We are pleased to present the third installment of PEA Soup's collaboration with Ethics, in which we host a discussion of one article from an issue of the journal. The article selected from Volume 120, Issue 3 is Mark van Roojen's "Moral Rationalism and Rational Amoralism" (open access copy here). We are very grateful that Russ Shafer-Landau has agreed to provide the critical precis of Mark's article, and his commentary begins below the fold.
I want to thank Dave Shoemaker from PEA Soup for coaxing me into making my blogosphere debut, and Mark van Roojen for writing such an interesting paper.
I’m very sympathetic with Mark’s line of argument, and have been encouraged to keep things short. So I will content myself with raising just three questions that might spark discussion, rather than trying to pinpoint lines of thought that strike me as mistaken.
First question: what is at stake in a defense of the two internalisms of the sort that Mark endorses? Take judgment internalism. In its classic, unsoftened form, judgment internalism asserted a necessary connection between sincere moral judgment and motivation: if an agent judges an available action to be morally right, then she is motivated to some extent to do it. As is well known, we can couple this internalist thesis with a Humean thesis about the motivational impotence of representational states to yield a noncognitivist conclusion. That’s interesting, and powerful.
But once we qualify judgment internalism, so that there is only normally a connection between sincere moral judgment and motivation among rational agents, I’m unsure of what role it can play in important metaethical arguments. Of course there is some intrinsic interest in whether the modified internalist thesis is true. But it would be good to hear about whether folks think that the addition of a plausible premise or two will yield a surprising and controversial metaethical conclusion.
I have the same sort of question about existence internalism. In its classic form, every moral obligation entails a decisive reason of compliance, so that those who act immorally thereby act irrationally. That’s a very cool thesis that lots of people hope is true, and lots of people are very skeptical about. (Sometimes these are the same people.) But once we enter Mark’s qualifications, by means of noticing the variety of forms of subjective irrationality, then it’s no longer clear who would reject existence internalism–or the rationalism that can accommodate it. After Mark’s treatment of these theses, we are left with the view that it can be perfectly subjectively rational to ignore or otherwise violate objective moral requirements (assuming there are any). But did anyone ever deny this?
Though I generally hate footnotes, I would have liked some citations to those whose interesting views entail the rejection of the internalisms that Mark ends up with. That probably would have given me a better sense of what is at stake here.
Second question: how plausible is the rationalism that Mark endorses? As I understand him, one commitment of his rationalism is the view that if an agent is morally required to do something, then it is (at least) objectively rational for the agent to do it. And this means that if the agent were a perfect reasoner (where perfection is characterized wholly without reference to moral features), then doing her duty would make sense to her.
The standard worry about rationalism is given by the traditional Humean: what is rational for me to do depends on what I want; people want different things, even after exposure to super-duper cognitive psychotherapy; therefore what is rational for my ideal counterpart to do may differ from what it is rational for your ideal counterpart to do in the same situation. If this familiar picture is correct in its essentials, then rationalism is true only if moral relativism is true.
Now I’m pretty sure that Mark isn’t a moral relativist. So he’s either got to reject rationalism, or one of the two Humean assumptions that sparked this worry. I assume he’d do the latter. But I don’t see that anything he says in the paper gives us a basis for rejecting either assumption.
Mark clearly thinks that his two duly qualified internalist theses help to support rationalism. Do they do anything to cast doubt on the familiar Humean picture just sketched? If not, then even if they do provide some support for a softened rationalism, they don’t go to what seems to me to be the heart of anti-rationalist concerns.
A final question: how are we to understand rationalism? Mark sometimes characterizes it as the view that the property of rightness is identical to the property of being rational. I was unsure about whether the rightness in question is meant to be moral rightness. Suppose it isn’t. If we aren’t talking of moral rightness, then it’s again not clear what is at stake here. Both internalist theses are theses about morality–either sincere moral judgments, or moral obligations. I don’t see how they support, or are supported by, a rationalist thesis that isn’t focused on moral rightness.
But if the rationalist thesis is about moral rightness, then it is false, since the property of being morally right is not identical to the property of making sense. Many beliefs, intentions and actions make sense, though they have nothing whatever to do with morality. Of course we might amend the thesis so that it is, or entails, the claim that what is morally right = what makes moral sense. There’s no denying that. But then there’s also no interest in the thesis, and it is hard to see who might be tempted to resist it. So on this reading, the thesis turns out to be either false, or of negligible importance.
I’ve left out most of the interesting stuff in Mark’s paper–-to read my comments, one would never know about his good work on Frege’s puzzle, or the varieties of subjective rationality, or the level of nuance present in his defense of the two internalist theses. I hope that other readers will address those aspects (and others) in their comments, and perhaps also offer replies that reveal my worries to be less serious than they might initially appear.
Russ Shafer-Landau
University of Wisconsin, Madison
shaferlandau@wisc.edu
I want to hear more about how Mark avoids the worry, “that [his view] undermines one of the main motivations for rationalism—that rationalism is well placed to secure the rational authority of morality and to explain why we rightly criticize those who act immorally.” (516)
If I accept Mark’s view, why not conclude that immoral but subjectively rational agents cannot be rightly criticized and are not subject to morality’s objective rational requirements?
Mark does say the following to help assuage the worry, as it arises in a specific context in the paper:
“An agent is only rarely in a position that requires choosing between doing what makes sense to do in light of her evidence, time, and so on and doing what she thinks would make sense to do if she had full information. In normal cases these questions collapse into one another from the first-person agential perspective; the agent is trying to do what she thinks is objectively right by doing what her subjective situation suggests it makes sense to do given that goal and her evidence, time, and so on.” (516)
and
“In an overwhelming majority of situations subjectively rational agents will and should intend to do what is objectively right. And this gives objective rightness its rational authority and grounds rational criticism of agents who don’t meet this requirement.” (516, ft. 25)
(1) Let’s grant that there are only a few cases in which agents believe that for them subjective rationality and objective rationality come apart, and in which the agents sensibly and “de dicto”, so to speak, intend to do what is subjectively rational and not objectively rational. Call these self-consciously sub-optimal agents “sub-ops”. I agree that in the overwhelming majority of situations agents are not sub-ops, and will grant the first-person perspective point that normally people do not have the split “de dicto” intentions characteristic of sub-ops.
