I'm really enjoying Scanlon's new book "Being Realistic About Reasons" (all citations below to this book), but I'm stumbling on the part about pure normative truths and the explanation of supervenience. Any help would be much appreciated.
Scanlon takes the reason relation to hold between tuples, (p, x, c, a), where p is a fact (a reason), x is an agent, c comprises some conditions (circumstances), and a is an act type. In a case where p = the metal is sharp, x = you, c = normal circumstances where you would avoid the pain by not pressing the metal, and a = not pressing the metal (3), Scanlon says the reason relation R(p, x, c, a) obtains (30).
He calls this a mixed normative fact because it contains normative and non-normative elements; this reason relation only obtains when the metal is indeed sharp. Aside from these, there are pure normative claims: "the essentially normative content of a statement that R(p, x, c, a) is independent of whether p holds. This normative content lies in the claim that, whether p obtains or not, should p hold then it is a reason for someone in c to do a. So I will take what I will call a pure normative claim to be a claim that R(p, x, c, a) holds (or does not hold), understood in this way" (36-37). Later he says the contingent non-normative elements are "subjunctivized away" in pure normative claims (40). So it looks like the pure normative claim for the sharp metal case is something like this:
(Pure?) It is normatively necessary that, for all agents, x, if it were the case that p and x were in c, it would be the case that x has a reason to a (40-41).
One of the advertised payoffs for separating the pure from the mixed is that it allows us to explain the supervenience of the mixed normative facts on the non-normative facts by appealing to the pure normative facts. But I am puzzled.
First the pure facts. As Schroeder's review in the AJP also points out, (Pure?) is an odd candidate for pure normative content. It looks like is is a claim about when pure normative content would hold (i.e., a condition under which there would be a reason). It not only subjunctives away p, but the reason relation itself. And it looks like a claim in need of a truth-maker.
I get my head around this by thinking of the pure normative content in terms of Ewing's moral laws (or Enoch's norms with modally maximal jurisdiction). True claims about them are necessarily true; they tell us, for any possible non-normative facts, whether they have normative significance (42); and they do not contain any natural elements. Whether they are purely normative themselves is questionable, though. They seem to be normative status conferring without themselves being normative (or natural). But I would love suggestions about how to understand pure normative content in other terms.
Second, explaining supervenience. Why can there be no change in the reason-relatedness status of the tuple (p, x, c, a) without some change in the non-normative features exhibited by (p, x, c, a)? The classical realist explanation is to posit real necessitation relations between non-normative properties and normative properties/relations (I think of a really strong towline). Ewing definitely wanted to avoid such "necessary synthetic connections", so he gave a different explanation: there are necessary moral laws (here, we can say normative laws) that somehow confer normative status on items with certain non-normative properties. Scanlon seems to opt for this second explanation, and finds supervenience less puzzling in light of it.
I am not so sure. First, I see dimly what it would be to have an explanatorily basic necessary synthetic connection between distinct existences (though I try to avoid them). But I do not see what an explanatorily basic necessary normative law would be. For a familiar law L (say, a traffic law), there are some non-legal facts in virtue of which L. But Ewing-Enoch-Scanlon necessary normative laws are not laws in virtue of anything. (Well, for Ewing, it might be in virtue of God's attitudes). For me, sui generis laws seem more puzzling than sui generis necessary synthetic connections.
Second, I see dimly what sui generis normative properties would be on the classic realist explanation (though I try to avoid them) - they are just intrinsically authoritative properties necessitated by certain non-normative properties. But I do not see what the normative statuses (properties?) conferred by necessary normative laws amounts to. Is the reason relatedness for a tuple *reducible* to facts about the necessary law? Is it that, for a tuple to be reason related just is for the pure normative law to decree that tuples like that are reason related? That makes the normative authority of the status (property?) somewhat elusive. But the alternative is to have non-normative properties, irreducible normative properties/relations that are intrinsically authoritative, and Scanlon's explanatorily basic necessary normative laws ensuring (describing?) the modally systematic connections between the two. I do not see how that is a better explanation than positing explanatorily basic necessary synthetic connections between the two. It can be hard to see how it is a different explanation.
Last, Blackburn's explanatory burden still lingers. Though there is no agreement about what that is, I think it comes into focus when we attend to the full statement of general strong supervenience.
(SS) Necessarily, if any tuple (p, x, c, a) exhibits the reason relation, R(p, x, c, a), then there are some non-normative features that the tuple exhibits, NN, such that, necessarily, any tuple (q, y, d, b) exhibiting NN bears the reason relation, R(q, y, d, b).
I take it that the wide scope necessity is conceptual and the narrow scope necessity is normative.
I think the burden is to explain how it is that we can know *as a conceptual truth* that there are normative necessities (synthetic relations or laws) dictating which non-normative properties are of normative significance when the normative necessities are not conceptual truths. If a normative necessity is to be discovered (perhaps a priori but not via conceptual analysis), how is it that we know *as a conceptual truth* that there is some necessity to be found prior to discovering it? Very odd. (BTW, my friend Cole Mitchell has a great paper on Blackburn's challenge.) This probably isn't on Scanlon's plate, so we can set it aside if you like.
This is too much, isn't it? OK. Thoughts?
Matt — Nice post. Very quickly…
You write:
“But I do not see what an explanatorily basic necessary normative law would be. For a familiar law L (say, a traffic law), there are some non-legal facts in virtue of which L. But Ewing-Enoch-Scanlon necessary normative laws are not laws in virtue of anything.”
I take it that anyone who accepts normative facts but denies that normativity has a “source” will think that there are the kinds of laws that you claim not to understand. So do you find the idea of sourceless normativity difficult to make sense of? What about the idea that there’s a source of normativity, but without any explanation of why that’s the source? What about the idea that there’s such an explanation, but no further explanation of why it is the correct explanation? And so on…
Also, a quick speculation about what Scanlon’s thinking: My bet is that he thinks that he has cleared up any genuine mystery about the connection between normative facts and non-normative facts by adverting to “pure” normative laws, and that, if you think there’s a *further* mystery of, say, a “Hume’s Dictum” sort, it’s because you’re assimilating metaphysics of the normative to the metaphysics of material objects, which he cautions us against earlier in the book. (Well, I guess he talks specifically about normative ontology, but you can see how maybe that would cross-apply to other issues in normative metaphysics.)
