Cuneo emphasizes certain similarities between ethics and epistemology, features that they share. There are also, without a doubt, differences between the two, and Cuneo mentions some of them.
It doesn't matter, for his core argument, if there are differences between ethics and epistemology as long as these differences are irrelevant. What do I mean be irrelevant? Two things. First, the differences can't play a role in any of the anti-realist arguments. It can't be that a difference between ethics and epistemology can explain why we would make an anti-realist argument in ethics but not in epistemology. That would sink the ship. Second, it can't be that a difference between ethics and epistemology overrides the anti-realist concerns, such that it's a positive reason to believe in epistemic realism.
So let's start cataloging some of the differences. I think Cuneo, to his credit, gets a lot of them.
DIFFERENCES BETWEEN ETHICS AND EPISTEMOLOGY:
1. (Cuneo p.98) "Moral motivation is more intimately connected with feelings of guilt than epistemic ones" or along the same line of thought "We do not typically claim that epistemic obligations are overriding, nor do we claim that failing to conform to them warrants guilt."
2. (Cuneo p.106) "The type of argument under consideration is one that concerns the explanatory role of moral and epistemic facts. It does not, however, deny that there may be interesting differences between the explanatory roles of moral concepts, on the one hand, and epistemic concepts, on the other. It might be, for example, that epistemic concepts are more entrenched in our best science and metaphysics than moral ones."
3. Even if virtue ethics/epistemology is right and far more things are evaluated then it typically thought in each realm, ethics primarily evaluates actions. Epistemology, on the other hand, primarily evaluates beliefs. This evaluation of beliefs is second order in a way that typical ethics (evaluating ethics) is not.
4. Consequentialism is far less attractive in epistemology than it is in ethics. People argue for it, but it seems far less attractive to say that you're justified in believing one false thing in order to believe 5 true things, than it is to say that you're choosing the course that will cause you to save more lives. This is a matter of debate, though.
5. On a related note, ethics regularly evaluates states of affairs that have nothing to do with humans or actors. We might say that war is an evil, that an operation is a good thing or that a certain weapon is an evil thing, or that the current healthcare system is morally problematic because it causes suffering, though we blame no particular actor. All of these could probably be reduced to claims about actors (when we morally evaluate the country's healthcare situation negatively we're really evaluating the performance of the country as a corporate actor). I'm not sure if this is a real difference, or if all these cases can be eliminated without trouble. But consequentialists, apparently, don't think that these things can be eliminated. Because they analyze what it means for an actor to do the right or wrong thing in terms of whether a situation of state of affairs is a good or bad one. So apparently there are states of affairs that are good or bad independent of actors. But is there any way to think of evaluating a situation or a state of affairs in an epistemic way? I'm not sure. Would that be like saying that a historical method is unjustified? Is this not really a difference at all? Maybe not. Hermmm.
Not sure if any of these are relevent (the one about cloesness to science, I think, could probably be interesting when we start to talk about forming an indispensability argument for epistemology, which will be very soon). Need to add to the list, refine the list, think about what the list means.
Thursday, January 7, 2010
Cuneo's 6 Arguments against realism: II
Now, onto the argument from queerness.
From Prof. Selim Berker's lecture notes on metaethics, I know that there are multiple versions of this argument. Following Prof Berker, here are two different ways to make the argument:
THE MOTIVATIONAL INTERPRETATION
1) If there are objective moral facts, then they are facts that motivate(/have to-be-done-ness built into them/reasons internalism is true)
2) No facts are motivating (Humean picture of motivation)
3) So there are no objective moral facts
THE DUTY INTERPRETATION
1) If there are objective moral facts, then they are facts that place upon us an obligation/restriction/duty/give us a reason to X
2) No facts are such
3) No objective moral facts
OK, OK, my premises 1) weren't really precise because I wasn't careful about what sort of thing brings upon the obligation. Is it knowledge of the fact? Something vaguer, like recognition? Mere belief (which is implausible in the reasons case but better in the motivational case)? I'm leaving that part out. Moving on,
So, Prof Berker, in his lecture, notes, makes a companion in guilt argument for both of these interpretations. For the first, he suggests that regular practical reasoning would be just as queer for the motivational interpretation. After all, we believe that there is a fact of the matter about what's in my best interest, and wouldn't that be just as queer? There's room to push back on Prof Berker. If one's unwilling to accept moral realism that makes best-interest realism unapealing. Say we were moral nihlists. When we consider what's in our best interest we usually discount options like "kill this guy and take his wallet" because doing so would be horribly wrong. Maybe morality is implicitly part of our consideration of what's in our best interests.
