This is the other point I saw a disagreement I had with Parfit - it all looked good until the argument against subjectivism. I take it that people don't have a *reason* to want to avoid future agony - rather, part of what makes it *their* future *agony* is that they *do* want to avoid it. The "Future Tuesday Indifference" sounds irrational because it sounds like it is based on a false belief that there is some difference in the pains. But if it's really just based on caring about some more than others, I find it hard to argue that it's irrational, and not just *different* (especially in combination with various formal arguments suggesting that it's impossible to come up with an interpersonal or intertemporal utility scale, without building in some assumption about trade-offs from the beginning).
Interesting! Does it help at all if we build in that (i) they do anticipate / identify with their future-Tuesday self, (ii) they know that, come Tuesday, they will be averse to the agony in the usual way (and deeply regret their past choice not to avoid it), and (iii) they *do* care in the normal way about future *Wednesday* agony, all along (and their future Wednesday-self also predictably regrets having suffered more than necessary on Tuesday)?
I think there's a tension of some sort between identifying with, knowing that they'll be averse to it, and not caring about it now. But I'm not convinced that we can take "identification with" as a primitive - in a sense, I'm thinking that knowing that there's an aversion and not caring about it basically *is* not identifying with it, even if someone makes verbal claims about identifying with it.
I'm a HUGE fan of Parfit and I do recommend everyone read his works so this is a really nice project! I might send a few of the articles to friends who have yet to read Parfit. Excited to see where this series goes!
The future Tuesday indifference idea seems to talk past the subject of its criticism. Hume was skeptical of the existence of any sort of non-instrumental rationality. Rationality always serves an end, but ends themselves, unless they are intermediate ends (hence also means), are not ultimately subject to rational criticism. The FTI idea, to fit in this frame, must either take the painful operation on Tuesday as an end or a means. For a choice to be preferred does not distinguish end from means. That is in the mind of the chooser. If he has an end which the painful operation will fail to satisfy, or will satisfy less well, this is instrumental irrationality, and Parfit fails. If Parfit is criticizing means, he is not contradicting Hume.
By stipulating that there is no reason for choosing the more painful option, Parfit tries to exclude the possibility of instrumental irrationality. But this goes too far. Wanting something is always a reason, though maybe not one that will satisfy others.
For Parfit's idea to work, the pain can’t be a means to something, it must be the end itself. But if we are comparing a painful but useful operation on Tuesday to a painless but otherwise identical operation on some other day, there is an end beyond the pain. Either we are criticizing ends, in which case the pain alone must be the end, or we are criticizing means, but Hume is fine with that. That’s to say, the Tuesday operation can be either an irrational means or an end. To criticize it as an end, it can’t have instrumental value.
I haven’t quite convinced myself. Let's take an even more extreme example. Suppose someone says they wish to experience pain, and that is their only end. They don’t enjoy the prospect, they don’t expect to enjoy the pain, and they don’t expect to gain something later as a result of experiencing the pain. They don’t wish to punish themselves or accomplish anything else by feeling the pain. This seems like an irrational end. But can we generalize from this that any other less extreme or more useful negative experience is irrational?
So maybe I disagree with both Parfit and Hume. One end is capable of being deemed irrational. We always seem to have a reason to avoid pain, so choosing to experience pain without some compensating side effect seems irrational.
But can we say the same about pleasure? Do we always have a reason to choose to experience pleasure, other things equal? Is it irrational to forego costless pleasure?
Maybe I can make one last quibble on behalf of Hume. The etymology of word “rationality” has to do with calculating correctly. But what is being miscalculated in this person's choice of pain? Desire is not the result of a calculation that can be mistaken. You either desire something or you don’t. As Hume said, if a desire is based on a mistaken fact, when the mistake is revealed the desire disappears. It was the fact that was mistaken, not the desire. We may not understand why people want what they do, but it doesn’t seem that we can claim that they don’t really want it.
I seem to remember Bernard Gert had a good discussion of rationality in his big book on morality. I don’t think he agreed with Hume. He had a list of evils, pain among them, that he thought is would be irrational to accept without compensation. I don’t remember if he got into the issue of instrumental rationality vs. other.
