Oct 24, 2022·edited Oct 24, 2022Liked by Richard Y Chappell
>"We should be open to considering scenarios in which nihilism turns out to be true..."
I find this paragraph confusing. "Scenario" to me seems to imply something like a "possible world" - but uncertainty about ethics (for a moral realist) seems to me quite different from uncertainty about which of a number of possible worlds we happen to be living in. It is more closely analogous to something like mathematical uncertainty (e.g. uncertainty about whether the Riemann hypothesis is true).
So your proposed "datum" that suffering is bad is something like the "datum" that some integers are prime. You would never say "we should be open to considering scenarios in which no integers are prime," and if you did encounter someone who said this, your response would not be to say "that's an interesting idea, let's explore the implications!", but to ask for clarification about what they think those words mean, since they are obviously using them differently from the way you would use them.
To put it slightly differently: "would suffering still matter, even if suffering didn't matter?" is the moral realist's equivalent of "would you still love me even if i was a worm?" Once you have accepted the question, any possible answer you can give is just going to get you in trouble.
Thanks for this. The realism argument strikes me as phenomenological—I.e. regardless of what’s true, I would be lying to myself if I were to claim suffering wasn’t bad.
Suffering being bad isn't the same as suffering being stance-independently bad. It's not clear to me why having the phenomenology of something being bad would in any way indicate that realism is true. I've tasted bad food before, but I don't take that to be an indication that gastronomic realism is true.
I wasn't sure about the following claim: 'It could be a serious moral error for one’s moral vision to be too restricted, whereas it’s harder to see such grave moral risks to being too self-sacrificing or having an otherwise expansive moral view.' What if realism is true and the true ethical theory features some sort of partiality or agent-relative reasons? Then by doing things like focusing on insect suffering or the welfare of people trillions of years in the future, we could be neglecting our communities or the near and dear, which might turn out to be wrong.
Yeah, there are *possible* risks there, but I think in practice the potential harms to people from refraining from (say) insect farming are unlikely to be all *that* great. Even if it turns out that we ought to do more to exploit others for the sake of our near and dear, it's hard to believe it would be a *grave* moral error to refrain from doing so. Not all wrongs are equal! (And of course many agent-relative reasons are thought to take the form of prerogatives or options rather than requirements.)
That's fair! Though I think the sort of case I was worried about takes the form of an agent neglecting the near and dear in order to benefit (in expectation) faraway moral patients, as opposed to *exploiting* faraway moral patients, as you suggest. E.g., someone who gives away enough of their resources to [insert your favorite consequentialist cause] that they can't pay for senior care for their parents when their parents come to need it. Maybe we'd be tempted to tell a story about how that wouldn't maximize EU in the actual world, given human psychology, signaling, etc., but let's just stipulate a case where acting in this way does maximize EU. Then the worry applies: if the true ethical theory says we have agent-relative duties (not just prerogatives) to care for eg our parents, then having given all our money away to [cause area] will have been the wrong move. So it's not always the case that the consequentialist-friendly move is the morally safe option, dominates other options in cases of ethical uncertainty, etc.
Thanks this is very interesting. I have a question about this bit: "Realists don’t hold that normative properties are what matter; their role is instead to mark which natural properties matter, objectively distinguishing them from the others that don’t."
Your point is that it is suffering (or pleasure) that matters, and not the fact that suffering is bad (pleasure is good). And you say that the property of being bad (good) marks the things that matter.
But don't you also want to say that it is because suffering is bad that something's involving suffering matters? And it is because pleasure is good that something's being pleasant matters.
Those normative properties don't just mark what matters, they explain why those things matter. No?
Yeah, I guess there is a kind of explanatory connection between the normative properties. Being good or bad is a way of mattering, after all. But I'd still be inclined to say, e.g., that bad things matter in a negative way *because* being bad means that it has other (natural) properties that fundamentally matter in a negative way. Such "buck-passing" higher-order explanations go by way of the underlying properties.