Hi Andrew,
I don’t find source-talk all that helpful. I’m fine with the idea that the explanatory buck must stop somewhere, but I’m wondering why it is better to stop with pure normative laws (or pure subjunctive principles or whatnot) rather than necessary synthetic relations. I don’t see how the laws or subjunctives explain more than the relations, or explain the same stuff better. And the laws or subjunctives seem to raise some explanatory puzzles that the relations do not. For example: what is it for a tuple to be reason related? I this distinct from the tuple and the pure stuff, or somehow reducible to the tuple and the pure stuff? Maybe just answering that would help me out. I’m guessing reducible. Is that right?
Hi Matt— to “stand in the reason relation” is just for the corresponding reason claim to be true. It is not true in virtue of any further fact. I am not sure what such a further fact could be. It could not be a non-normative fact. I argue that it can’t be a fact about rationality (understood in a sense that does not presuppose facts about reasons); so it would have to be just another fact about reasons, which would leave us in the same place. I think the main difference between us lies here: you think that normative facts would have to be explained in terms of some further facts. I think that cannot be done and there is no reason to think that it is needed.
I am pleased that you found what I said of interest. Here are a few thoughts in response, which may not satisfy you.
First, although it may not matter, I was not thinking of pure normative claims as having the form “It is normatively necessary that for all agents p if it were the case that p and p were in c then x would be a reason for p to do a.” What I had in mind was just the part of this claim that comes after the italicized bit. I said I thought that such claims were normatively necessary, but I was not thinking of that claim of necessity as being part of every pure normative claim. Maybe I should think of pure normative claims in that way, and maybe it makes no difference, but I thought I would mention it.
Leaving that aside, however, I disagree with you and Mark Schroeder about the normative character of pure normative claims. I believe that they are fully normative and are what accounts for the normative content of mixed normative claims. One thing you could have in mind when you say that a pure normative claims seem to you to be in need of a truth maker is that the fact that p, if it is a fact, and the fact that x is in c, if it is a fact, are what makes it the case that p is a reason for x to do a. But this seems to me to rest on a failure to distinguish clearly between pure and mixed normative claims. The non-normative facts just mentioned are part (but not all) of what makes it the case that x has reason to do a, and that p is a reason for p to do a. But since these facts are non-normative, they cannot be all of that makes this the case. The facts that p and that x is in c are one thing; the normative fact that p is a reason for x to do a is something else, a normative fact. There is a difference between the (non-normative) fact, p, that is a reason and the (non-normative) fact that p is a reason. (I believe that normative facts can be reasons too. I am just focusing here on non-normative facts to make things clearer.)
So in my view, normative facts are just facts about which things are reasons. Pure claims about such facts, when true, are necessary. I don’t believe that the pure normative content of such claims can be “understood in other terms.” For a tuple
You might ask why I prefer a view that is committed to sui generis normative facts rather than one that is committed to necessary connections between non-normative properties and normative properties. My reason is that the necessary connections in question seem to me to be normative, rather than necessary connections of some other kind between non-normative and normative properties. Facts about which non-normative facts are normatively significant are just normative facts about when reason relations hold, not facts of some other kind about the relation between these two kinds of facts.
What I found puzzling about supervenience was why, if normative truths are not logically or conceptually derivable from non-normative truths, they nonetheless “march in lock step with them.” My explanation is that this is true of only some normative truths, namely the mixed ones, and the linkage holds in virtue of pure normative facts. You seem to be asking how one can know, “as a matter of conceptual necessity,” that there are any pure normative truths. I have also worried about this. The normative claims we commonly make are mixed normative claims. And as I say in Lecture 5, we are often somewhat unclear about exactly how the “conditions c” would need to be spelled out. But I am inclined to think that this can in principle be done. And if all of the non-normative facts that a reason relation depends on are captured in p and c, then there is no further dependence of R(p, x, c, a) on non-normative facts. I don’t know that this is a conceptual truth, but it seems very plausible to me. There is then the further question whether, if pure normative truths do not depend on non-normative facts, they nonetheless “could have been otherwise” in some other way (i.e. are not necessary.) I would not say that it is a conceptual truth that they could not be otherwise, but, thinking about it carefully, I don’t see how they could. That, it seems to me, is how we arrive at the view that many other kinds of truths are necessary. But I could, of course, be mistaken about this.
Hi Tim,
Thanks for the reply! There was too much in my initial post, so let me try to home in on the main stumbling blocks for me. I’m not asking for an explanation of normative facts in terms of other facts. And I’m clearly distinguishing the fact that p and the fact that p is a reason.
What I’d like to understand is what pure normative facts are (it doesn’t have to be in non-normative terms) such that 1) they explain why the mixed normative facts supervene on the non-normative, and 2) they provide a different and better explanation than necessitation relations.
One problem with stating the pure normative facts with subjunctive claims is that those claims, when additionally understood as necessary, are too close to a statement of supervenience to count as an explanation of it. We don’t want to say that the pure facts are facts such that, if they are necessary, the mixed supervenes. Hence the suggestion that we think of the claims as a) about necessitation relations between properties or b) about normative laws about the conditions under which mixed facts obtain.
It looks like you go for (b), and say it is better than (a) because “the necessary connections in question seem to me to be normative, rather than necessary connections of some other kind between non-normative and normative properties.” But the fan of necessary relations could say with equal plausibility that these relations are “normative, rather than necessary connections of some other kind”. You also say “Facts about which non-normative facts are normatively significant are just normative facts about when reason relations hold, not facts of some other kind about the relation between these two kinds of facts.” But the fan of necessary relations could say with equal plausibility that facts about which non-normative properties necessitate which normative relations “are just normative facts about when reason relations hold, not facts of some other kind about the relation between these two kinds of facts.” So these considerations cannot help us pick between the views.