(2) But notice that most of the defectors from morality that we worry about are not sub-ops. They may, “de re” intend the merely subjectively rational action and not the objectively rational one, but they do not intend that “de dicto”. On the contrary, these agents “de dicto” intend to act objectively rationally – it is just that (if rationalism is true) they have false beliefs about what it is objectively rational to do. Given that they inhabit a normal first person perspective, we can assume that many Freshman Egoists, Huck Finnish agents, Nietzscheans, Susan Wolfish agents, etc, (a) intend to act objectively rationally and (b) believe that some moral actions are not objectively rational. And it might be that many of these people’s false beliefs about objective rationality are themselves subjectively rational, for the sorts of reasons that Mark insightfully helps us to identify.
In these cases, I do not see how Mark can claim that the actual objectively rational thing to do has authority over these agents or that we may criticize these agents for violating the actual objective rational requirements. After all, they aim to act objectively rationally, have subjectively rational beliefs about what it is objectively rational to do, and act on those beliefs. The mere fact that their beliefs about what it makes most sense to do are false does not support the claims about authority and criticism. So I am left wondering how Mark can support the traditional rationalist projects.
I have a quick and somewhat vague answer to one of Russ’s questions.
The simple answer is: a metaethical theory should explain why there is this ‘normal’ connection. Take some other type of judgment — say, judgment about negative electrical charge. (For a realist, I take it, the judgment should be typed by its subject matter.) There is no necessary connection between this type of judgment and motivation, even if we restrict our scope to what happens normally and to rational agents. Why does the connection obtain when we move from negative electrical charge to moral wrongness?
I don’t say there’s no good answer to this question for a realist, only that it’s something that has to be explained. For that matter, irrealists have to pay attention, too — they have to be careful not to be explaining too much or too little.
I also had a similar worry as Russ’s last one. I had slightly different worries on both horns of the dilemma. So, assume that ‘all things considered rightness’ is identical with objective rationality. This would mean that also non-moral considerations are relevant to what what is objectively rational/right. In this case, the property identity would not explain why moral requirements are requirements of reason. It would if the objective rationality were identical with moral rightness. But, then taking a worse bet out of two would be morally wrong.
Secondly, I actually think that in certain rare cases moral reasons are outweighed by reasons of special relationships. A parent has most reason to cheat her child to a life-saving surgery even if this cannot be justified to other parents whose child dies as a result. So, I might think that in this case it would be irrational for the parent to do the right thing (it would not make sense). Now, on Mark’s view, I think I’m in this case subjectively rational but not objectively rational. But, I really would love to hear what piece of rational reasoning I’m missing out on that shows that in this case too moral reasons outweigh the rest.
I guess I would have also liked to have heard more about the semantics in virtue of which ‘right’ and ‘objectively rational’ came to co-refer. Two semantic frameworks are discussed. On Fregean view, the terms have the same reference but different senses. On the Millian view, the terms have the same meaning because they co-refer. I have general worries about the latter framework (for instance, what does it take to be able to use a term if knowing the meaning of the term cannot be required). But, anyway, in the example used (Evening Star/Morning Star) there is a story of how these terms came to co-refer – we were in a causal relation to one object whilst coming up with the names. Yet, properties like being right and being objectively rational don’t seem to get their reference in the same way. So how did they end up co-referring if not in that way?
There is a suggestion that this is because the actions and practices of most normal speakers. They treat rightness as sufficient for rationalising and justifying an action – this makes the terms co-refer. This seems like a controversial empirical claim. Also, wouldn’t property-identity make rightness also necessary for rationalising an action? But, do the actions and practices of most normal speakers support that necessity? Don’t many speakers accept that the mob hitman acts rationally even if he does wrong? Finally, if the reference of ‘right’ gets fixed by the actions of most ordinary speakers don’t we get the Moral Twin-Earth problem? Assume a community in which people don’t think that rightness is sufficient for rationality (say that their rationality is that of prudence). Would it really be the case that we couldn’t disagree with them about what is right given that we talk about different properties?
Finally, I wish Mark hadn’t use ‘existence internalism’ for the claim that having a moral obligation is necessarily a reason to do it. This term seems to have a more popular use when we think about whether the existence of reasons/obligations depends on our motivations. The claim Mark discusses is often just called Moral Rationalism (the term Mark uses for the claim that requirements of morality are requirements of reason – which seems like the same claim as his existence internalism). In any case, the terminology is confusing with these terms generally.
I’m not sure if it is good form to jump in so soon, but there are already lots of comments to respond to, so good form or not, I’ll start.
First, though, I should thank Russ for his nicely focused and helpful comments, and also the folks behind the curtain here at Pea Soup and also at Ethics for putting this together. It is hard to express just how pleased I am to get the chance to do this.
I’ll probably wind up dividing my comments into different postings, but here is a first shot at one of the issues raised.
In his first comment Russ wonders about the interest of the two sorts of internalism I focus on once they are qualified to get moderate internalism:
But once we qualify judgment internalism, so that there is only normally a connection between sincere moral judgment and motivation among rational agents, I’m unsure of what role it can play in important metaethical arguments. (Quoting just what Jamie also quoted above)
I think that Jamie’s brief comment roughly captures how I think about this. First off does seem to be some connection between moral judgements and motivation that is special to such judgements. And, the various arguments that I steal from Jamie, Hare, and Horgan and Timmons, seem to me to show that we think this connection is necessary. That is, we don’t feel justified in translating a term with a moral term of ours unless some such connection is there to underwrite the translation. We want to capture exactly what we think is required as best we can, even if it turns out that connection is not as robust as some of its partisans and detractors had thought it was.
Secondly, even if no specific family of metaethical theories needs to deny the connection once it is suitably modified, the connection may favor some positions over others. When it comes to general kinds of philosophical theories we’re not so much in the business of refuting one or another genus by confronting it with phenomena it has to deny as we are in the business of showing how one theory does a particularly good job of explaining this or that phenomenon for which we’d like an explanation. So in this paper I claim that rationalism can explain the true version judgement internalism – moderate judgement internalism. It has that going for it. Some other theories will also have explanations on offer, whereas certain others may not. This doesn’t show that those others are out of the running, but I take it that it is a reason to prefer one theory over another that it offers the best explanation of a certain range of relevant and true phenomena. Since I like the explanation I give, I’m going to think that rationalism of this sort does pretty well by this test.