Do you think there are explanatory connections here that bypass the underlying properties altogether?
Thanks. I guess I was assuming that mattering in a negative way and being bad were the same normative property. If we distinguish them, then presumably they are necessarily connected.
But my question was really about whether the normative property of mattering in a negative way (or of being bad) does more than just mark out certain things. Isn't it supposed to explain why things that involve suffering are to be avoided/eliminated/etc?
Let's assume that being pinched is always painful. That's a contingent matter, I guess. Suppose we are told that pinching is rarely permitted and Jones asks why. Perhaps Jones does not realise that, as a matter of fact, being pinched is always painful. Pointing this out might satisfy Jones.
But suppose Jones does know this, and still asks why pinching is rarely permitted.
I was assuming that an appeal to the badness of suffering is meant to answer Jones by saying more than was said in merely pointing out that being pinched is painful. Pinching is rarely permitted, the answer would go, BECAUSE being pinched is painful and pain is a form of suffering and suffering matters in a negative way (though this can be outweighed).
Of course, it would not be contingent that pain is a form of suffering nor that suffering matters in a negative way (is bad).
//I’ve long been sympathetic to the argument that one “might as well” be a normative realist, as it can’t very well be a normative fault to believe in normative facts.//
Perhaps there’s more nuance to this, but this seems potentially false on the face of it. If you’re a normative subjectivist it could be inconsistent with your subjective normative standards to endorse normative realism, in which case you could be (by your own lights) at fault for believing in normative facts. But even if you aren’t a normative subjectivist (I’m not a normative subjectivism), I wouldn’t want to be a normative realist unless I thought it served my goals to be a normative realist (and even then, if I didn’t think it was true, I’m not sure I could just opt to believe it). So why should I “might as well” be a realist?
The next remark is where I take more serious issue:
//(Either you’re right, or it doesn’t matter that you’re wrong.) //
This doesn't seem right to me. I’m a normative antirealist, and it does matter to me whether I’m wrong. I have an antirealist conception of things “mattering.” If normative realism is false, it only doesn’t stance-independently matter whether normative realism is false, it doesn’t just not matter, full stop, unless one presumes that things can only “matter” in a realist sense, which would seem a bit question-begging.
It almost looks like there's an equivocation here between normative claims about things mattering and metanormative claims about the way in which they matter (i.e., stance-independently). For instance, it does not follow that if normative realism is false that “nothing matters.” It only fails to matter specifically in the way in which realists think things matter. An antirealist need only deny the latter, not the former.
As far as classical logic goes, any conditional with necessarily false antecedents or necessarily true consequents is necessarily true. If P is necessarily true, then "If not-P, then P" is also necessarily true.
Hence since necessarily suffering matters, then even if suffering did not matter, it would still matter.
Consider a more modest version of the wager. Suppose you're 50% confident in moral realism. Someone offers you 5 dollars if moral realism is true, but if it's not, you'll have to pay 5 dollars and one cent. It seems like you should take the deal.
I disagree with Carlsmith about the wager -- though, given uncertainty, it would depend on the stakes. If suffering isn't really bad, then we're mistaken about its badness properties. This means we have no reason to care about suffering -- in the wager, we have some reason to care about money conditional on realism, but no reason to care about suffering conditional on anti-realism. I think that the reason this seems unintuitive is the sheer obviousness of the badness of suffering. But if our pain is really not bad (somehow!) then it's more analogous to things that only seem bad. If a person thinks that what happens after death matters, but it really doesn't, then that would give no genuine reasons -- the same would be true in this case about pain.
If moral anti-realism is true, nothing matters. Moral anti-realism eliminates genuine reasons, replacing them with merely descriptive claims about desires.
//If moral anti-realism is true, nothing matters.//
This is false. If moral antirealism is true, then nothing *stance-independently* matters.