To take stock, I’m working within theories that are realistic about reasons, and just wondering what distinguishes the pure fact view from other views and why to go in for it. The subjunctive claims don’t give me a good handle on what the difference is – if there are normative necessities between properties the subjunctive claims will be true. And all the realists are saying that there are non-normative facts (that it would hurt), mixed normative facts (that it would hurt is a reason . . .), and then explaining supervenience with normative necessitation. Why go in for the normative law view of that necessitation rather than the relational view? They seem explanatorily on a par.
. . . Unless, the law view reduces mixed facts such that what it is for p to be a reason for x to phi in c is for pure normativity to say thus-and-such about p, x, phi and c, and for p, x, phi, and c to obtain. That suggestion hasn’t been adopted yet. To the extent I understand it, I see problems with it. But without it, I don’t see the explanatory payoff of pure laws.
Thanks again for the post!
Matt–
I am getting a bit unclear as to what the disagreement is about. You write:
One problem with stating the pure normative facts with subjunctive claims is that those claims, when additionally understood as necessary, are too close to a statement of supervenience to count as an explanation of it. We don’t want to say that the pure facts are facts such that, if they are necessary, the mixed supervenes.
You may not want to say this, but I am inclined to do so. It seems to me that supervenience is puzzling if one thinks of it as a having to do with a linkage between two kinds of facts, non-normative and normative. My point is that this linkage (which non-normative facts are linked with which normative facts) is itself a set of normative truths. This does not explain the phenomenon if by that you mean explaining why the particular linkages that hold do hold. But it seems to me to redescribe what is going on in a way that makes it less puzzling.
Once one has accepted this way of looking at the matter, I am not certain what the difference is between saying that the linkages are a matter of normative laws or a matter of normatively necessary connections. It is particularly hard to tell the difference given what you say about the necessitation view in your fourth paragraph. In particular, just calling that view a “necessary relations” view makes it hard for me to see what the important difference is supposed to be.
Tim,
Thanks for pushing me on this. I’m starting to see that the key for you is thinking of the linkages as themselves normative. So maybe it matters less what form those linkages take so long as they can be seen as themselves normative. I’ll mull over that.
In the meantime, regarding the difference between the necessary relation view and the law view, I was thinking that the former view has it that some facts about tuples involving p necessitate (mixed) facts about p being a reason. This would be an explanatorily basic necessitation relation between distinct sorts of facts. (I was thinking it could be thought of as a normative necessitation relation, which made it look more like your view. That might have been hasty – more on this shortly). Any law or subjunctive asserting the conditions under which p is a reason would only be a description of the patterns of mixed facts secured by this necessitation relation. So the truth of any such law or subjective would not help to explain or make it the case that the mixed facts supervene, nor would they explain or make it the case that the necessitation relations hold.
Alternatively, the law view says that truth of the subjunctive claims asserting the conditions under which p is a reason pulls explanatory weight. These are not mere descriptions of the patterns secured by an explanatorily basic necessary relation between two kinds of fact. Somehow the true subjunctives (laws or principles) help to make it the case that the mixed facts pattern with the non-normative facts as they do.
Maybe I’m wrong that this distinction makes sense. But if it make sense, maybe I was too hasty in thinking that the necessary relation view could say the relation between non-normative and (mixed) normative facts is itself normative. Maybe that relation can only be normative in the anemic sense that what it necessitates is normative. Maybe there is a deeper way in which normative laws are normative, contra what Mark and I were thinking. I need to think about that more, and how it might translate into a better explanation of supervenience for the mixed facts.
Thanks very much, Tim, and I should thank Barry Maguire for some helpful discussion off-blog.
Matt,
Think of the circumstances, c, as extremely specific. Indeed, make them so specific that the slot for the person is no longer necessary: the relation is tertiary, even though we say of a person that she has the reason, when she is in the circumstances (and the proposition is true). When you think of it that way, there is no question of why the normative status of the triple cannot change without a change in the non-normative features of the triple. The non-normative features of that triple cannot change! If you change the circumstances in any way, you have a different triple. So the triples have their non-normative features necessarily.
If I were a Reasons Fundamentalist, I would say that the basic facts, the explainers, are the particular (necessary) reason facts, like the fact that in my exact circumstances the further fact that you speak English is a reason for me to write in English. (I think this is necessary – have I missed something?) The laws, if there are any, are merely summaries of these facts, not explainers of them. Of course, I am not a Reasons Fundamentalist.
I think of the burden a bit differently from you. The way we know conceptual truths is pretty much always the same! The question, I think, is how the necessity embedded in the conceptual truth can be true. In the consequent of the quantified, conceptually necessary conditional you have as SS, there is another necessitation between something (in this case a quadruple) having a natural property and its having a normative property (in this case R). How does that work, that necessitation? How does the NN *make the quadruple have R? Or, more simply, I wonder why that quadruple couldn’t have failed to have R. Things, including quadruples, do not usually have their properties necessarily; when they do we are entitled to ask why, I think. Of course, there may be no explanation. But a theory that can provide no explanation is in that way worse than a theory that offers one. My dissatisfaction with Reasons Fundamentalism is that it doesn’t offer explanations for so many of the features of the normative world.
Hi Jamie,
I’m not sure if I understand your suggestion. If we make the non-normative specification of the circumstances in the tuple maximal, I can see that the non-normative features of the tuple could not change without changing the tuple. But that does not entail that the tuple has its normative features necessarily. It seems to be an extra claim that tuples, even ones with maximally specific circumstances, have their normative features necessarily. I’m not questioning that further claim, just wondering how maximally specifying the circumstances of the tuple eases the explanatory burden for the reason fundamentalist, as opposed to merely identifying the thing to be explained as strong global supervenience (in my formulation of SS I leave open whether NN could be maximally specified non-normative features of a centred world). (Maybe you weren’t suggesting that it eases their burden? Am I missing part of your point?)
Then, one explanation offered by the reason fundamentalist is that there are these explanatorily basic in-the-world (metaphysical? normative?) necessitation relations between tuples (with max c) and reason facts. (We probably agree in not being too satisfied with that explanation).