In any event, Prof Berker suggests that in the reasons interpretation of the argument epistemology will be a companion in guilt. Because, say what you will about epistemology, if you think it's a normative domain then it gives you reasons to believe (at the very least), and if it's queer that there are facts that give you reasons to act (at the very least) then how is that any less queer than a reason to believe? A reason's the weird thing, no?
Cuneo presents a motivational interpretation of the argument from queerness, but then suggests that epistemology is a proper companion in guilt to ethics. But since he's doing the motivational interpretation of the argument he can't simply appeal to the fact that epistemology gives one a reason to believe something. Motivation is motivation to act, and so he has to show that epistemology gives one a reason to act. He convinced me that earlier that virtue epistemology is right, that we evaluate actions epistemologically as well. Following that, he says "Consider the fact that Sam's maintaining a high level of confidence in a proposition is unwarranted because his doing so is intellectually foolish. If facts such as these were to exist, then they would seem to have a motivational magnetism simliar to that of moral facts."
What's wrong with this? Well, if we're simply talking about motivations (and not reasons) it seems that ethics has a much different claim on motivation than epistemology does. In Cuneo's terminology (which he has plenty of!) this is summed up by saying that morality has normative force, while epistemology doesn't. That is, "We do not typically claim that epistemic obligations are overriding, for example. Nor do we claim that failing to conform to them warrants guilt." Why is this a concern? Cuneo is attempting to say that the weirdness of ethics is shared by epistemology. But it seems that epistemology doesn't motivate quite in the way that ethics does. Perhaps we should say that epistemology isn't motivating, that the queerness belongs only to ethics. But I wonder if Cuneo isn't on better ground then this. Isn't epistemology motivating, even if it's not an overriding consideration, as it is with ethics? The objection only seems to stick if we think that epistemology doesn't motivate at all. Essentially, Cuneo's responses are ways of saying that ethics might be a stronger or different kind of motivation, but that doesn't mean that epistemology isn't motivating. But it is worth taking stock of an important difference between ethics and epistemology--the normative force of ethics is lacking in epistemology. Whether this pulls the two apart in any significant way is to be seen.
Personally, I'm feeling sympathetic to Brink's arguments that motivation internalism is right, in which case the better argument is the one presented by Garner, the reasons interpretation of the queerness argument. If that's the case though, as Prof Berker pointed out, epistemology seems like an even better companion in guilt since there are epistemic reasons as well.
In short: I'm sympathetic to Cuneo's first two arguments. The normative nature of ethics seems to be behind the supervenience objection, as well as the argument from queerness. Epistemology is also a normative domain, so I don't have any objections to Cuneo's claim of having found a companion in guilt.
From Prof. Selim Berker's lecture notes on metaethics, I know that there are multiple versions of this argument. Following Prof Berker, here are two different ways to make the argument:
THE MOTIVATIONAL INTERPRETATION
1) If there are objective moral facts, then they are facts that motivate(/have to-be-done-ness built into them/reasons internalism is true)
2) No facts are motivating (Humean picture of motivation)
3) So there are no objective moral facts
THE DUTY INTERPRETATION
1) If there are objective moral facts, then they are facts that place upon us an obligation/restriction/duty/give us a reason to X
2) No facts are such
3) No objective moral facts
OK, OK, my premises 1) weren't really precise because I wasn't careful about what sort of thing brings upon the obligation. Is it knowledge of the fact? Something vaguer, like recognition? Mere belief (which is implausible in the reasons case but better in the motivational case)? I'm leaving that part out. Moving on,
So, Prof Berker, in his lecture, notes, makes a companion in guilt argument for both of these interpretations. For the first, he suggests that regular practical reasoning would be just as queer for the motivational interpretation. After all, we believe that there is a fact of the matter about what's in my best interest, and wouldn't that be just as queer? There's room to push back on Prof Berker. If one's unwilling to accept moral realism that makes best-interest realism unapealing. Say we were moral nihlists. When we consider what's in our best interest we usually discount options like "kill this guy and take his wallet" because doing so would be horribly wrong. Maybe morality is implicitly part of our consideration of what's in our best interests.
In any event, Prof Berker suggests that in the reasons interpretation of the argument epistemology will be a companion in guilt. Because, say what you will about epistemology, if you think it's a normative domain then it gives you reasons to believe (at the very least), and if it's queer that there are facts that give you reasons to act (at the very least) then how is that any less queer than a reason to believe? A reason's the weird thing, no?