It looks to me like Brad and Jussi’s comments are more closely connected with Russ’s other points, so I’ll leave this bit as a single comment and follow up on other points in comments below.
I think I can quickly address Russ’s second question before I have to run out temporarily. Russ is right that I’m not a Humean about reasons, so I don’t think I need to be a relativist about morality in order to defend rationalism. I suppose though, that it is worth saying something about how I’m not a Humean – I think there are reasons beyond those Humeans think there are. Some of the reasons we have do not require any antecedent motivational ground of the sort Humeans think there are for all reasons. But I don’t deny that there are reasons of the sort that Humeans think there are. This means that there can be room for some kinds of relativity to an agent’s psychology in what makes the most overall sense to do. And, given my (highly controversial and as yet undefended) tendency to think the reduction of morality to rationality is a straightforward identity this will let some relativity into morality as well.
Russ writes:
Mark clearly thinks that his two duly qualified internalist theses help to support rationalism. Do they do anything to cast doubt on the familiar Humean picture just sketched? If not, then even if they do provide some support for a softened rationalism, they don’t go to what seems to me to be the heart of anti-rationalist concerns.
I didn’t think of myself as arguing against relativism in this paper. Maybe I should have. But I tend to think of relativism as a kind of minimal rationalism. Someone like Gil Harman thinks roughly that an action is morally right for an agent only if she either already accepts a requirement to do it or fails to accept it only due to a failure of rationality or information. Since on his view rationality is hypothetical, he thinks that people with different psychologies can have different basic moral demands that apply to them. So when I think of my debate with this kind of relativist, I don’t tend to see them as anti-rationalist opponents. I tend to see the disagreement as being about what rationality requires.
I need to think more about whether the kind of overall story I’m telling here (about terms getting their referents from a connection with other speakers who use the term to refer to the same thing) would be more difficult for a relativist view of this sort.
I’ll address Russ’s third family of comments and their overlap with Brad and Jussi’s concerns after an errand and some more thought. One thing I will need to explain is why I tend toward the unpopular and counter-intuitive view that ‘morally right’ and ‘rationally right’ refer to one and the same property. Everyone so far is rightly worried about that. And in fact so am I.
Russ’s third point is about how I understand rationalism, is it an identity thesis or is it something else? And he objects that if it is an identity thesis it is false,
since the property of being morally right is not identical to the property of making sense. Many beliefs, intentions and actions make sense, though they have nothing whatever to do with morality.
Jussi has a similar worry:
This would mean that also non-moral considerations are relevant to what what is objectively rational/right. In this case, the property identity would not explain why moral requirements are requirements of reason. It would if the objective rationality were identical with moral rightness. But, then taking a worse bet out of two would be morally wrong.
And he expands:
I actually think that in certain rare cases moral reasons are outweighed by reasons of special relationships. A parent has most reason to cheat her child to a life-saving surgery even if this cannot be justified to other parents whose child dies as a result. So, I might think that in this case it would be irrational for the parent to do the right thing (it would not make sense). Now, on Mark’s view, I think I’m in this case subjectively rational but not objectively rational. But, I really would love to hear what piece of rational reasoning I’m missing out on that shows that in this case too moral reasons outweigh the rest.
The paper is neutral between two different identity theses, one of which is the one that raises the worry and the other which I think avoids the complaints. The first version just says moral rightness is identical to rational rightness. And it is subject to the worries Russ and Jussi highlight because it means that every reason is a moral reason. Thus, as Jussi points out, the view is committed to holding that when it is rational to cheat to save one’s own child, it is also morally right to save one’s child. The second version identifies moral rightness with the a restricted range of grounds counting in favor of an action’s being rationally right. I think this version doesn’t cause the problems Russ and Jussi point out insofar as it allows there to be rationalizing features of an action that aren’t moral grounds and which even oppose moral grounds.
I think that the view which restricts moral reasons to a subset of rationalizing reasons is very attractive in many ways. Most people seem to make a distinction between moral reasons and other sorts of reasons in at least some contexts, and we understand what they mean when they do so. Furthermore, it would make the task of explaining Frege puzzles more tractable.
Unfortunately, I can’t see any way to draw a single moral/non-moral distinction that comports with the way people talk when they make such a distinction in ordinary speech. And I don’t think there is any natural (as opposed to gruesome) kind of ground for rationalizing action which could be divided off as moral as opposed to non-moral. Often when people distinguish moral reasons from other reasons they mean to distinguish non-self interested reasons from self-interested reasons. At other times they mean to single out altruistic reasons from other reasons. People sometimes include religious prohibitions of arbitrary shape as moral reasons, even though they have nothing to do with the interests of anyone. And eudaimonists famously think moral reasons to be partly grounded in self-interest. This raises the question of how we so easily know what people meant to exclude or include when they make such a distinction. My answer to this is that in a particular conversation it is often very easy to see what a speaker intends to get across when she makes such a distinction. If you say to me, “That’s not a moral reason!” when I cite my interest in support of a course of action, I know roughly how you mean to draw the line in this conversation. So we can communicate despite there being no real underlying difference between moral reasons and other reasons that people are all tracking.
There still might be a line that people don’t all recognize, if for instance there was a natural kind of grounds that explained why people were apt to make a distinction between moral and other reasons. That people don’t track the extension of a term perfectly need not show that there isn’t a fact of the matter about the referent of the term. But when I try to come up with a way of picking a subset of grounds for acting that both constitutes something like a natural subset of the grounds for action generally and which could be part of the explanation of why people make such a distinction, I cannot come up with one. Many grounds which seem at first like they are irrelevant to the requirements of morality seem to me to be obviously morally relevant when we are assessing the stringency of certain moral requirements. That a certain course of action would require too much by way of personal sacrifice seems to me a reason to think that some moral requirements are less stringent than some think. But if that is true we have a case where certain sorts of self-interested reasons change the shape of moral requirements. And I think that makes them moral reasons – reasons relevant to determining what is morally right to do.