I find these arguments very unpersuasive: the argument against egoism just fails, whereas the argument about subjectivism is good but doesn’t establish anything important. The rational egoist values their own welfare but not that of others. They care about their future pain because it is _their_ pain. We might say that their evaluation of pain is depends on who has it, but not when. And the major argumentative move here is a complete non-sequiter: the felt badness of pain IS dependent on who feels it! When someone else is in pain, I don’t feel anything. So there is a clear asymmetry. What a terrible argument!
The case against subjectivism is less obviously bad, but all it establishes is that, barring some further work, a subjectivist should accept that a moral agent could have preferences that don’t hang together well — are ‘irrational’ in a certain sense. But that doesn’t establish that subjectivism is false or implausible. Subjectivism is an account of what morals are: it isn’t committed to the idea that no possible moral agent could have silly preferences. It’s not hard to offer an account of why actual moral agents would be expected to care about their future selves!
You seem confused. "The felt badness of pain" refers to an intrinsic qualitative property of the pain itself; it's not a claim about *who* is feeling it. (Obviously an egoist doesn't care about the felt badness of pain when it's someone else doing the feeling. But that's just to say that they aren't responding to the intrinsic badness of pain, but only to their *relation* to it. As you say, "When someone else is in pain, I don’t feel anything." But it's equally true that when your future self is in pain, your momentary self doesn't feel anything. So the egoist has no non-question-begging basis for criticizing the presentist—who cares only about their *present* self—as irrational.)
Note that the argument is NOT that any person who has the concerns of the egoist must change their concerns. It's that the thesis of rational egoism—according to which egoism is the ONLY rationally permissible pattern of concern—is not a principled, well-supported thesis about the scope of rationality. A more principled view would either be more broadly permissive (incl. permitting narrowly presentist concerns) or else require broader concern (as in forms of rational altruism).
N.B. "Subjectivism" (like "internalism") can mean different things in different contexts. Here I am not talking about the "account of what morals are", but "the view that we have reason to do whatever will fulfill our ultimate (non-instrumental) desires".
Ah, that’s interesting that the argument isn’t actually that egoism is irrational, only that the view which says that egoism is obligatory is irrational. I didn’t realize the scope was intended to be so limited.
It seems to me that the argument fails with even the limited scope. If, as you say, the egoist is asserting that one ought to care about one’s own pain (that is only the intrinsic bads they will have a personal relation to), there is still also an asymmetry between the later time version of their self and the later time version of a different person. The asymmetry is that they will have the relation they care (experiencing the pain) about to the future pain of themself, but not to the other person’s pain. It’s just false to say that the egoist has no question begging basis for criticizing the presentist: their basis is exactly that they can rationally anticipate that they WILL feel their own future pains and WON’T feel others’. There is nothing in rational egoism that prevents them anticipating future events. In the case of another person’s future pain, they know they will never experience, which is what they care about. In the case of their own future pain, they know they will. Therefore the fact that they care about the pain they experience gives them a clear rational basis for caring about their own future pains but not those of others.
Just for fun:
P1: (Egoism) I have reason to care about all and only the pains I experience.
P2: I will experience my future pains.
C1: I have reason to care about my future pains.
P3: I will not experience the future pains of other people.
C2: I have no reason to care about the future pains of other people.
I'm enjoying this. In a somewhat-unrelated note, I use your IEP article on utilitarianism with my business ethics students. One thing that I've always wondered about is the four-part definition you give at the start. It seems to preclude sub-types of utilitarianism, which I also think fit to teach (rule, cooperative, etc.) Can you peel back the curtain on that particular choice?
We can criticize desires without declaring them irrational, can’t we? It is one thing to want something, another to act. I can refrain from acting without having to declare my non-enacted impulses irrational.
Excited to read this, though I was already convinced by Parfit's arguments. Is the pre-print's text and page numbering identical to the published version?
Pagination is very different, and there might be some slight wording tweaks from the copyediting. If you need access to the published version for citation purposes, feel free to shoot me an email.