I agree with you that the reasons fundamentalist is better off making the laws (or whatever the pure facts are that articulate under what conditions the reason relation holds) just summary descriptions, rather than helping to explain supervenience. But I think that Scanlon, and others like Enoch and Ewing, are articulating a different way of thinking about this. (Maybe I’m wrong about that.)
As for your understanding of the burden of explaining supervenience, I agree that part of the burden is to explain why it is that a given tuple that is reason related could have failed to be reason related. But I think the puzzle is cranked up a notch with the additional claim that we know as a conceptual truth they have their normative properties necessarily. The burden is not to explain how we know conceptual truths; as you say, the way we know conceptual truths is always the same. Instead, the burden is to explain how in the world could we know *in that way* a claim like this: For all tuples T, either necessarily R(T) or necessarily~R(T). Outside of normativity, supposing something x has an essential property p, it is not usually a conceptual truth that either x essentially has p or x essentially lacks p. Usually, as a purely conceptual matter, x could have p contingently. Normativity seems to be exceptional here. Doesn’t it? (I think this brings us back to Blackburn’s point, which I wasn’t pressing with Scanlon.)
Sorry, in my last analogy I meant to be speaking of claims about essential properties that are not conceptual truths or analytic. Of course, if it is a conceptual truth that x is essentially p, it is a conceptual truth that either x is essentially p or x is essentially not p.
If I don’t write this without skipping thinking about the comments or even reading them carefully I won’t write it at all. So I’m skipping all of that and writing something off the cuff in response to the very last bit. Matt writes:
I think the burden is to explain how it is that we can know *as a conceptual truth* that there are normative necessities (synthetic relations or laws) dictating which non-normative properties are of normative significance when the normative necessities are not conceptual truths. If a normative necessity is to be discovered (perhaps a priori but not via conceptual analysis), how is it that we know *as a conceptual truth* that there is some necessity to be found prior to discovering it?
One way to take Quine’s arguments against the analytic/synthetic distinction is as suggesting that some of the truths that we think of as built into our concepts or meanings get their *justification* not in virtue of their being built into those meanings but in virtue of some independent source. Only when they are independently justified do they get taken up into the meanings of the terms we use. Truths which are like that, may well be analytic (insofar as they truths of meaning) but their status as justified rests ultimately on something synthetic, that seemed important enough to think of as part of the meaning of the relevant term. So as Gil Harman has suggested (if my memory is not totally unreliable), the upshot of Quine’s arguments was that there is no class of truths that has the epistemic status that analytic truths were supposed to have on the picture Quine questioned. It isn’t the fact that analytic truths are part of the meaning of the relevant claims that justifies us in accepting such truths, it is rather that they are justified that makes for little risk in building such claims into the meanings of the relevant terms.
So similarly, in this context, you could think that the supervenience becomes a conceptual truth because we think conceptual competence with moral terms requires recognition of some substantive moral truth.
It is very late and I can’t tell for sure if this suggestion is to the point or not in response to your last comment. Nor am I completely sure it is coherent. I may regret not deleting this comment and hitting post instead. I guess that is one of the risks of commenting on a blog late at night.
Tim says: “I would not say that it is a conceptual truth that they could not be otherwise, but, thinking about it carefully, I don’t see how they could. That, it seems to me, is how we arrive at the view that many other kinds of truths are necessary. But I could, of course, be mistaken about this.”
I think this is the nub. I find it easy to “conceive” circumstances in which the pure reasons facts are otherwise than they are — circumstances in which, e.g., the fact that and act would make X famous is an underived reason for X to do A. (This might be a world in which the normative assumptions of the Iliad are correct.) Is this “conceivable” world a metaphysically possible world? Well, if you’re a reasons fundamentalist (as I sometimes am), you can’t block the inference from conceivability to possibility in any of the Kripkean ways. So why not acquiesce in it and say, “Fine, the pure normative truths could have been as the Iliad says they are, so the pure normative truths are contingent”?
This is not to say that the normative truths are contingent “on” the natural facts. The basic ones aren’t contingent on anything (besides themselves). But still they could have been otherwise, so they’re contingent. (In this respect they’re like basic non-Humean laws of nature.)
I think it’s totally harmless for Tim’s sort of reasons fundamentalist to take this route. Everyone should agree that Iliad-worlds are conceivable in a pretty strong sense. It doesn’t matter for ethics or metaethics whether we call them “metaphysically possible”, and the best account metaphysical possibility (I think) nudges us in that direction.
*caveat* I may be misunderstanding something.
I want to jump a bit on Mark’s bandwagon here (i.e. addressing the quote from Matt in Mark’s comment) with a slightly different point. Surely one view that has the combination of commitments Matt finds puzzling is a way of taking Kripke/Putnam, etc on the necessary a posteriori. There is a conceptual cum metaphysical truth about identity (say, that whatever water is, it necessarily is) which in conjunction with other facts yields a metaphysical truth (water is necessarily H20) which is not discovered through conceptual analysis nor in anyway a conceptual truth.
Matt,
That bit (about how the ordered tuples have their non-normative properties necessarily) was just a response to this:
There’s something wrong with that way of putting it, since there cannot be any change in the non-normative features of the finely-specified quadruples.
I don’t think the conceptual necessity is a particular issue for any non-naturalist. So, no, it doesn’t seem to me hard (for anyone) to explain how we know a priori that for each t either necessarily R(t) or necessarily ~R(t). It’s like knowing that each number is either necessarily prime or necessarily not prime. Of course, in that case we have a better story about how numbers have their factors necessarily… but that’s the part I do think is important. That second necessitation operator.
Compare a different kind of necessity: causal necessity. Think of killings. It’s a conceptual necessity that each killing causally necessitates a death. How do we know this? Just by familiarity with the concepts of killing and death. But we can ask of each killing how it necessitates a death, and the answer is not at all trivial. Likewise, I believe, it may be a trivial conceptual truth that non-normative facts necessitate normative ones*, but still a substantive and difficult question how they do it in each case.
*I admit that if it is indeed a trivial conceptual fact, it’s a bit surprising that Gideon denies it!