Given what I just said, it is pretty obvious how I’d describe Jussi’s example. If it really is true that I have sufficiently weighty overall reason to make cheating to save my child rationally required, then it is also morally right that I do so.
I don’t expect people to find my conclusion here very plausible. So I should highlight that the problematic commitment is an optional component of my opinion about these matters. It isn’t required by any of what I say in the paper. And I’d be (mostly) happy if someone could give me a way of drawing a line between moral reasons and other sorts of reasons that seemed worth drawing and seemed to explain why people are apt to make such a distinction. The rest of the view that I defend in the paper doesn’t depend on it and would in some ways be easier to defend as I note in one footnote to the paper.
I still need to say quite a bit more about the comments so far. But I need to take a break, at least for much of this evening.
Brad, your comments about the authority of objective deserve a long response. You summarizes the overall worry as:
If I accept Mark’s view, why not conclude that immoral but subjectively rational agents cannot be rightly criticized and are not subject to morality’s objective rational requirements?
And then you focus on a particular sort of subjectively rational agent who doesn’t in fact do what is objectively right:
. . . it is just that (if rationalism is true) they have false beliefs about what it is objectively rational to do. . . . [they] (a) intend to act objectively rationally and (b) believe that some moral actions are not objectively rational. And it might be that many of these people’s false beliefs about objective rationality are themselves subjectively rational.
Given that setup, the specific worry then is:
I do not see how Mark can claim that the actual objectively rational thing to do has authority over these agents or that we may criticize these agents for violating the actual objective rational requirements. After all, they aim to act objectively rationally, have subjectively rational beliefs about what it is objectively rational to do, and act on those beliefs. The mere fact that their beliefs about what it makes most sense to do are false does not support the claims about authority and criticism.
As I get the force of the objection, it isn’t really an objection to moral rationalism itself. Whether rationalism is true or not there can be people who aim to do what it is objectively rational to do and who have false beliefs about what that is. And these beliefs themselves can be subjectively rational (which here means rational given everything they are in a position to know). So as I see the objection it is an objection to the use of this objective/subjective distinction while maintaining the thought that objective rationality is authoritative. And the point is that once we realize that it can be subjectively rational to believe this about what objective rationality demands people who don’t do what is actually objectively rational for this reason are beyond criticism.
My response is that these people are trying to do the objectively rational thing by their lights. And they are failing. Criticism comes in lots of flavors. Given that they faultlessly believe the wrong things about objective rationality, I’d agree we shouldn’t blame them. (Remember that we’ve stipulated that their false beliefs are themselves rational.) But we surely can point out that they’ve got something wrong. And we can point out that they don’t succeed in doing what they are trying to do. I’d call that a criticism. Furthermore, if I point this out and perspicuously share evidence of their mistake with them I have changed their epistemic situation. They are not now rational if they persist in their false beliefs. Now I might well criticize them in a way that includes holding them responsible for their errors in ways that mere ignorance does not normally support. This is all I think we should want for the authority of objective rationality.
This all sounds OK to me when we are talking about rationality. And it still sounds OK to me when as a rationalist I add that each of these objective and subjective kinds of rational status are at the same time a similar kind of moral status because moral requirement, permissibility and forbiddenness are identical with rational requirement, permissibility and forbiddenness. So I could say the same thing substituting the corresponding moral terms and what I say would remain true. And, just as the authority discussed in the previous paragraph is all the authority I think it reasonable to want for objective rationality, I think the same thing about objective moral rightness. If an action is one that I ought morally to choose given all the facts and I try to do what is morally required given all the facts, and if I err through no error of reasoning but only due to epistemic factors beyond my control, I can be criticized for doing the wrong thing. It can be pointed out that I didn’t achieve my aim. But holding me blameworthy in any more stringent way seems morally inappropriate.
I should note that there are cases where subjective rationality tells us not to give full authority to objective rationality. In other words, these are cases in which an agent making the best of imperfect information will not be trying to do the objectively right thing by doing the subjectively right thing. These sorts of cases are those such as the mine shaft cases and Jackson’s disease examples (which Janice Dowell discussed in her post last week). In these cases, because the cost of going wrong in one sort of way is really high, we should do what reduces the likelihood of that sort of error, even when reducing that likelihood rules out getting the optimum outcome which one would aim at if one had full information. I think that this shows that you don’t want to overdo the authority of objective rationality/objective rightness. Sometimes, given your evidence you must do something you know would not be the thing to do if you had full information.
What a wonderfully subtle and discriminating paper! While (for now) I’m convinced of your thesis–that moral rationalists can allow for rational amoralism–here’s a few small qualms I have.
(1) As Jussi points out, your case for moral rationalism depends on a controversial empirical claim–one which I don’t personally find very plausible. What you’re proposing to me, if I understand, is that I and others who reject moral rationalism don’t know what moral rightness really is, or what normal people mean by “morally right”. It’s not that we merely have a faulty metaethical theory about our own practice, but that we’re like those who are ignorant that Hesperus and Phosphorus are the same “star”. To take the Fregean option, our mode of representation of moral rightness is something like “whatever it is that those normal people are referring to when they use the words ‘morally right'”. I find that incredible.
(2) Small quibble: on p. 513, you suggest that it can’t be rational to act in a way that you think is irrational. But what if someone thinks that something terrible will happen unless he manages to do something that is irrational, and so does A, because he thinks that doing A is irrational? His belief might be mistaken, but couldn’t this still qualify as a case of rationally doing something he thinks is irrational? Or is he necessarily irrational if he acts in this way?
(3) On p. 514, you write that it is only irrational to act in a way that is irrational given your actual situation. But I wonder: does any agent ever perform an action that made no sense to her at the time she performed it? I don’t think I’ve ever experienced anything like that. (That’s not to deny that we might have trouble after the fact identifying the sense we saw in an action). If your answer is “no”, does that mean that irrationality is impossible?
Mark,
This:
“The second version identifies moral rightness with the a restricted range of grounds counting in favor of an action’s being rationally right. I think this version doesn’t cause the problems Russ and Jussi point out insofar as it allows there to be rationalizing features of an action that aren’t moral grounds and which even oppose moral grounds.”
avoids some of the problems me and Russ pushed but not all.
