality requires, but the moral requirement will still maintain normative authority in the face of competing requirements. That’s all I meant by “winning”.
Q3. First, given MC and NMA, moral rationalism at least entails a comparative claim; that when moral requirements compete with the requirements of other authoritative non-moral domains, morality maintains rational authority in comparison to the authority of other domains. Second, again, MC doesn’t say anything about the conflict between moral reasons and other sorts of reasons, but instead about moral requirements and other sorts of (authoritative) requirements. You might deny this, of course, which would be interesting to find out—though I suspect you would not. By the way, I was thinking that arguments of the sort you mention do provide an explanation, so I didn’t take myself to be ruling that sort of thing out in the paragraph you mention.
Hi Jussi,
Thanks again. This is very good stuff, and it’s making me think hard about further implications of my picture. But here goes. You say: “Thus, before we have views on these issues, we cannot do the radical translation. The question is: why think this?” I guess my thought is that if we have NO VIEWS on the moral domain, its content, its force, etc., etc., then I am quite skeptical that any sort of translation would get off the ground. But I’m generally happy to allow, as I suggested to Kathryn, that moral inquiry can proceed by taking some very plausible starting points—morality is motivational, maybe, or morality is forceful, or concerned about interests, and so forth. Maybe a bunch (a cluster?) of them. This can allow us to engage in radical translation, but will also allow us to revisit not only the starting points but also our translations as we learn more about whether these starting points are actually principles of the moral domain. And I think this is plausible. Let’s say that we take a “morality cares about everyone’s interests” as a starting point. But after substantive investigation, we choose to reject, or to substantially qualify that claim. I would think that the right thing to do would then be to revisit whatever radical translations we’ve made taking that starting point as a given.
Is this responsive to your question?
Hi Dale,
Before I respond substantively, I want to be clear on my understanding of all your metaphors, which I’m having trouble with. Here are my attempted translations of your metaphorical phrases into non-metaphorical language. Please let me know if these translations are correct, and if not, what the correct translations into non-metaphorical language are.
(T1) S’s moral requirements are in “competition” with, say, S’s prudential requirements if and only if there is some φ and ψ such that S is morally required to φ, is prudentially required to ψ, but does not have the option of both φ-ing and ψ-ing.
(T2) S’s moral requirement to φ “wins” in the face of competing requirements (such as, a prudential requirement to ψ, where S’s both φ-ing and ψ-ing is not an option) if and only if S has decisive reason, all things considered, to φ.
(T3) The moral requirement that S φ “maintains rational authority in comparison to the authority of other domains” if and only if, despite there being requirements from these other domains to do things that preclude S’s φ-ing, S has decisive reason, all things considered, to φ.
Hi Doug,
Sorry for my imprecise language. Those sound OK to me if we take it as given that one has practical reason to conform both to the prudential and moral requirements.
Hi Dale,
So moral rationalism (MR) is the view that, as a matter of conceptual necessity, if S is morally required to φ, then S has decisive reason, all things considered, to φ. And you say that this “seems hard to square with the truth of MC and NMA.” But I’m still not getting why. So one can easily accept both MC and MR. For instance, one who accepts MR can also accept that we have nonderivative (and not necessarily decisive) reasons to conform to prudential requirements and thereby accept MC as well. And one can easily accept both NMA and MR: One could accept, for instance, that sometimes I’m morally required to φ even though this precludes my doing what I’m prudentially required to do — viz., ψ-ing. One can accept these two so long as one holds that the moral reasons that I have for φ-ing decisively oppose the prudential reasons that I have for ψ-ing such that I have decisive reason, all things considered, to φ. Moreover, it seems that one can accept MR, NMA, and MC together. So where’s the difficulty in squaring MR with the truth of NMA and MC?
Now, I do see that the comparative normative significance
of moral reasons over say non-moral reasons — where this means something like moral reasons always decisively oppose non-moral reasons — is not something that we could ascertain simply by coming to a conceptual understanding of the nature of a moral requirement. But I don’t see why we can’t ascertain via the conceptual connections among moral requirements, having sufficient reason (all things considered), and being blameworthy that an agent can only be morally required to do that which she has decisive reason (all things considered) to do given that one can, under suitable conditions, be blameworthy for doing something morally wrong but cannot be blameworthy for doing what one had sufficient reason, all things considered, to do.
(Yes, Dale, sorry I did mean felled by the *failure* of The Project!)
Hi Doug –
Second paragraph first. I agree with you there—I didn’t mean anything in the paper to rule that possibility out. (I think I must have messed up in making myself clear; if so I apologize.) I took the connection between moral requirements and blameworthiness to be an explanatory distinguishing feature. To say that the connection is conceptual isn’t to say that it isn’t explanatory. Though I think that argument isn’t successful, I think it’s a pretty damn good argument and I don’t think it fails because of anything to do with MC or NMA.