Hi Jamie,
Suppose you met a tribe of people like me: English-speakers who are all happy to say that in remote possible worlds –e.g., Iliad worlds — the non-normative facts are just as they are but the normative facts are different (because the pure normative facts are different). They are happy to concede that there is a sense in which these worlds are impossible. They’re inconsistent with the pure normative truths, so they’re impossible in a restricted sense akin to physical impossibility. (Fine would say that they are “normatively” but not metaphysically impossible.) Do you think these guys are all incompetent with the concept of metaphysical possibility? I lose my grip on “conceptual possibility” when people say that whole tribes of clear-headed apparently sensible people could be conceptually incompetent in this way.
Mark,
If I understand you, you suggest that competence requires recognition of some substantive moral truth. I’m suspicious of that, though I know there are folks out there who think supervenience only follows from some first-order commitments. My last para below, in response to Jamie, picks up on this.
Gideon,
I agree that Iliad pure normative standards are conceivably true. I take it that the reason fundamentalist wants to say that if those pure standards are actually false, then they are not normatively possible (though still conceivable). The pure normative truths aren’t contingent in that way. Right? I’m asking because it looks like you are tempted to say that if those standards are actually false, they are still metaphysically possible. Then I just wonder what we accomplish by distinguishing metaphysical and normative possibility in this way. In other words, why are we talking about three kinds of possibility? The data involve two: one sense in which the pure norms are possible, and one sense in which they are not possible. I’m not up on my metaphysical vs. normative possibility, so apologies if I’m missing something.
Ah! Is the idea that, when we combine normative possibility talk with possibility talk in other domains we find some normative-but-not-metaphysical possibilities (maybe worlds where taking scalps is bad but water is not H2O), or some metaphysical-but-not-normative possibilities (worlds where water is H2O but taking scalps is good)?
Jack,
I think there are explanations available for necessary a posteriori truths about water and H2O and the like that are pretty compelling (and respectful of Kripke-Putnam points about facts external to cognition helping to fix reference), but those explanations are unavailable to the reason fundamentalist. You might disagree with me there.
Jamie,
Just to clarify: I would not deny that it is a conceptual truth that non-normative facts necessitate normative ones. I just think that is an interesting that such content is a conceptual truth.
You say “It’s like knowing that each number is either necessarily prime or necessarily not prime. Of course, in that case we have a better story about how numbers have their factors necessarily… but that’s the part I do think is important. That second necessitation operator.” But that’s it. The stories about the second necessity operator that help make sense of how we could know them in the first-necessity-operator-way seem to me unavailable to the reason fundamentalist.
I’m not sure about the killings and the numbers examples. I want an example (non-normative) where it is not an analytic or conceptual truth that a is F, not a conceptual truth that a is not F, but it is an analytic or conceptual truth that either necessarily a is F or necessarily a is not F (these last two necessities not being conceptual/analytic). Maybe there are examples, but then I want to know what the explanation is for that, and whether the explanation is available to reason fundamentalists. I think not – I don’t think there is an explanation for them.
I think some of these examples proceed by first coming to know that necessarily a is F, and then building that into the concept so that it is a conceptual truth that necessarily a is F (or, we can tack on, necessarily a is not F). (Maybe this is part of Mark v. R.’s point). But one striking thing about strong supervenience in normativity is that we can know the conceptual truth without having a clue as to what the pure normative facts are (the normative truth). (I’m pretty clueless there :)). That makes you examples about killing and numbers different, I think.
Tim,
Apologies for straying from your main points a bit, but this has been fun stuff. I’m warming up to the idea that there are ways of thinking of the pure normative facts that might be more illuminating/explanatory than thinking of them as facts about brute necessitation relations between two sets of facts. Not convinced, but warming up.
Gideon, you lose your grip on “conceptual possibility” when people say that whole tribes of clear-headed apparently sensible people could be conceptually incompetent in this way? But couldn’t they just have a slightly different concept? I don’t get your test. No matter what conceptual impossibility you come up with, I can describe a tribe of people who will be insisting on it if we stipulate that they are using the same concept. I can make them as (otherwise) clear-headed as you like. So there will be no such thing as conceptual impossibility, by this test.
Matt, I guess I don’t see what this has to do with non-naturalism. Either we can have a concept that works like R in the relevant way (viz., that for maximally specific tuples it’s a conceptual truth that either NEC R(t) or NEC ~R(t)), or we can’t. It seems pretty obvious that we can, since we do. So shouldn’t a non-naturalist just say, it’s built into our concept and needs no metaphysical explanation? (And of course, if it turns out it’s not possible to have a concept like that, again there’s no special problem for a non-naturalist.)
Sorry, Jamie. I was too cryptic. The test is just the usual one: When the tribe disagrees with us about the use of the word, do we take that to show that they mean something different (or that they’re confused)? I’m thinking that someone who calls the Iliad-world “metaphysically possible” (because he can see no reason why things could not have been that way) might mean *exactly* what we do by the technical term and needn’t be confused — unlike someone who insists on calling married men ‘bachelors’. That’s not to say he’s right; it’s just to say that his mistake isn’t conceptual.
I say this in part because I like an analysis of metaphysical possibility — Kit Fine’s— which together with non-naturalism pretty much entails the falsity of supervenience. The analysis says that a world is metaphysically possible iff it’s consistent with the essential truths, and non-naturalism (pretty much) says that Iliad-world and our world are both consistent with the essential truths. So given this package, both worlds are metaphysically possible.
All I claim for this package now is that it’s not ruled out on conceptual grounds.
Matt,
Here’s how I think of this. There’s a sense of “possible” in which it turns out to be impossible for water to be an element. Call that “metaphysical possibility”. There may also be a sense of “possible” in which it’s impossible for a world just like ours in all non-normative respects to be a world in which glory is to be pursued for its own sake. Call this “normative possibility”. Iliad-world is normatively impossible but conceptually possible. The question is whether it’s metaphysically possible — i.e., whether it has the same metaphysical modal status as (say) “There are two cats on the mat”. I say that if non-naturalism is true, then Iliad-world is metaphysically possible and supervenience in the usual sense is false. (As I say in my response to Jamie, I say this because I like an analysis of metaphysical possibility that has this result. On this account, metaphysical possibility and normative possibility are very different beasts: a proposition is metaphysically possible iff it’s consistent with the essence of things; normatively possible (at w) iff it’s consistent with the most fundamental normative principles (of w) . Iliad-world is metaphysically possible but normatively impossible at our world in these senses.)