Rationalism, as traditionally conceived, is the thesis that acting morally is rationally required *overall, all things considered, on the resultant level when all reasons have been taken into consideration*. Now, your identity-thesis, as defined above, only gives you the thesis that acting morally is *pro tanto, to a degree rational*. This is a thesis on a different level. And, you agree that there are other reasons, other considerations on the same level of moral reasons. So, even if we accept your identity-thesis, the question of moral rationalism is still left open.
The fact that there’s no sharp line between moral and non-moral reasons and that ‘intuitively’ moral reasons are overriding get you somewhere towards moral rationalism (and these are things many others have said too), but not quite there yet in terms of having an argument for moral rationalism. That argument for me would have to be a substantial argument about what the moral reasons are. Scanlon makes a good attempt on this.
Hi Mark,
Thanks. That clarifies the modest rationalism you aim to defend.
I think I have a more restrictive conception of criticism than you do.
I would not criticize a subjectively rational Oedipus, for example; instead, I would sympathize with his regret.
I’ve still got a number of Jussi’s first comments to respond to but I’m jumping ahead to respond to Jussi’s new comment and Steve’s comments.
Jussi,
I think I’m not that worried by the part of the problem you think remains on the second (subset of reasons) version of the view. I agree that if we define rationalism to require overriding strength for moral reasons the the view should not be labeled rationalism. But I wouldn’t put that into a definition myself – I think of rationalism as the view that the demands of morality are the demands of practical reason. And a pro-tanto view seems to me to capture this, since I read that as compatible with thinking that demands can be overridden. But at this point the dispute would just be about terminology.
Of course this worry doesn’t arise for the unqualified universal identity claim, the one I tend towards because I can’t find a way to divide strictly moral considerations from other considerations. I agree that Scanlon does very nice work. But if your idea is that we can use his reasonable rejection principle to find a dividing line between moral considerations and other considerations, I think he himself doesn’t buy it. He proposes that principle as covering the part of morality that consists in what we owe to each other. So I’m not seeing it yet.
Steve,
Thanks! I think the first comment is most troublesome for me so I’ll get to it last. The sentence you focus on on 513 is a first gloss which gets modified. As it stands it is incorrect. (Millians will have to say that it can be rational to aim at something one believes irrational, due to Frege’s puzzle considerations.) But examples of your sort (intentional irrationality in service of a good cause) weren’t exactly the sort of examples that lead to my qualifications. I need to think about these cases some more. I think that what we should say about the examples might turn on wrong kinds of reasons type worries.
The comment about p. 514 seems a problem for what I said (though we do find people claiming to do things because they are irrational, at least in books and classrooms). The thing I was trying to get at was something about what a view like mine with multiple kinds of subjective rationality should say is the important level to look at, or which one we do tend to regard as most important and hence tend to be talking about when we call someone irrational. And the thought was just that even though we can idealize an agent’s position so that there are more objective notions (such as what she would do if she was reasoning correctly and knew certain facts she currently does not know) when we call an agent irrational we are usually using one of the more subjective senses of rationality which make only accessible information relevant to the judgement.
Now for the hardest challenge:
What you’re proposing to me, if I understand, is that I and others who reject moral rationalism don’t know what moral rightness really is, or what normal people mean by “morally right”. It’s not that we merely have a faulty metaethical theory about our own practice, but that we’re like those who are ignorant that Hesperus and Phosphorus are the same “star”. To take the Fregean option, our mode of representation of moral rightness is something like “whatever it is that those normal people are referring to when they use the words ‘morally right'”. I find that incredible.
Well, with a largish quibble that is what I’m proposing, at least insofar as I’m proposing that it is like Hesperus and Phosphorus in one important respect.
The largish quibble is that I don’t think that people who don’t know that the morning star is the evening star don’t know what Hesperus is. They can know quite a lot about Hesperus. They know it is a star, that it is very far away, that it is largely made of hot gasses, that it is self-identical. There is a big difference in the ways we establish empirical identities of the Hesperus Phosphorus sort, and the kinds of identities we’re worried about in metaethics, but people can know a lot about the targets of such analyses without knowing enough to accept an identity claim as part of an analysis. There is lots you know about moral rightness even if the analysis is correct and you disbelieve the analysis. One of those things might be the one codified in your Fregean mode of representation. There will be various other things people agree on about moral rightness despite disagreeing about philosophical analyses; there will be certain things that your end-relational view will say that the correct analysis will also say, and likely more besides. If these claims are true and trusting that your views are well supported by decent arguments I think you know them. And your not knowing the correct analysis doesn’t undermine that claim.
This doesn’t sound so bad to me. It looks like a defender of any philosophical analysis not accepted by most people has to make similar claims.
Thanks Mark!
(1) Regarding the possibility of irrationality, I agree with you about the relevant level for those judgements; my worry (which is a case of pursuing my own rather solitary crusade) is that so clarified, irrationality turns out to be impossible. Usually I try to keep these worries to myself as they’re not widely shared, but I thought your presentation in terms of what makes sense brought out the problem with unusual clarity. (Note that you may be able to blunt the objection by allowing that irrationality may involve merely acting in ways that don’t seem to make the MOST sense).
(2) Agreed about Hesperus and Phosphorus, so your correction is well taken. What I find implausible, though, is the implication that moral rationalism’s opponents only know moral rightness by some of its accidents, and not by its essence. (This needs some argument, though, which I try to provide in my book).
But I disagree that “a defender of any philosophical analysis not accepted by most people has to make similar claims”. My objection wouldn’t have much force if it implied that the correct metaethical theory is uncontroversial. I think there’s an important distinction between the implicit knowledge of a property involved in first-order judgment, and the reflective knowledge we seek in metaethical theories. The Open Question Argument, for example, is designed to show metaethicists that their reflective theories are belied by their implicit knowledge. The difference between your position and mine is that I only attribute my philosophical opponents a mistaken metaethical theory, not a lack of implicit knowledge about what they’re talking about in first-order moral judgments. (Although some people seem to find that just as uncharitable, or even more so!)