First paragraph. (Surely you don’t mean that moral rationalism is the view that as a matter of CONCEPTUAL NECESSITY, if S is morally required to phi then S has decisive reason to phi. That may be one way to be a moral rationalist, but it isn’t the only way). Why would that sort of a view run afoul of NMA and MC? I think the problem lies here: “One can accept these two so long as one holds that the moral reasons that I have for φ-ing decisively oppose the prudential reasons that I have for ψ-ing such that I have decisive reason, all things considered, to φ.” But if what we’re referring to is the narrow sense of morality, I don’t see how one could come to that conclusion simply given a conceptual understanding strictly of moral requirements themselves. You actually have to do some work to figure out whether that’s a true claim about normativity. One COULD come to this understanding if one said, for instance, that moral requirements just are whatever we have all things considered decisive reason to do, but this would be the broad rather than narrow sense of morality that isn’t under discussion here. When we’re talking about the narrow sense of morality, the notion of a decisive or practical reason is itself over and above merely the notion of a moral reason or a moral requirement.
But as I say, I take your argument to be perfectly compatible with everything I want to say in that paragraph, given that you not only rely on a conceptual analysis of a moral requirement, but also on substantive claims about the limits of blameworthiness and the connection between that concept and practical rationality.
Hi Doug (one more time) –
I have some doubts about what I just wrote, and I feel like I need to be a little more clear. I have to do take care of some child care things right now, but I want to say more in a little while.
Ok. Back again. Just to sum up my response, I think the key is that in the paragraph you cite, I’m dealing with views according to which MR can be established SIMPLY by a conceptual understanding of moral requirements themselves. It’s possible for MR to be conceptually true without MR being true simply as a result of the concept of a moral requirement, in particular. And there, for the reasons I cited two comments back, MC and NMA are in tension with such a view. But I recognize the need to provide more substantive argument against views according to which other concepts are in play—like, e.g., blameworthiness, motivation, and so forth. I hope to have provided such, but YMMV.
Hi Dale,
You write: “I think the problem lies here: ‘One can accept these two so long as one holds that the moral reasons that I have for φ-ing decisively oppose the prudential reasons that I have for ψ-ing such that I have decisive reason, all things considered, to φ.’ But if what we’re referring to is the narrow sense of morality, I don’t see how one could come to that conclusion simply given a conceptual understanding strictly of moral requirements themselves.”
Okay, I think that I agree that we cannot, simply given a conceptual understanding of moral requirements themselves, know such a conclusion. But once we realize this, I think that MC become contestable. To see why, consider the view that prudential requirements are normatively supreme (call this prudential supremacy or ‘PS’) such that S has decisive reason, all things considered, to φ if and only if S is prudentially required to φ. If PS and MR are both true, then moral requirements will be a proper subset of prudential requirements. And thus moral requirements and prudential requirements will never conflict, calling MC into question.
So you claim that the idea that the proposition ‘if S is morally required to φ, then S has decisive reason (all things considered) to φ’ just “is a conceptual truth of the nature of morality or moral requirement…[is] hard to square with the truth of MC and NMA.” On the one hand, I think that it’s certainly possible that they are all compatible. But, on the other hand, that will depend on just how powerful the reasons from other authoritative domains (such as prudence) are. Because if they are very powerful, then we will be unable to square MR with the “truth” of MC, but only because MC isn’t true.
One more thing: So I agree that, simply given a conceptual understanding strictly of moral requirements themselves, we cannot conclude that the moral reasons that I have for φ-ing decisively oppose the prudential reasons that I have for ψ-ing such that I have decisive reason, all things considered, to φ. But that’s not what I was doing. I was concluding that from BOTH a conceptual understanding of moral requirements and MC.
Hi Doug,
Thanks for that. I may very well be misreading you, but I’m not sure I want to disagree. Ultimately, I think MC is plausible, and I was resting the argument there on its first-order plausibility. But you may be right that once we get down to the nitty-gritty, we find (perhaps to our surprise) that both PS and MR are true, and hence we need to revisit MC. I certainly don’t want to rule out that possibility, but that would be an a posteriori exercise, I should think.
I think I should maybe try to say more about how I take the argument to go—I really wasn’t thinking I had established anything super-significant here. An important part that I haven’t yet stressed is that I’m really referring to the narrow sense of morality, the sense that—unlike the broad sense—isn’t an “all inclusive” guide to conduct. In other words, the sense of morality that doesn’t just deliver verdicts about how we are to act all things considered. So in this sense I’m cooking the books before I even begin. But one COULD say that you still don’t need any further explanation of MR even given just a basic conceptual understanding of narrow morality if either MC or NMA were false. If MC were false, MR would be trivially true given that no authoritative requirements would tell us to do something other than what morality tells us to do. If NMA were false, again, MR would be trivially true given that morality’s requirements are the only ones one has practical reason to conform to. So it’s important that these are both true. But I think they’re both true. And so a pure conceptual understanding of the narrow sense of morality doesn’t deliver the goods—we need some further explanation.
Okay, thanks Dale. That’s helpful. I’m curious, though, what you mean by “trivially true.” I gather that you don’t mean “uninformative.” For even if MC were false, learning that MR is true would not be uninformative — especially, if one didn’t know that MC is false.
Hi Doug,
Yeah, I don’t mean “uninformative”. I think that sentence might better be rephrased like this: “If MC were false, MR would be a straightforward consequence given that no authoritative requirements would tell us to do something other than what morality tells us to do.”
Hi all,
It looks like the discussion may have come to something of a close. Anyway, I’d just like to thank Daniel Star, the rest of the PEA Soup team, Kathryn, Doug, and Jussi for a very helpful and thought-provoking discussion.