But if that’s your analysis, then it’s pretty clear we are using slightly different concepts. Unless you mean a real analysis, which is more obscure to me in this context than it is in other contexts.
Gideon, I’m working on a paper about Enoch’s explanation of supervenience (which I take to be pretty similar to Scanlon’s, using somewhat different terminology and lacking the direct focus on reasons), and in it I push that for explanation to work we need moral principles to be responsible for the distribution of moral properties. In order for this to work, we need to accept a particular picture of moral properties and moral principles. Moral properties would have to be (certain kinds of) relational properties between moral principles and whatever is the supervenience base. This would be an essential truth about moral properties on this picture. Add in (motivated) essential truths about the principles to elaborate the other side of the mechanics, probably including that the principles have their contents essentially, and we have a non-naturalism that doesn’t just allow for supervenience, but requires it–even on your preferred use of possibility.
Now, I don’t think that non-naturalism is forced to follow this story. If a non-naturalist wants to deny supervenience (or is motivated to deny supervenience), she can reject many parts of my story. But this shows that non-naturalism plus Finean possibility don’t entail the falsity of supervenience. (Also, is that paper in the pipes somewhere?)
Jamie, regarding your comment above to the effect that conceptual commitments can rule out the need for metaphysical explanation. I’m not convinced. I’m skeptical that this ever happens, but I’m most concerned when the conceptual truth has implications for relations between entities in the world. So, I don’t think that a conceptual truth about supervenience rules out the need for metphysical explanation. But this will apply more broadly. What do you think about the following case (inspired by you killing-dying case):
There is a general conceptual truth that all causes have effects. There is a specific question for each event of causation, how the cause causes its effect. But just because we have the general conceptual truth about all causes having effects, doesn’t mean that there isn’t an important and distinct general metaphysical question that remains to be answered. The question is “how/why do causes have effects?” Or “how does causation work?” Or “what is causation?” The answer to these questions is a theory of causation: it may be that causation works through activation of powers, or causation may be counterfactually robust regularities. The conceptual truth that causes have effects doesn’t answer this question for us, so much as it raises it. The conceptual truth entails that there is some way in which causation works, but it doesn’t tell us how (I hear Mark’s voice in my head telling me to qualify this: “if there is causation”). Then, once we have an answer to this general metaphysical question, we are able to give answers to the questions about specific instances. This cause had this effect because the appropriate power was activated; this cause had this effect because they were an instance of the right kind of counterfactually robust regularity.
I the same should be said for explaining supervenience. The conceptual truth that moral properties supervene on something raises but does not answer the metaphysical question “why would moral properties supervene on something/anything?” Then, when we have an answer to this, we can give answers about the specific instances, why this moral property supervenes on the natural property that it does
(yikes: html fail!)
No, I didn’t say that conceptual commitments can rule out the need for metaphysical explanation. On the contrary, I said there is a metaphysical relation that I don’t think non-naturalists have an explanation for. SS has two boxes. The first is conceptual, and it will not get a metaphysical explanation. The second is metaphysical and if (I really should say “since”, pace Rosen) SS is true, that second one needs a metaphysical explanation.
The example of killings and deaths is supposed to make essentially the same point that you want to make with causes and effects.
Does this tag close the italics?
Jamie, I was trying to say that even if there is a conceptual general supervenience claim, where the first box is conceptual, there is also a metaphysical general supervenience claim (this one is entailed by the conceptual claim plus existence) where both boxes require metaphysical explanation.
In the analogy, the claims of theory of causation are not conceptual claims. Humean theories offer one metaphysical account of causation, and anti-humean accounts offer a different one. But neither theory (should be) offered at the level of conceptual truth. The debate on the nature of causation is a substantive metaphysical one, not a mere conceptual confusion. The debate is over how to explain (or understand, or account for) the first box, metaphysically construed. “What is it metaphysically that makes it necessary that all causes have effects?” I take this to be another way of asking “what is…” or “how does causation work?” The second box comes in when we wonder about how any particular cause causally necessitates its effect. “How did this killing have the effect of this person dying?” This is not the subject matter of a theory of causation, but the answer will be informed by it.
So, I think there are three separate questions, two of which are metaphysical. I take you to deny that the general metaphysical supervenience question is even a question, not merely that we don’t need to answer it. Is that your view?
(sorry for breaking the italics, everybody)
That general metaphysical causal question should read “what is it metaphysically that makes it necessary that all causes necessitate some effect?” Now we have both boxes in there explicitly: the metaphysical one first, then the causal one.
Aaron: try to go back your initial comment with italics and close them there, then repost. I would do it but I don’t have access to the right computer (I’m out of country).
Jamie, looking back at what you said before, I realize that I was half talking past you because I misunderstood your claims above.
I was taking as a given that nonnaturalists can explain the specific supervenience question (why wrongness supervenes on the properties it does), given a suitable elaboration of Enoch’s proposal for example. So, I misunderstood that you were suggesting that NN can’t explain *that* part of supervenience.
What I am suggesting is that there are parallel conceptual and metaphysical general supervenience questions to answer, where general questions are asking about the first box. This means I take the explanatory burden to be larger than you do (which is why I was only *half* talking past you).
Another analogy: it’s a conceptual truth that bachelorhood supervenes on conjunctions of being unmarried and being a man (say). Is there a metaphysical explanation on the same level of generality? Yes: bachelorhood (the property) is a complex property constituted by the properties being unmarried and being a man. Since constituted properties supervene on their constituent properties, and since constitution (a relation between properties, not concepts) is a metaphysical relation, what I said is a metaphysical explanation of that supervenience fact–the metaphysical supervenience fact entailed by the conceptual supervenience fact.
David, I don’t know how to go back and edit a comment. But I would correct the close italics command if I can do so.
Hi Aaron,
I would feel better about your explanation about bachelorhood if I were more comfortable with the idea of property constitution. I think some naturalists and some non-naturalists believe that moral properties are constituted by natural ones. If natural constitution is common ground between naturalists and non-naturalists, I don’t feel like I have a good grip on what it is.