Cuneo presents a motivational interpretation of the argument from queerness, but then suggests that epistemology is a proper companion in guilt to ethics. But since he's doing the motivational interpretation of the argument he can't simply appeal to the fact that epistemology gives one a reason to believe something. Motivation is motivation to act, and so he has to show that epistemology gives one a reason to act. He convinced me that earlier that virtue epistemology is right, that we evaluate actions epistemologically as well. Following that, he says "Consider the fact that Sam's maintaining a high level of confidence in a proposition is unwarranted because his doing so is intellectually foolish. If facts such as these were to exist, then they would seem to have a motivational magnetism simliar to that of moral facts."
What's wrong with this? Well, if we're simply talking about motivations (and not reasons) it seems that ethics has a much different claim on motivation than epistemology does. In Cuneo's terminology (which he has plenty of!) this is summed up by saying that morality has normative force, while epistemology doesn't. That is, "We do not typically claim that epistemic obligations are overriding, for example. Nor do we claim that failing to conform to them warrants guilt." Why is this a concern? Cuneo is attempting to say that the weirdness of ethics is shared by epistemology. But it seems that epistemology doesn't motivate quite in the way that ethics does. Perhaps we should say that epistemology isn't motivating, that the queerness belongs only to ethics. But I wonder if Cuneo isn't on better ground then this. Isn't epistemology motivating, even if it's not an overriding consideration, as it is with ethics? The objection only seems to stick if we think that epistemology doesn't motivate at all. Essentially, Cuneo's responses are ways of saying that ethics might be a stronger or different kind of motivation, but that doesn't mean that epistemology isn't motivating. But it is worth taking stock of an important difference between ethics and epistemology--the normative force of ethics is lacking in epistemology. Whether this pulls the two apart in any significant way is to be seen.
Personally, I'm feeling sympathetic to Brink's arguments that motivation internalism is right, in which case the better argument is the one presented by Garner, the reasons interpretation of the queerness argument. If that's the case though, as Prof Berker pointed out, epistemology seems like an even better companion in guilt since there are epistemic reasons as well.
In short: I'm sympathetic to Cuneo's first two arguments. The normative nature of ethics seems to be behind the supervenience objection, as well as the argument from queerness. Epistemology is also a normative domain, so I don't have any objections to Cuneo's claim of having found a companion in guilt.
Cuneo's 6 Arguments against realism: I
As far as I can tell, the heart of Cuneo's argument...well, that's not quite right. Let's try that again. As far as I can tell, Cuneo's argument has two hearts. One of the hearts is that anti-realism in epistemology is untenable. The second heart is that the arguments against realism in ethics apply just as well to realism in epistemology. He considers 6 arguments against moral realism, and shows how they extend into epistemology. I'll now write 6 posts, one going through each argument and trying to see how well it holds up. And, cuz I'm an undergrad and need practice with this sort of thing, I'll also recap the argument for antirealism Cuneo's dealing with. (I'm not sure I've got much interesting to say on all of them, including the first one below).
The first argument he considers is Blackburn's argument from supervenience. In short, Blackburn's argument is another way of trying to be precise about what's weird about ethical realism. It's generally believed that non-ethical propositions can't entail ethical ones, on the conceptual level. That is, is doesn't imply ought. Now that there is lack of entailment of ethical propositions from the plain old natural ones, does this mean that ethics is an independent realm of mysterious stuff that has nothing to do with the natural world? No, of course not. If two situations are identical naturally, then they are identical morally. And the only way for there to be a moral difference between two situations is if there is also some natural difference which underlies the moral difference. Let's call this relationship between ethics and the natural features of the world a supervenience relation. So ethics supervenes on the natural. Further, this supervenience relation seems to be a conceptual, analytic one: if somebody thought that you could get an ethical difference without some corresponding natural difference they would be simply misunderstanding what we mean by the ethical. (I guess that's true).
Blackburn argues that this supervenience relationship is odd. Note, for example, that it's quite unlike chemistry supervening on physics. Physics propositions can entail chemical ones, and this is what makes the supervenience relationship more easy to swallow. In a certain sense, chemistry propositions just are propositions about particles and atoms. That is, chemistry can be reduced to physics. But what we're claiming in ethics is that there is a supervenience relationship even though ethics cannot be reduced to the natural (unless you're a naturalist about ethics in which case they can be reduced!). This relationship seems mysterious. And if you want to cache out this mysterious relationship in a more precise way, we can talk about mixed worlds: since there's a supervenience relation, it's a necessary fact that in a universe, if the natural facts underly some ethical property for some object (they're hitting a baby makes them evil) then whenever any object has those natural properties in this universe they have that ethical property (ANYONE who hits a baby is evil). In addition, since there is lack of entailment on the conceptual level of ethical claims from natural ones, it's possible that in some alternate universe that set of natural properties won't underly an ethical property (in some universe hitting a baby doesn't make you evil). But how can we explain the fact that it's forbidden, by the supervenience relation, for there to be some universe where sometimes hitting a baby makes you evil, but sometimes it doesn't? Blackburn wants an explanation of this, and doesn't think it's possible for the realist to give a good one.