// Moral anti-realism eliminates genuine reasons, replacing them with merely descriptive claims about desires.//
An antirealist can just conceive of descriptive claims about desires as "genuine reasons," and maintain that only antirealists endorse genuine reasons, while realists like you endorse some kind of confused, magical type of pseudoreason.
Using words like "genuine" isn't going to cut any ice. When you use the term, it's either just a stand-in for "stance-independent," in which case it's a misleading bit of rhetoric to use a term like that. Realists don't get to own words like "genuine", or, it doesn't mean the same thing, in which case it's not clear what you do mean but it's probably an instance of equivocation.
Genuine mattering cannot come from descriptive facts about what we care about. When you say something matters, your just referring to your desires. Well, that's not what I mean by mattering.
You can call them genuine reasons all you want -- they're not genuine reasons. They do not count in favor of things -- they're just descriptive claims about what people want.
Thanks. The statement seems ambiguous to me, so it's hard for me to tell what, exactly, the claim is that's supposed to be obviously true.
Is it just a normative claim, or is there a metaethical presupposition behind the claim that isn't being made explicit? I'm not sure what it means to say I "should" be unconditionally opposed to gratuitous suffering. Do you mean "should" in a realist sense, in an antirealist sense, or in some other sense?
Okay, thanks. I still don't quite know what the claim means, unfortunately. I'm not sure, for instance, how to make sense of it if one is a noncognitivist.
An expressivist might say that they are *for* norms that treat suffering as unconditionally bad. When they imagine themselves engaging in torture -- even when imagining that this version of themselves is subjectively pro-torture -- they are *against* this. Boo torture!
Alright, that makes sense. I suppose it may take a bit of reframing depending on the metaethical account in question but I can imagine this sticking on most views. Maybe not some kinds of antirealism, though, which worries me a bit. A discussion for another time though.
>"We should be open to considering scenarios in which nihilism turns out to be true..."
I find this paragraph confusing. "Scenario" to me seems to imply something like a "possible world" - but uncertainty about ethics (for a moral realist) seems to me quite different from uncertainty about which of a number of possible worlds we happen to be living in. It is more closely analogous to something like mathematical uncertainty (e.g. uncertainty about whether the Riemann hypothesis is true).
So your proposed "datum" that suffering is bad is something like the "datum" that some integers are prime. You would never say "we should be open to considering scenarios in which no integers are prime," and if you did encounter someone who said this, your response would not be to say "that's an interesting idea, let's explore the implications!", but to ask for clarification about what they think those words mean, since they are obviously using them differently from the way you would use them.
To put it slightly differently: "would suffering still matter, even if suffering didn't matter?" is the moral realist's equivalent of "would you still love me even if i was a worm?" Once you have accepted the question, any possible answer you can give is just going to get you in trouble.
Thanks for this. The realism argument strikes me as phenomenological—I.e. regardless of what’s true, I would be lying to myself if I were to claim suffering wasn’t bad.
Suffering being bad isn't the same as suffering being stance-independently bad. It's not clear to me why having the phenomenology of something being bad would in any way indicate that realism is true. I've tasted bad food before, but I don't take that to be an indication that gastronomic realism is true.
Perhaps, I shouldn’t have conflated the two because I agree. By definition, phenomenology does not indicate truth.
I wasn't sure about the following claim: 'It could be a serious moral error for one’s moral vision to be too restricted, whereas it’s harder to see such grave moral risks to being too self-sacrificing or having an otherwise expansive moral view.' What if realism is true and the true ethical theory features some sort of partiality or agent-relative reasons? Then by doing things like focusing on insect suffering or the welfare of people trillions of years in the future, we could be neglecting our communities or the near and dear, which might turn out to be wrong.
Yeah, there are *possible* risks there, but I think in practice the potential harms to people from refraining from (say) insect farming are unlikely to be all *that* great. Even if it turns out that we ought to do more to exploit others for the sake of our near and dear, it's hard to believe it would be a *grave* moral error to refrain from doing so. Not all wrongs are equal! (And of course many agent-relative reasons are thought to take the form of prerogatives or options rather than requirements.)