But probably we don’t disagree about much. What do you think about a view that just denies strong supervenience, but endorses weak supervenience? With no operator for metaphysical necessity, is there still a metaphysical explanation required?
Jamie, I take your point about property constitution being uncomfortable to some. But I wasn’t trying to suggest that non-naturalists explain supervenience via constitution by natural properties (that would make them natural in my book). This was merely an analogy to another case of a conceptual truth of supervenience not being explanatorily sufficient.
But I should have presented it as a dilemma. If we take a sparse theory of properties that doesn’t include bachelorhood, then the predicate ‘bachelor’ refers to a conjunction of properties (unmarried, man) but not to a third property (bachelorhood). In this case there is no metaphysical explanation for why bachelorhood supervenes on unmarried and man, because there is no such property. So, here we stop at the conceptual explanation. But I don’t think that the non-naturalist can accept this (I actually don’t think any realist can accept this), because they want to have non-natural properties not just non-natural predicates. So, if I can say that we have a disjunction between the conceptual-only story here and the conceptual-and-metaphysical story above, then the non-naturalists have to go with the above, and give a metaphysical explanation.
I don’t have much to say about property constitution, but I don’t think that non-naturalists are forced to say normative properties are composed properties. They just have to be genuine properties that are numerically distinct from their natural supervenience base. (And for the naturalist, general metaphysical supervenience is explained by the general metaphysical truth that normative properties are identical to natural properties–so property identity, not just property constitution can serve as a metaphysical explanation)
I think we also need to do this for supervenience of determinates on determinables, and that it is the metaphysical relation between those kinds (determination?) that is explanatory. Metaphysical relations galore!
I haven’t thought about this would apply to explanations of weak supervenience, but I think in principle we should look for general metaphysical ones (in addition to the specific metaphysical ones). But there aren’t weak supervenience claims that are conceptual truths, are there? I’m thinking that pain might supervene on c-fiber firings in some world, but that wouldn’t be conceptual. And we might explain this by appealing to functional roles being filled in this world only by c-fiber firing. But that looks like weak specific supervenience. When we ask why would pain supervene on anything at all, it seems like we would say because pain is [favorite relation] to what plays the functional role. But *that* is a strong supervenience explanation (whether specific or otherwise), isn’t it?
So, I don’t know.
Well, whenever there is a strong supervenience conceptual truth, there is a weak supervenience conceptual truth, for sure (because the latter follows logically from the former). I guess you want an example of conceptual weak supervenience without strong supervenience. Gideon thinks the moral/natural case is like that (right?). Or here’s a nice Lewisian one: It is a conceptual truth that being Lonely (that is, being the only object in the world) and its complement property supervene weakly on intrinsic properties, but clearly not strongly.
I think Russ Shafer-Landau might be a naturalist in your book. Doesn’t he say that moral properties are wholly constituted by natural ones?
Gideon and Jamie,
Thanks for that explanation, Gideon. That helps. So there are these issues: i) whether the best theory of metaphysical possibility is Fine-style essence consistency, ii) what the essences of things are, and iii) whether normative variation is consistent with the essences of things. I don’t think the conceptual truth of metaethical supervenience is going to say the normative is necessitated Fine-style metaphysically. I take it you and Jamie (and I) agree over that.
Jamie wants to use the term ‘metaphysically necessitates’ to express the embedded necessitation in metaethical supervenience. I don’t have strong opinions here, but I don’t think it is built into the conceptual truth that the metaethical necessitation is of a kind with necessitations about water and so forth. The inner necessitation in SS could be a different sort of necessity than the one that holds for those classic cases, so long as it is not conceptual (and I’d want to hear reasons for granulating necessity spaces in one way rather than another). Jamie, would you take issue with that? Or do you think that part of what is conceptually true is that the necessitation is of a kind with classic cases of metaphysical necessitation?
Aaron and Jamie,
I’m not sure that I understand the burden as Aaron does. For me, the issue might be captured by comparing the claims of X and Y. They are both dualists about classes of properties A and B (real, distinct existences). They both say that the A properties necessitate the B properties, but aren’t sure which A properties necessitate which B properties (maybe the Bs are phenomenal properties and the As non-phenomenal properties). X additionally says the existence of a necessitation relation here is a conceptual truth. Y does not; for him it is conceptually open that further investigation could prove that there is no necessary relation between these distinct existences after all.
I think X has more explaining to do than Y. How is it that the concepts of As and Bs guarantee no change in one without change in the other? Shouldn’t the presence of a dependency be up to the world to decide – the nature of the As and Bs and the world beyond our concepts? That’s what Y thinks, and he seems to have the less perplexing position.
In the normative case, X is right, perplexing as the view is. Still, don’t we want to explain how the presence of a dependency is a conceptual truth if we can? Like Blackburn, I think the expressivist has a better explanation of how this is a conceptual truth than does the reason fundamentalist. It is easier because he doesn’t start off by talking about distinct existences and necessary connections. Basically, I think an explanation firstly occurs in a meta-language about the pragmatic function of the discourse, or maybe in the metasemantics about rules of use for normative terms, and then by pragmatic/semantic descent we get to assert SS in the object language. That’s sketchy, but something along those lines.
Matt, an explanation for a conceptual truth should be at the level of concepts, or language. Expressivists do have a good story about how this works, but I don’t see any big obstacle to realists providing one. (We aren’t irrealists about bachelors or planets, after all.) My view is that expressivists do not have a good story at all about the second necessity operator, so they might be in the same boat as the non-naturalists.
I think that second operator is the necessity that goes with “could not but have been”. That is, what’s striking, to me, about the relation between the moral and the natural is that there could not have been a person just like you actually are, JUST like you in all descriptive respects, but much worse. And I am sympathetic to Hare’s view about the status of supervenience: someone who thinks people with longer hair than you are thereby morally better is very bad at ethics, but someone who thinks there could be (or could have been) people like you in every descriptive respect but much worse are just confused.
I can see I’ll have to have more convincing things to say than this. I’ll think about it.