They can know quite a lot about Hesperus. They know it is a star…
Uh, Mark, I hate to break it to you…
Thanks Mark. I’m not sure that the pro tanto view captures the idea that the demands of morality are demands of practical reason. I think most people read demands in that formulation as overall demands. That’s why many would think that the pro tanto view falls short of rationalism proper. Of course it’s interesting that reason requires moral action to a degree but this isn’t the original goal-post.
I also cannot see how the fact that one cannot divide strictly moral considerations from other considerations would support the unqualified identity thesis. Sure there are considerations of which it is hard to tell. But, then there are considerations of which we can tell that they are non-moral. And, there will be cases in which only these are in question. Is it rational to take a 5:1 or 9:1 one bet for Nadal winning the US open if given a choice? Unqualified identity makes taking the worst bet wrong and not merely identical. And it’s hard to see how this could be a question of us not knowing whether the question is moral or not.
Oh yeah – the reference about Scanlon wasn’t to the thought of how to divide moral and non-moral considerations. Rather, what I had in mind, and what I should have written, is that I quite like his argument that, if one is to argue for anything like rationalism, it is going to require substantial moral arguments. We better look at what the substantial reasons are to do the right thing, what other substantial reasons there might be in conflict with these reasons, and how these reasons weight up. This seems right to me. I remain unconvinced that any formal, metaethical, or philosophy of language accounts can justify this conclusion.
Mark,
As you know, I love this paper and am very glad Pea Soup chose it to discuss but wanted to join in offering some questions.
Let me first try to characterize the argumentative strategy of a part of the paper (p. 498-9) and, supposing I have that right, ask a question.
So I hear you as saying that the fact that 1 is “sufficient” to support 2 provided that you can make good on claim 3.
1) “When someone asks for a reason to do something, it is appropriate and not obtuse to explain that the action in question is morally right and to offer an explanation for why it is right. No further answer to the why question would normally appear to be needed.” (498)
2) “It is plausible that having a moral obligation to do something is necessarily a reason to do it…” (498)
3) The fact that rational people can doubt that they have a reason to do something even after learning that morality requires it is compatible with the claim that they really do have such a reason. (attempted paraphrase of 498)
If I have the argumentative strategy right, I am not sure about it. My worry about it is more about whether the conclusion follows rather than about the premises. I guess most generally I think a response being “appropriate and not obtuse” can do less work than the above argument gives it. I take it that it would also be not obtuse to reply to the why question by saying 1) it would be fun, 2) it would be the bold choice, 3) it would be the safe choice, 4) it is the choice Jesus would make, and so on for a very wide range of considerations.
So first I wonder if you agree that such responses would also not be obtuse? If so, I wonder if you think that all of the sorts of replies (such as the above) that would be “appropriate and not obtuse” should also be seen to make plausible corresponding claims about what is necessarily a reason? If not, then I take that as bad news for this argumentative strategy. If so (which would surprise me as I would have expected some non-obtuse answers to why questions to end up being false) I wonder if what this argument at most shows is that morality’s reasons are just another kind of reason in a great big pile of reasons and not yet shown to be especially significant.
OK, most of these are going to take some thought. Jamie’s comment on the other hand requires only that I concede that what I wrote about Venus was totally wrong.
These are all good comments. In the order which responses come to me (and probably putting off some for later):
Jussi,
Thanks for the clarification as to which part of Scanlon you meant to reference. I think I agree that justifying rationalism will require substantive normative reasoning. But I think that theory choice is holistic. So I think we can do that while also noticing that other, non-first order, arguments also favor or count against the position.
You write:
I also cannot see how the fact that one cannot divide strictly moral considerations from other considerations would support the unqualified identity thesis.
I think there are reasons that favor some sort of identity thesis. Among the options are different qualified versions and the unqualified versions. I can’t find a qualified version that seems plausible to me because I can’t find a good dividing line. So I opt for the unqualified version.
YMMV, after all this way of looking at things depends on assumptions I don’t think you share with me.
With respect to your more detailed argument, which turns on whether we think certain cases can or cannot plausibly be construed as involving moral rightness, I’m willing to think a false but widespread view that moral reasons are of a different kind than other practical reasons explains why people tend to regard these as counter-examples. My worries aren’t that the border is fuzzy. My worry is that there is no good candidate for a border that classifies the moral reasons as members of a non-gruesome kind of which non-moral reasons are not members. To put it another way, the only natural kind of which the moral reasons are all members also includes the other practical reasons. And, I’m a bit partial to natural kinds as reference magnets. All of that is controversial and not obviously correct, I agree.
Steve,
You write:
What I find implausible, though, is the implication that moral rationalism’s opponents only know moral rightness by some of its accidents, and not by its essence. (This needs some argument, though, which I try to provide in my book).
(I should probably read that book, and if my view is correct I probably ought morally to read that book.)
I’m not sure that I’m committed to saying that the opponents don’t know its essence. As long as I don’t know that what I know about it is all there is to the essence of a thing, I can be unsure whether it is the same as another thing which I know just as much about.
To be slightly more concrete, I can know that a, b and c are properties of X, and I can know that a, b and c are properties of Y. But I can fail to know that they are essential properties of X and Y. (Dalton would have identified the wrong features as essential to atoms, though some of what he knew was essential to them.) And I can rationally fail to know whether X and Y are identical, even though they share a, b, and c. Even if I know that a, b and c are essential to each of X and Y, I can still fail to recognize that they are identical, so long as I don’t know that these are the only essential properties of each. (This is still pretty abstract and I may be missing the point entirely.)
I’m pretty sure that for many people who are (by my lights) on the wrong end of this sort of metaethical disagreement, there are things they know about rightness which are essential to it.
You elaborate:
But I disagree that “a defender of any philosophical analysis not accepted by most people has to make similar claims”. My objection wouldn’t have much force if it implied that the correct metaethical theory is uncontroversial. I think there’s an important distinction between the implicit knowledge of a property involved in first-order judgment, and the reflective knowledge we seek in metaethical theories.
I think that perhaps my Millian tendencies are coming in here. I’m willing to say that a person knows something even when they only recognize what they know when it is put a certain way and are willing to deny the very same fact when they encounter it in a different way. It seems open to say that if you implicitly know something, you know it. And that remains true even when you go on to reject some explicit formulation of the very same fact.