Now, what's doing the work in this argument? (Incidentally, "doing the work" is one of my least favorite philosophy phrases, though I have trouble dispensing with it). That is, in merit of what aspects of the nature of ethics does this argument work? What feature of ethics gets ethics into trouble? There's nothing too strange about a supervenience relation, but what is strange is the supervenience relation with the lack of entailment. So what is responsible for the lack of entailment? In short, it's the normative character of ethics; you can't get "ought" from "is."
So I don't have much to quibble with when Cuneo suggests that epistemology has the same problem. As long as epistemology is normative, then it too will have lack of entailment. And so it will suffer from the same problems that ethics does.
Of course, the problem for realism isn't just that something weird needs to be explained, but rather that no good explanation seems forthcoming for the realist. Is there any reason to suppose that there will be a difference for epistemic realism? The epistemic projectivist would still have available the answer that expressing an epistemic attitude is just expressing an attitude in response to natural features, and so I'm not sure I have much to add here. It's the bigness of the natural realm that makes it so hard to conceptually doubt the supervenience relation in ethics, and I think that as long as epistemology isn't natural then you have the same problem. I'm moving on, though I reserve the right to come back and update the post if I think of something better to say.
UPDATE: Well, I might as well add some stuff later about attempts to argue against the supervenience relation in epistemology by Lehrer (Cuneo writes "But if Lehrer is right, then it will also be true that moral facts do not supervene on non-moral facts"). Note also that we have to be careful here about whether we're talking about supervenience on non-moral/epistemic or on the non-normative/natural facts. There can be differences there. Also, Blackburn argues that the mental/physical and other supervenience relations don't supervene conceptually, and it's worth trying to explain more carefully why the epistemic will be conceptually supervening while the mental doesn't. I'll do that later.
The first argument he considers is Blackburn's argument from supervenience. In short, Blackburn's argument is another way of trying to be precise about what's weird about ethical realism. It's generally believed that non-ethical propositions can't entail ethical ones, on the conceptual level. That is, is doesn't imply ought. Now that there is lack of entailment of ethical propositions from the plain old natural ones, does this mean that ethics is an independent realm of mysterious stuff that has nothing to do with the natural world? No, of course not. If two situations are identical naturally, then they are identical morally. And the only way for there to be a moral difference between two situations is if there is also some natural difference which underlies the moral difference. Let's call this relationship between ethics and the natural features of the world a supervenience relation. So ethics supervenes on the natural. Further, this supervenience relation seems to be a conceptual, analytic one: if somebody thought that you could get an ethical difference without some corresponding natural difference they would be simply misunderstanding what we mean by the ethical. (I guess that's true).
Blackburn argues that this supervenience relationship is odd. Note, for example, that it's quite unlike chemistry supervening on physics. Physics propositions can entail chemical ones, and this is what makes the supervenience relationship more easy to swallow. In a certain sense, chemistry propositions just are propositions about particles and atoms. That is, chemistry can be reduced to physics. But what we're claiming in ethics is that there is a supervenience relationship even though ethics cannot be reduced to the natural (unless you're a naturalist about ethics in which case they can be reduced!). This relationship seems mysterious. And if you want to cache out this mysterious relationship in a more precise way, we can talk about mixed worlds: since there's a supervenience relation, it's a necessary fact that in a universe, if the natural facts underly some ethical property for some object (they're hitting a baby makes them evil) then whenever any object has those natural properties in this universe they have that ethical property (ANYONE who hits a baby is evil). In addition, since there is lack of entailment on the conceptual level of ethical claims from natural ones, it's possible that in some alternate universe that set of natural properties won't underly an ethical property (in some universe hitting a baby doesn't make you evil). But how can we explain the fact that it's forbidden, by the supervenience relation, for there to be some universe where sometimes hitting a baby makes you evil, but sometimes it doesn't? Blackburn wants an explanation of this, and doesn't think it's possible for the realist to give a good one.