Matt, how much do you think that Y has to explain?
She should explain why the class of A properties necessitates some member or other of the class of B properties.
She should also explain why the particular A property, a1, necessitates the particular B property it does, say b5.
(I think this way of putting things is misleading because ‘necessitates’ means something like ‘implicates’ here, right? But that means it’s epistemological not metaphysical. If wrongness supervenes on causing pain, then wrongness necessitates causing pain. But there being wrongness isn’t metaphysically responsible for there being causing pain. It merely implicates that there is causing pain.)
The first one is general metaphysical supervenience, and the second one is specific supervenience. I think Y has to explain both of these, and my claim was that X has to explain both of these too. X can’t reject the general metaphysical question by saying it’s conceptually entailed.
Maybe X additionally has to explain why our concepts has this entailment, but as Jamie says, this is a phil language issue. The explanation that has to be given about concept creation or concept acquisition. But Y needs one of these stories too. They could even accept the same general story, and disagree about the details regarding A concepts.
Jamie, you’re right that Shafer-Landau says that normative properties are constituted by natural properties. I’ll admit that I don’t have a great understanding of the distinction between non-naturalism and non-reductive naturalism. I took non-reductive naturalism to be a claim about normative concepts not being analytically connected to natural concepts, while still referring to a natural property. Non-naturalism then thinks the normative concepts refer to irreducibly normative properties.
So, the question is what makes something a natural property vs a normative property. Whatever else he says, I would think that S-L would want to say that bachelorhood is natural in a way that wrongness is not. I would want to hear something about what would underlie the difference if they are both properties constituted entirely by natural properties.
Aaron, I think the reason fundamentalist explanation of supervenience appeals to metaphysical (or normative) necessitation. We could talk about grounding if you like. One thing. You say “If wrongness supervenes on causing pain, then wrongness necessitates causing pain.” But that is not true. The necessitation goes the other way – causing pain necessitates wrongness. I agree with you that both X and Y need to explain general and specific supervenience. In the earlier post I was just focusing on what I take to be an additional explanatory burden for X.
I can see no one is on board with me about the realist burden of explaining how the existence of a metaphysical (or normative) necessitation relation (grounding) between distinct existences is a conceptual truth. That makes me sad. One last stab. It looks as fishy to me as saying that the existence of God follows from our concept of Him. It is the concept of a perfect being, and he’d be less perfect if he didn’t exist, so it is the concept of something that exists. I don’t deny that explaining why our concept is like that is a phil language/mind issue; I just think if you get to introduce God that way you have some explaining to do. Whether he exists is up to the world, not our concepts. Similarly, the reason fundamentalist, like everyone else, needs to explain why supervenience is a conceptual truth, and that is a phil language/mind issue. But if s/he thinks we get to introduce grounding relations between distinct existences that way, s/he has some explaining to do; whether there are such relations is up to the world, not my concepts. (Again, that’s not all there is to explain. We’d like to explain the inside necessity (grounding) if we can, both general and specific, as you and Jamie point out.)
Well, bachelors are in the world. We don’t appeal to metaphysics to explain why all of them are unmarried, but the concepts don’t tell us whether all men are married.
Matt,I thought your examples of the As and Bs was in the direction that As supervene on Bs. I still hold that the supervening property necessitates it’s base, but that’s in terms of implication, or epistemological necessitation. If wrongness is around, then we know that causing pain must be around. But you’re right that the metaphysical necessitation does the other way. Causing pain (partially) necessitates wrongness. If you’re reading the ‘necessitates’ in “moral concepts necessitate that moral properties supervene” as ‘grounds,’ then I’ll deny that that claim is true or conceptually entailed. (And I like non-naturalism.) The concepts don’t do anything to the properties they refer to.
But the way you put your second paragraph makes me think that we’re exactly agreeing with one another. The concept of God doesn’t tell us *that* God exists or is a perfect being, it only tells us that *if* God exists it’s a perfect being. This conceptual truth plus the existential assumption (which we grant to non-naturalists when asking them to account for supervenience, so we do so here by analogy) then entails that God is a perfect being. But this conceptual (plus existential) entailment doesn’t let us avoid the metaphysical question of why God is a perfect being. This still needs to be answered. So, we agree. But my suggestion above was that (by analogy) the person who thinks that God exists and is a perfect being BUT DENIES that it’s part of the concept of God that ‘if it exists, its a perfect being’ has to explain the exact same things as the person who accepts the conceptual claim. (And they each owe us a story about how we got this concept for God, and why it entails what it does, but that’s all the phil language stuff.)
This is a limited analogy because I don’t think there are embedded necessities here.
Maybe what I want is a three box sentence: “it’s conceptually necessary that (it’s metaphysically necessary that (natural properties necessitate moral properties)).” The most embedded claim is the specific supervenience claim, the middle one is the general metaphysical supervenience claim, and the first one is the conceptual general supervenience claim.
Jamie, what if I put it this way? Our concept of bachelor entails the metaphysical truth (given existential assumption) that the property bachelorhood supervenes on *some* properties. This metaphysical claim needs explanation. The answer is that bachelorhood is a constituted property, and so it supervenes on the constituent properties. This is the general level (contra to how I put it before–your last comment cleared this up for me). Then we ask the next supervenience question: what specifically does bachelorhood supervene on? The answer will be whatever properties constitute it. Unfortunately, I think none of us knows exactly what properties those are. We know it’s not just male and unmarried (e.g. boys). Adding age doesn’t do it either (e.g. priests). Maybe eligibility does it, but this is likely to be another constituted property, so we make little progress here. FINALLY, we ask the question you just raised, what is it about each person who is a bachelor that gives them the constitutive properties? That’s not a metaphysical question (probably), but it’s the only one since the conceptual claim that isn’t.
Now, replace “bachelorhood” with “wrongness” and “constitution” with whatever metaphysical relation between wrongness and the natural properties you think grounds the one going where the other one goes, and we can ask all the same questions.
Why does our concept “wrong” entail that wrongness supervenes on something?
Why (metphysically) does wrongness supervene on anything?
Why (metphysically) does wrongness supervene on what it does?
Why (??) is this action wrong?