But I’m not sure that I’m tracking your main concern here.
David,
The summary of the structure looks right and I feel the force of the worry. I’m going to mull this over a bit before I say anything more specific. Good question to which I’ll return after some thought.
Thanks to all!
…
…
David,
Having mulled your question over, I think I can say more. I think your setup was fair to the argument and I’ll repeat your main worry here for context:
I guess most generally I think a response being “appropriate and not obtuse” can do less work than the above argument gives it. I take it that it would also be not obtuse to reply to the why question by saying 1) it would be fun, 2) it would be the bold choice, 3) it would be the safe choice, 4) it is the choice Jesus would make, and so on for a very wide range of considerations.
I think you’ve shown I need to say more than just that it is appropriate and not obtuse to answer with a claim that the favored option is morally right because . . . whatever. Two or three patches come to mind, none of which I’m completely sure of.
Patch one: I think it is always appropriate and not obtuse to give the answer that it is morally right, whereas the candidates you offer are either often or usually (depending on which ones) but not always appropriate and not obtuse.
Patch two: Emphasize the “no further answer” clause in the condition more heavily than I have been. I guess I do think there are challenges to the candidate answers you offer, that their proponent would be obtuse if they didn’t answer in a way they wouldn’t be if they just didn’t answer a similar challenge to the “It’s morally right,” answer. I should clarify the sorts of challenges I have in mind. Any and all answers are open to a challenge that they are false. But challenges which grant the truth of the answer but go on to ask for a reason to regard the answer as sufficient to answer the why question seem to me more pressing with the examples offered than that sort of challenge would be where we’ve already been told that the recommended course is morally right and why it is right.
Patch three: The idea for this is pretty sketchy yet, but it has to do with the order in which these answers are open to challenge by what sorts of facts. If I tell you the favored option is morally right and you answer, “Yeah, but the alternative would be more fun,” my reaction is that you are being obtuse (or actually I would suspect you of making a joke). But if I said, the favored action is the thing to do because it is fun, it seems perfectly appropriate for you to respond, “Yeah, but it isn’t morally right to do that.”
FWIW, I’m using ‘because it is fun’ as the target because I think it is more general than the others. (The Jesus answer involves complications about constitutive and epistemic reasons that I’d rather ignore for now.)
I suspect that I see this as some good reason to accept internalism (once it is properly qualified) because I already have some partiality towards the view. I think that’s legitimate, but I also think that someone who started with different credences could legitimately think that even if the patches work the revised argument doesn’t provide strong enough reason to accept the conclusion. So I think this means that in the overall scheme of argument about the point, more arguments of a different sort would help make the case stronger and it would not be unreasonable to say that such arguments are needed to make the case fully convincing.
I’ll need top think about this more.
Jussi, in your first comment you asked to know more about the semantics. (FWIW, the paper was too long and some of the semantic details got cut to make it of publishable length.)
Specifically, you say:
But, anyway, in the example used (Evening Star/Morning Star) there is a story of how these terms came to co-refer – we were in a causal relation to one object whilst coming up with the names. Yet, properties like being right and being objectively rational don’t seem to get their reference in the same way. So how did they end up co-referring if not in that way?
I think that all of us have first personal thoughts about what we ought to do. It seems to us that some actions make sense to do and others don’t. And many of us find moral terminology appropriate for expressing those thoughts. “I wonder which choice is the right one to make,” etc. Somewhat speculatively, I think that many of these terms were either coined to express such thoughts, or were appropriated from some similar discourse by people to express these thoughts. I think for these people, and for other people whose use of these terms depends on these people’s use, the term refers to the property things have when they make sense to do. (Were such people in causal contact with the property? Well, they did have causal contact with instances of the property which I think should be good enough. I have worries about causal contact as the only reference enabling relation, but that’s going pretty far afield here.)
Terms like ‘objectively rational’ are philosophical technical terms. And they shift around quite a bit even in philosophical uses. That’s why the paper has a short section giving my use of ‘rational’ the “makes sense” gloss. I guess the thing to say here is that this means that some people doing philosophy wanted to be in a position to talk about the property things have when they make sense to do. They may have wanted to be able to ask whether right actions always made sense to do. And so they stipulatively defined a term for that, probably also borrowing a term from some existing use that may or may not have been similar to the stipulated use.
This answer may look like it is too easy. But it does seem sufficient to me to explain how the terms come to co-refer. So I’m wondering if there is another worry connected to this one. Is there also a worry lying behind your comment about how the two terms could have the same referent while speakers remain ignorant of that fact?
Mark,
I like the direction of your patches. Not that I think any of them are slam-dunks but they seem to me the right directions for you to move in–roughly all strengthen a bit the first premise. The first premise will then, as you point out, be more debatable, but the form of rationalism you are trying to vindicate will flow more validly from such versions of the first premise. Wondering about such versions of the first premise seems to me a fruitful area of inquiry for all of us.
Hi Mark,
Thanks for the excellent paper. I’m sorry that I’m late to the discussion and that I have, at this point, so little to add to what others have already said.
You write above: “That a certain course of action would require too much by way of personal sacrifice seems to me a reason to think that some moral requirements are less stringent than some think. But if that is true we have a case where certain sorts of self-interested reasons change the shape of moral requirements. And I think that makes them moral reasons – reasons relevant to determining what is morally right to do.”
I think that this last statement is false, for I think that it’s important to distinguish a moral reason (a reason that counts morally in favor of performing some act) from a morally relevant reason (a reason that’s relevant to determining the deontic status of some act). Some reasons (e.g., those concerning the personal sacrifice of physical pleasure) may have purely moral justifying strength. That is, they may justify not doing something that would otherwise be morally obligatory without themselves counting morally in favor of the act that they justify. I tend to think that a moral reason is one that can do more than merely justify not doing what would otherwise be morally obligatory. More specifically, I tend to think that a moral reason would have to have the “power” to make an act morally required (or, at least, morally supererogatory) in the absence of countervailing reasons. I think that some morally relevant reasons don’t have this “power.” For instance, the reason that I have to avoid sacrificing a great deal of physical pleasure does seem to be a morally relevant reason in that it can justify not performing an act that would otherwise be morally obligatory. But it doesn’t seem to be a moral reason. The fact that I would incur a lot of physical pleasure by performing x doesn’t make x morally obligatory or morally supererogatory even in the absence of countervailing reasons.