Now, what's doing the work in this argument? (Incidentally, "doing the work" is one of my least favorite philosophy phrases, though I have trouble dispensing with it). That is, in merit of what aspects of the nature of ethics does this argument work? What feature of ethics gets ethics into trouble? There's nothing too strange about a supervenience relation, but what is strange is the supervenience relation with the lack of entailment. So what is responsible for the lack of entailment? In short, it's the normative character of ethics; you can't get "ought" from "is."
So I don't have much to quibble with when Cuneo suggests that epistemology has the same problem. As long as epistemology is normative, then it too will have lack of entailment. And so it will suffer from the same problems that ethics does.
Of course, the problem for realism isn't just that something weird needs to be explained, but rather that no good explanation seems forthcoming for the realist. Is there any reason to suppose that there will be a difference for epistemic realism? The epistemic projectivist would still have available the answer that expressing an epistemic attitude is just expressing an attitude in response to natural features, and so I'm not sure I have much to add here. It's the bigness of the natural realm that makes it so hard to conceptually doubt the supervenience relation in ethics, and I think that as long as epistemology isn't natural then you have the same problem. I'm moving on, though I reserve the right to come back and update the post if I think of something better to say.
UPDATE: Well, I might as well add some stuff later about attempts to argue against the supervenience relation in epistemology by Lehrer (Cuneo writes "But if Lehrer is right, then it will also be true that moral facts do not supervene on non-moral facts"). Note also that we have to be careful here about whether we're talking about supervenience on non-moral/epistemic or on the non-normative/natural facts. There can be differences there. Also, Blackburn argues that the mental/physical and other supervenience relations don't supervene conceptually, and it's worth trying to explain more carefully why the epistemic will be conceptually supervening while the mental doesn't. I'll do that later.
Wednesday, January 6, 2010
Quick thoughts on why ethics and math would be difficult
I started this project looking at similarities between ethics and math, and now I'm looking at similarities between ethics and epistemology. Why the switch? I think I'm able to explain more clearly why I shifted focus.
Let's say you have some problematic discourse, X. It has problematic features and properties, and these problematic features have lead some folks to suggest that this discourse X is not real--doesn't describe or pick out anything out in the world. How can one respond to these concerns?
I can think of at least two ways (I'd be really excited if I could think of more). One way is to directly engage with the "problematic features" and attempt to show that they are not problematic, or that they are not features of X. For the latter you have to argue that X doesn't have the features ("There's no widespread disagreement in ethics"). For the former you could directly argue that these properties are simply unproblematic ("So what if there's widespread disagreement in ethics?! Who cares?"). But another way to answer in the former way is to cheat a bit. You can cheat by finding companions in guilt: "X may contain these features, but so do a bunch of other unproblematic, real discourses, and so these features are not problematic." Companion in guilt arguments can be convincing, but in order to be convincing eventually they have to come around to explaining why the features are not problematic. It's insufficient to simply find a companion in guilt--because maybe what we've learned is that this other discourse Y really should be problematic, just as X is. True, we're more confident in Y than we are in X, but in order to be really convincing we need to learn how come these features aren't problematic.
I want to make one more distinction. The following distinction isn't a really tight one, because you can easily collapse it, but I think what I mean will be clear. There's a distinction between finding a companion in guilt between two objects that shares a problematic feature, and two arguments that share a problematic feature. On the one hand we can try to find other objects that are queer in the sense that ethical objects are, but we can also try to find arguments that are problematic in the same way that arguments that justify ethics are problematic.
When I started, I was looking at math, and specifically the Quine-Putnam indispensability argument for mathematical realism. I was, essentially, doing the second of these two options, looking for a companion in guilt for an ethical argument in a mathematical argument. I wasn't looking for a companion in guilt for ethical objects in mathematical objects. I think the search for an analog of the indispensability argument for ethics didn't really go anywhere, and here's why. The indispensability argument in math is strong, if it is strong, because it stays close to something like IBE. The argument, in other words, is strong because it doesn't invoke much beyond IBE. This, however, means that if I'm looking for an argument in ethics that I can find a companion in guilt in math for, I need to restrict myself to IBE and not much more. This lead me to Sturgeon and Ethical Naturalism at first. That didn't seem interesting to me. Then it lead me to Enoch.
Another way that I could've gone was to look at mathematical facts or statements and attempt to show that these objects are themselves companions in guilt to ethical statements. Justin Clarke-Doane has been working on this for disagreement, arguing that if disagreement is problematic than it's not just ethics, but (perhaps even more so) math that is problematic.