That's fair! Though I think the sort of case I was worried about takes the form of an agent neglecting the near and dear in order to benefit (in expectation) faraway moral patients, as opposed to *exploiting* faraway moral patients, as you suggest. E.g., someone who gives away enough of their resources to [insert your favorite consequentialist cause] that they can't pay for senior care for their parents when their parents come to need it. Maybe we'd be tempted to tell a story about how that wouldn't maximize EU in the actual world, given human psychology, signaling, etc., but let's just stipulate a case where acting in this way does maximize EU. Then the worry applies: if the true ethical theory says we have agent-relative duties (not just prerogatives) to care for eg our parents, then having given all our money away to [cause area] will have been the wrong move. So it's not always the case that the consequentialist-friendly move is the morally safe option, dominates other options in cases of ethical uncertainty, etc.
Thanks this is very interesting. I have a question about this bit: "Realists don’t hold that normative properties are what matter; their role is instead to mark which natural properties matter, objectively distinguishing them from the others that don’t."
Your point is that it is suffering (or pleasure) that matters, and not the fact that suffering is bad (pleasure is good). And you say that the property of being bad (good) marks the things that matter.
But don't you also want to say that it is because suffering is bad that something's involving suffering matters? And it is because pleasure is good that something's being pleasant matters.
Those normative properties don't just mark what matters, they explain why those things matter. No?
Yeah, I guess there is a kind of explanatory connection between the normative properties. Being good or bad is a way of mattering, after all. But I'd still be inclined to say, e.g., that bad things matter in a negative way *because* being bad means that it has other (natural) properties that fundamentally matter in a negative way. Such "buck-passing" higher-order explanations go by way of the underlying properties.
Do you think there are explanatory connections here that bypass the underlying properties altogether?
Thanks. I guess I was assuming that mattering in a negative way and being bad were the same normative property. If we distinguish them, then presumably they are necessarily connected.
But my question was really about whether the normative property of mattering in a negative way (or of being bad) does more than just mark out certain things. Isn't it supposed to explain why things that involve suffering are to be avoided/eliminated/etc?
Let's assume that being pinched is always painful. That's a contingent matter, I guess. Suppose we are told that pinching is rarely permitted and Jones asks why. Perhaps Jones does not realise that, as a matter of fact, being pinched is always painful. Pointing this out might satisfy Jones.
But suppose Jones does know this, and still asks why pinching is rarely permitted.
I was assuming that an appeal to the badness of suffering is meant to answer Jones by saying more than was said in merely pointing out that being pinched is painful. Pinching is rarely permitted, the answer would go, BECAUSE being pinched is painful and pain is a form of suffering and suffering matters in a negative way (though this can be outweighed).
Of course, it would not be contingent that pain is a form of suffering nor that suffering matters in a negative way (is bad).
//I’ve long been sympathetic to the argument that one “might as well” be a normative realist, as it can’t very well be a normative fault to believe in normative facts.//
Perhaps there’s more nuance to this, but this seems potentially false on the face of it. If you’re a normative subjectivist it could be inconsistent with your subjective normative standards to endorse normative realism, in which case you could be (by your own lights) at fault for believing in normative facts. But even if you aren’t a normative subjectivist (I’m not a normative subjectivism), I wouldn’t want to be a normative realist unless I thought it served my goals to be a normative realist (and even then, if I didn’t think it was true, I’m not sure I could just opt to believe it). So why should I “might as well” be a realist?
The next remark is where I take more serious issue:
//(Either you’re right, or it doesn’t matter that you’re wrong.) //
This doesn't seem right to me. I’m a normative antirealist, and it does matter to me whether I’m wrong. I have an antirealist conception of things “mattering.” If normative realism is false, it only doesn’t stance-independently matter whether normative realism is false, it doesn’t just not matter, full stop, unless one presumes that things can only “matter” in a realist sense, which would seem a bit question-begging.