Doug,
Thanks! I think what you suggest in your could be a way to divide moral considerations from other considerations:
I think that it’s important to distinguish a moral reason (a reason that counts morally in favor of performing some act) from a morally relevant reason (a reason that’s relevant to determining the deontic status of some act). Some reasons (e.g., those concerning the personal sacrifice of physical pleasure) may have purely moral justifying strength. That is, they may justify not doing something that would otherwise be morally obligatory without themselves counting morally in favor of the act that they justify. I tend to think that a moral reason is one that can do more than merely justify not doing what would otherwise be morally obligatory. More specifically, I tend to think that a moral reason would have to have the “power” to make an act morally required (or, at least, morally supererogatory) in the absence of countervailing reasons.
Whether it’s right or not will, I think, depend on stuff about how reasons work that I’m not settled about. That’s partly why I’m very interested in Mark Schroeder’s stuff on weighing reasons. And I think complications about different ways that reasons can determine what makes sense are also relevant. I’m thinking here of Josh Gert’s distinction between reasons which (normally) require and those (“purely justificatory reasons”) which permit us not to do what would otherwise be required. I think this is the sort of issue that illustrates Jussi’s suggestion above that at least some metaethics depends on substantive theorizing about normative matters. We’ll need to do some work to see if the kind of distinction you suggest can be carried over so as to generate a general division between moral and non-moral considerations. (That’s not meant to be a problem for your suggestion; it just indicates my current agnosticism about using it to generate a division.)
Suppose we get the sort of general account I think we’d need. I now wonder if that will automatically give us an account of moral rightness that makes it distinct from other sorts of rightness. In many cases, whether an action is morally required will still be a function of all the reasons that apply to it, including those we have classed as non-moral. And in many cases, whether an action is morally permissible will similarly depend on all the reasons. (I’ve been ignoring the issue of whether rightness involves requirement or just permissibility, partly because I’m not settled about it.) So I don’t immediately see how we can get a distinction between rational permissibility and moral permissibility, or rational requirement and moral requirement from a division of reasons in to moral and non-moral. I can kind of see that we could generate some divisions between requirements and permissions based on the sorts of reasons that ground them. And we might stipulate some labels for different sorts grouped in that way. But I’m less sure that the natural language terms ‘morally required’, ‘nonmorally required’, ‘moral permissibility’, and ‘nonmorally rational permissibility’, and so on, will map onto those divisions in such a way that it would be plausible that these are already their meanings.
But this is worth further consideration, especially as I think it all turns on lots of details.
Jussi , you were/are unhappy with my use of terminology:
I wish Mark hadn’t use ‘existence internalism’ for the claim that having a moral obligation is necessarily a reason to do it. This term seems to have a more popular use when we think about whether the existence of reasons/obligations depends on our motivations. The claim Mark discusses is often just called Moral Rationalism (the term Mark uses for the claim that requirements of morality are requirements of reason – which seems like the same claim as his existence internalism). In any case, the terminology is confusing with these terms generally.
At least one other knowledgeable person has found my use here somewhat confusing. I’ll explain why I think it is apt, while admitting that not everyone currently speaks this way. My thinking about it is somewhat framed by having been introduced to the terms by Michael Smith, and also some of Darwall’s nice discussion of the different sorts of internalism.
As I recall Michael used the term ‘internalism’ as a label for the views that postulated a necessary or conceptual connection of certain sorts involving either motivation or reasons as one of the relata. One such connection is the view he labels ‘rationalism’. It amounts to the claim that necessarily if something is right for you to do in a circumstance then there are reasons for you to do it in that circumstance. (TMP, 62) And, as you note, that’s what I’m also sometimes calling “Moral Rationalism.” Thus, I get from Michael the thought that rationalism is a kind of internalism.
But it is only one kind of internalism. And it seems to me that we can ask a bit more about what kind it is. Obviously it is a kind of morals/reasons internalism to use Darwall’s useful slash name labels. But we can go on to ask what kind of morals/reasons internalism it is.
Another distinction between kinds of internalism is the one that Darwall has highlighted when talking about reasons/motives internalism. Darwall used the labels ‘judgement internalism’ and ‘existence internalism’ to mark the following distinction. Some internalisms postulate a necessary connection between a judgement of a certain sort and motives, whereas another kind postulates a necessary connection between the existence of a reason (whether recognized or not) and motives. Darwall never (so far as I know) explicitly makes this distinction when addressing morals/reasons internalism, but the same distinction is there to be made. An internalist of the morals/reasons sort might think it necessary that a belief about moral obligation requires or generates a reason, or a morals/reasons internalist might think it a necessary condition on the existence of a moral requirement that there be such a reason. This is the same kind of distinction as we can make with reasons/motives internalism. It seems to me that it would be confusing to introduce new terms for this distinction when we encounter it with morals/reasons internalism rather than just carry the same terms over from reasons/motives internalism to mark the same sort of distinction. So that’s what I’ve done. In fact, I think this is probably what Darwall had in mind. In Impartial Reason (p. 54) he makes the distinction with respect to both reasons/motives internalism and morals/motives internalism. It seems very natural to extend the idea to the one remaining sort of internalism, morals/reasons internalism. (There’s also relevant discussions of the judgement/existence distinction in his paper in Darwall, Gibbard, and Railton at around pp. 306ff. )
I should note that rationalism is one kind of morals/reasons existence internalism. But there might be other sorts of morals/reasons existence internalism. So there is a point to using the words ‘existence internalism’ in addition to the word ‘rationalism’.
I think that it would be a really good thing if the field came to always use Darwall’s slash names whenever they were talking about internalism, since even before we get to the judgement/existence distinction, there are three candidates for what we might mean by ‘internalism’ (leaving aside the uses outside metaethics). And, since I think the judgement/existence distinction cuts across all of these categories, we could use that distinction to generate roughly six types of internalist view, each of which is of some interest. But I don’t have the sociological heft to get the field to adopt this practice.