So how would I continue if I thought more about math? First would be to try to pivot off of IBE in some other way. Note, however, that thinking about IBE is no longer really thinking about math. But maybe IBE is rich enough a principle that it actually justifies ethics somehow, either directly or by some sort of reflection through what justifies IBE itself. So Enoch and Sturgeon are still possibilities, but they have nothing to do with math (not that there's anything wrong with that! but I'm explaining my change in focus). Rather, IBE is an epistemological principle and that means that if I'm going to evaluate these arguments I need to learn about epistemology.
I could also reexamine the Quine-Putnam argument and try to see if it reaches beyond IBE. Maybe the resources that it needs to justify the reality of mathematical discourse goes beyond IBE? The problem is that to argue in this way is actually to substantially weaken the indispensability argument. What makes the argument strong is that it doesn't reach beyond IBE, that someone who believes in electrons should believe in numbers. Of course, it's possible that belief in electrons reaches beyond IBE, but then this is to weaken scientific realism in order to find a companion in guilt for ethics. Possible, but not what I was trying to do, and not something I think would go over very well (but maybe necessary to try at some point, of course!).
Finally, I could try to look for problematic features that math shares with ethics. Disagreement was mentioned. What else? The thing is that mathematical statements and her putative facts are not normative, at least I don't think that they are. If they were, then that would be a significant insight and would make for interesting analogies with ethics. But that's certainly not obvious. Is math queer? Maybe, but not in the sense that ethics is, because the queerness of ethics comes from normativity. Is ethics irrelevant for explanations? Harman argues "no," and maybe we could contend with that. Does math supervene on natural facts, or any other facts? Presumably not. What ethics and math definitely have in common is that they are not traditionally empirical domains. That may or may not be a helpful connection. I need to think about this some more, maybe there are some similarities that I didn't think of, but I don't think that they share a lot of the problematic features.
Ethics and epistemology, on the other hand, share normativity, and that's really special and important, since a lot of the arguments against ethics become much weaker if there's nothing wrong wit normativity (Blackburn's supervenience argument, I think, takes advantage of a special feature of normativity, queerness more or less trades on that, etc.) This is why I shifted focus--becuase epistemology seems more promising than math in this regard. It's not an excuse for stopping to think about math, but it is an excuse for putting it on the back burner.
Let's say you have some problematic discourse, X. It has problematic features and properties, and these problematic features have lead some folks to suggest that this discourse X is not real--doesn't describe or pick out anything out in the world. How can one respond to these concerns?
I can think of at least two ways (I'd be really excited if I could think of more). One way is to directly engage with the "problematic features" and attempt to show that they are not problematic, or that they are not features of X. For the latter you have to argue that X doesn't have the features ("There's no widespread disagreement in ethics"). For the former you could directly argue that these properties are simply unproblematic ("So what if there's widespread disagreement in ethics?! Who cares?"). But another way to answer in the former way is to cheat a bit. You can cheat by finding companions in guilt: "X may contain these features, but so do a bunch of other unproblematic, real discourses, and so these features are not problematic." Companion in guilt arguments can be convincing, but in order to be convincing eventually they have to come around to explaining why the features are not problematic. It's insufficient to simply find a companion in guilt--because maybe what we've learned is that this other discourse Y really should be problematic, just as X is. True, we're more confident in Y than we are in X, but in order to be really convincing we need to learn how come these features aren't problematic.
I want to make one more distinction. The following distinction isn't a really tight one, because you can easily collapse it, but I think what I mean will be clear. There's a distinction between finding a companion in guilt between two objects that shares a problematic feature, and two arguments that share a problematic feature. On the one hand we can try to find other objects that are queer in the sense that ethical objects are, but we can also try to find arguments that are problematic in the same way that arguments that justify ethics are problematic.
When I started, I was looking at math, and specifically the Quine-Putnam indispensability argument for mathematical realism. I was, essentially, doing the second of these two options, looking for a companion in guilt for an ethical argument in a mathematical argument. I wasn't looking for a companion in guilt for ethical objects in mathematical objects. I think the search for an analog of the indispensability argument for ethics didn't really go anywhere, and here's why. The indispensability argument in math is strong, if it is strong, because it stays close to something like IBE. The argument, in other words, is strong because it doesn't invoke much beyond IBE. This, however, means that if I'm looking for an argument in ethics that I can find a companion in guilt in math for, I need to restrict myself to IBE and not much more. This lead me to Sturgeon and Ethical Naturalism at first. That didn't seem interesting to me. Then it lead me to Enoch.
Another way that I could've gone was to look at mathematical facts or statements and attempt to show that these objects are themselves companions in guilt to ethical statements. Justin Clarke-Doane has been working on this for disagreement, arguing that if disagreement is problematic than it's not just ethics, but (perhaps even more so) math that is problematic.