It almost looks like there's an equivocation here between normative claims about things mattering and metanormative claims about the way in which they matter (i.e., stance-independently). For instance, it does not follow that if normative realism is false that “nothing matters.” It only fails to matter specifically in the way in which realists think things matter. An antirealist need only deny the latter, not the former.
As a matter of logic, the answer is still yes.
As far as classical logic goes, any conditional with necessarily false antecedents or necessarily true consequents is necessarily true. If P is necessarily true, then "If not-P, then P" is also necessarily true.
Hence since necessarily suffering matters, then even if suffering did not matter, it would still matter.
Consider a more modest version of the wager. Suppose you're 50% confident in moral realism. Someone offers you 5 dollars if moral realism is true, but if it's not, you'll have to pay 5 dollars and one cent. It seems like you should take the deal.
I disagree with Carlsmith about the wager -- though, given uncertainty, it would depend on the stakes. If suffering isn't really bad, then we're mistaken about its badness properties. This means we have no reason to care about suffering -- in the wager, we have some reason to care about money conditional on realism, but no reason to care about suffering conditional on anti-realism. I think that the reason this seems unintuitive is the sheer obviousness of the badness of suffering. But if our pain is really not bad (somehow!) then it's more analogous to things that only seem bad. If a person thinks that what happens after death matters, but it really doesn't, then that would give no genuine reasons -- the same would be true in this case about pain.
Why would I take that deal? It's a losing proposition, even if trivially.
Because money only matters if things matter.
I agree. Please explain why I should take the deal.
If moral anti-realism is true, nothing matters. Moral anti-realism eliminates genuine reasons, replacing them with merely descriptive claims about desires.
//If moral anti-realism is true, nothing matters.//
This is false. If moral antirealism is true, then nothing *stance-independently* matters.
// Moral anti-realism eliminates genuine reasons, replacing them with merely descriptive claims about desires.//
An antirealist can just conceive of descriptive claims about desires as "genuine reasons," and maintain that only antirealists endorse genuine reasons, while realists like you endorse some kind of confused, magical type of pseudoreason.
Using words like "genuine" isn't going to cut any ice. When you use the term, it's either just a stand-in for "stance-independent," in which case it's a misleading bit of rhetoric to use a term like that. Realists don't get to own words like "genuine", or, it doesn't mean the same thing, in which case it's not clear what you do mean but it's probably an instance of equivocation.
Genuine mattering cannot come from descriptive facts about what we care about. When you say something matters, your just referring to your desires. Well, that's not what I mean by mattering.
You can call them genuine reasons all you want -- they're not genuine reasons. They do not count in favor of things -- they're just descriptive claims about what people want.
Can you clarify what you mean by a "moral datum"?
A moral claim I take to be obviously true, without need of further argument. That which an adequate theory needs to be able to accommodate.
Thanks. The statement seems ambiguous to me, so it's hard for me to tell what, exactly, the claim is that's supposed to be obviously true.
Is it just a normative claim, or is there a metaethical presupposition behind the claim that isn't being made explicit? I'm not sure what it means to say I "should" be unconditionally opposed to gratuitous suffering. Do you mean "should" in a realist sense, in an antirealist sense, or in some other sense?
Just a normative claim!
Okay, thanks. I still don't quite know what the claim means, unfortunately. I'm not sure, for instance, how to make sense of it if one is a noncognitivist.
An expressivist might say that they are *for* norms that treat suffering as unconditionally bad. When they imagine themselves engaging in torture -- even when imagining that this version of themselves is subjectively pro-torture -- they are *against* this. Boo torture!
See, e.g., https://peasoup.princeton.edu/2014/08/expressivism-subjectivism-and-reasons/
Alright, that makes sense. I suppose it may take a bit of reframing depending on the metaethical account in question but I can imagine this sticking on most views. Maybe not some kinds of antirealism, though, which worries me a bit. A discussion for another time though.