So how would I continue if I thought more about math? First would be to try to pivot off of IBE in some other way. Note, however, that thinking about IBE is no longer really thinking about math. But maybe IBE is rich enough a principle that it actually justifies ethics somehow, either directly or by some sort of reflection through what justifies IBE itself. So Enoch and Sturgeon are still possibilities, but they have nothing to do with math (not that there's anything wrong with that! but I'm explaining my change in focus). Rather, IBE is an epistemological principle and that means that if I'm going to evaluate these arguments I need to learn about epistemology.
I could also reexamine the Quine-Putnam argument and try to see if it reaches beyond IBE. Maybe the resources that it needs to justify the reality of mathematical discourse goes beyond IBE? The problem is that to argue in this way is actually to substantially weaken the indispensability argument. What makes the argument strong is that it doesn't reach beyond IBE, that someone who believes in electrons should believe in numbers. Of course, it's possible that belief in electrons reaches beyond IBE, but then this is to weaken scientific realism in order to find a companion in guilt for ethics. Possible, but not what I was trying to do, and not something I think would go over very well (but maybe necessary to try at some point, of course!).
Finally, I could try to look for problematic features that math shares with ethics. Disagreement was mentioned. What else? The thing is that mathematical statements and her putative facts are not normative, at least I don't think that they are. If they were, then that would be a significant insight and would make for interesting analogies with ethics. But that's certainly not obvious. Is math queer? Maybe, but not in the sense that ethics is, because the queerness of ethics comes from normativity. Is ethics irrelevant for explanations? Harman argues "no," and maybe we could contend with that. Does math supervene on natural facts, or any other facts? Presumably not. What ethics and math definitely have in common is that they are not traditionally empirical domains. That may or may not be a helpful connection. I need to think about this some more, maybe there are some similarities that I didn't think of, but I don't think that they share a lot of the problematic features.
Ethics and epistemology, on the other hand, share normativity, and that's really special and important, since a lot of the arguments against ethics become much weaker if there's nothing wrong wit normativity (Blackburn's supervenience argument, I think, takes advantage of a special feature of normativity, queerness more or less trades on that, etc.) This is why I shifted focus--becuase epistemology seems more promising than math in this regard. It's not an excuse for stopping to think about math, but it is an excuse for putting it on the back burner.
Tuesday, January 5, 2010
Cuneo on what makes ethics and epistemology similar
Cuneo spends some time drawing out the similarities between ethics and epistemology. Here are the similarities that he points to:
1) means that both epistemic and ethical facts "are, imply, or indicate reasons for agents to behave in certain ways regardless of whether these agents care about conducting their behavior in a rational/moral way, whether they belong to a social group of a certain kind, of whether they have entered into social agreements with others." (59) I take it that this is an explication of what it means to say that epistemology is a normative domain.
2) means that facts of both kinds come in the same kinds. Some ethical statements are evaluative ("Xing is good"), while others are deontic ("You ought to perform X"). Likewise epistemic statements are either evaluative ("Believing X is rational") or deontic ("You ought to believe X"). Further, ethical facts are either particular ("That deed is evil") or general ("Murder is wrong"), and the same can be said for epistemic facts which are either particular or general.
3) expresses the idea that "institutions, persons, intentions, actions, propositional attitudes (beliefs, acceptances, inquiries and hopes), character traits, emotions, policies, ways of viewing things, ways of finding things out, and so forth, are all plausibly thought to be subject to moral and epistemic norms."
4) means that "in some cases there is no obvious way to disentangle (ontologically at least) their moral and epistemic dimensions; some failing...appear to be both moral and epistemic failings." For example, "failing to treat the testimony of another with sufficient care and attention."
Are these similarities impressive? Certainly I find the first impressive, but I consider it just an explication of what it means to say that ethics and epistemology are both normative discourses. The second doesn't seem so exciting to me, though I'm having trouble picking out exactly why. I think because it seems to me (WARNING: not an argument) very difficult to imagine a normative domain that could be either deontic but not evaluative, or the other way around. The fourth way I find very impressive, though I'm not sure what to make of it if moral and epistemic domains are quite distinct. After all, it's often difficult to separate the normative and descriptive dimensions of a term. For example, take the word "murder." It doesn't seem to be murder if you kill during war, or kill an evil perpetrator, etc. So part of what it means to murder someone is to do something wrong. So there is significant entanglement between the normative and descriptive domains in these terms. Nevertheless, it's unclear to me what this goes to show.
Let's look at the third similarity. Cuneo writes "it would be incorrect to say that moral and epistemic facts are interestingly disanalogous because epistemic facts concern only 'theoretical' reasons, while moral ones concern only 'practical' reasons." He argues, and I'm convinced, that some moral reasons are theoretical and some epistemic reasons are practical.
I might be missing something, but can there be epistemic evaluations of states of affairs beyond actors? Because there can be moral evaluations not just of actors, but of situations and states of affairs. Is it obvious that the same can be done in epistemology? What is an epistemically good outcome that doesn't have to do with the actor? We can morally evaluate, for example, the goodness of evil of a war. A tax can be bad. Can epistemology evaluate anything outside of an actor? Cuneo says that institutions and policies can be epistemically and morally evaluated. But these are tangled up with actors--institutions, such as countries, companies, boards or clubs, can work together and reach decisions and act as a corporate. And policies can be morally evaluated epistemically because they are the decisions of actors. But what about plain-old boring states of affairs. I think they can be morally evaluated but not epistemically evaluated (what would that even mean?).
Is this important at all? I'll see if I can make the case for this doing some work in the next post, unless I change my mind about the previous few paragraphs.
1) Moral and epistemic reasons are categorical
2) Moral and epistemic facts are structurally isomorphic
3) Moral and epistemic facts normatively appraise objects of the same type
4) It's often the case that it's difficult to disentangle the moral and epistemic parts.
1) means that both epistemic and ethical facts "are, imply, or indicate reasons for agents to behave in certain ways regardless of whether these agents care about conducting their behavior in a rational/moral way, whether they belong to a social group of a certain kind, of whether they have entered into social agreements with others." (59) I take it that this is an explication of what it means to say that epistemology is a normative domain.
2) means that facts of both kinds come in the same kinds. Some ethical statements are evaluative ("Xing is good"), while others are deontic ("You ought to perform X"). Likewise epistemic statements are either evaluative ("Believing X is rational") or deontic ("You ought to believe X"). Further, ethical facts are either particular ("That deed is evil") or general ("Murder is wrong"), and the same can be said for epistemic facts which are either particular or general.
3) expresses the idea that "institutions, persons, intentions, actions, propositional attitudes (beliefs, acceptances, inquiries and hopes), character traits, emotions, policies, ways of viewing things, ways of finding things out, and so forth, are all plausibly thought to be subject to moral and epistemic norms."
4) means that "in some cases there is no obvious way to disentangle (ontologically at least) their moral and epistemic dimensions; some failing...appear to be both moral and epistemic failings." For example, "failing to treat the testimony of another with sufficient care and attention."
Are these similarities impressive? Certainly I find the first impressive, but I consider it just an explication of what it means to say that ethics and epistemology are both normative discourses. The second doesn't seem so exciting to me, though I'm having trouble picking out exactly why. I think because it seems to me (WARNING: not an argument) very difficult to imagine a normative domain that could be either deontic but not evaluative, or the other way around. The fourth way I find very impressive, though I'm not sure what to make of it if moral and epistemic domains are quite distinct. After all, it's often difficult to separate the normative and descriptive dimensions of a term. For example, take the word "murder." It doesn't seem to be murder if you kill during war, or kill an evil perpetrator, etc. So part of what it means to murder someone is to do something wrong. So there is significant entanglement between the normative and descriptive domains in these terms. Nevertheless, it's unclear to me what this goes to show.
Let's look at the third similarity. Cuneo writes "it would be incorrect to say that moral and epistemic facts are interestingly disanalogous because epistemic facts concern only 'theoretical' reasons, while moral ones concern only 'practical' reasons." He argues, and I'm convinced, that some moral reasons are theoretical and some epistemic reasons are practical.
I might be missing something, but can there be epistemic evaluations of states of affairs beyond actors? Because there can be moral evaluations not just of actors, but of situations and states of affairs. Is it obvious that the same can be done in epistemology? What is an epistemically good outcome that doesn't have to do with the actor? We can morally evaluate, for example, the goodness of evil of a war. A tax can be bad. Can epistemology evaluate anything outside of an actor? Cuneo says that institutions and policies can be epistemically and morally evaluated. But these are tangled up with actors--institutions, such as countries, companies, boards or clubs, can work together and reach decisions and act as a corporate. And policies can be morally evaluated epistemically because they are the decisions of actors. But what about plain-old boring states of affairs. I think they can be morally evaluated but not epistemically evaluated (what would that even mean?).
Is this important at all? I'll see if I can make the case for this doing some work in the next post, unless I change my mind about the previous few paragraphs.
Subscribe to:
Posts (Atom)