This is the other point I saw a disagreement I had with Parfit - it all looked good until the argument against subjectivism. I take it that people don't have a *reason* to want to avoid future agony - rather, part of what makes it *their* future *agony* is that they *do* want to avoid it. The "Future Tuesday Indifference" sounds irrational because it sounds like it is based on a false belief that there is some difference in the pains. But if it's really just based on caring about some more than others, I find it hard to argue that it's irrational, and not just *different* (especially in combination with various formal arguments suggesting that it's impossible to come up with an interpersonal or intertemporal utility scale, without building in some assumption about trade-offs from the beginning).
Interesting! Does it help at all if we build in that (i) they do anticipate / identify with their future-Tuesday self, (ii) they know that, come Tuesday, they will be averse to the agony in the usual way (and deeply regret their past choice not to avoid it), and (iii) they *do* care in the normal way about future *Wednesday* agony, all along (and their future Wednesday-self also predictably regrets having suffered more than necessary on Tuesday)?
I think there's a tension of some sort between identifying with, knowing that they'll be averse to it, and not caring about it now. But I'm not convinced that we can take "identification with" as a primitive - in a sense, I'm thinking that knowing that there's an aversion and not caring about it basically *is* not identifying with it, even if someone makes verbal claims about identifying with it.
I'm a HUGE fan of Parfit and I do recommend everyone read his works so this is a really nice project! I might send a few of the articles to friends who have yet to read Parfit. Excited to see where this series goes!
I'm starting to like this Parfit guy...
The future Tuesday indifference idea seems to talk past the subject of its criticism. Hume was skeptical of the existence of any sort of non-instrumental rationality. Rationality always serves an end, but ends themselves, unless they are intermediate ends (hence also means), are not ultimately subject to rational criticism. The FTI idea, to fit in this frame, must either take the painful operation on Tuesday as an end or a means. For a choice to be preferred does not distinguish end from means. That is in the mind of the chooser. If he has an end which the painful operation will fail to satisfy, or will satisfy less well, this is instrumental irrationality, and Parfit fails. If Parfit is criticizing means, he is not contradicting Hume.
By stipulating that there is no reason for choosing the more painful option, Parfit tries to exclude the possibility of instrumental irrationality. But this goes too far. Wanting something is always a reason, though maybe not one that will satisfy others.
For Parfit's idea to work, the pain can’t be a means to something, it must be the end itself. But if we are comparing a painful but useful operation on Tuesday to a painless but otherwise identical operation on some other day, there is an end beyond the pain. Either we are criticizing ends, in which case the pain alone must be the end, or we are criticizing means, but Hume is fine with that. That’s to say, the Tuesday operation can be either an irrational means or an end. To criticize it as an end, it can’t have instrumental value.
I haven’t quite convinced myself. Let's take an even more extreme example. Suppose someone says they wish to experience pain, and that is their only end. They don’t enjoy the prospect, they don’t expect to enjoy the pain, and they don’t expect to gain something later as a result of experiencing the pain. They don’t wish to punish themselves or accomplish anything else by feeling the pain. This seems like an irrational end. But can we generalize from this that any other less extreme or more useful negative experience is irrational?
So maybe I disagree with both Parfit and Hume. One end is capable of being deemed irrational. We always seem to have a reason to avoid pain, so choosing to experience pain without some compensating side effect seems irrational.
But can we say the same about pleasure? Do we always have a reason to choose to experience pleasure, other things equal? Is it irrational to forego costless pleasure?
Maybe I can make one last quibble on behalf of Hume. The etymology of word “rationality” has to do with calculating correctly. But what is being miscalculated in this person's choice of pain? Desire is not the result of a calculation that can be mistaken. You either desire something or you don’t. As Hume said, if a desire is based on a mistaken fact, when the mistake is revealed the desire disappears. It was the fact that was mistaken, not the desire. We may not understand why people want what they do, but it doesn’t seem that we can claim that they don’t really want it.
I seem to remember Bernard Gert had a good discussion of rationality in his big book on morality. I don’t think he agreed with Hume. He had a list of evils, pain among them, that he thought is would be irrational to accept without compensation. I don’t remember if he got into the issue of instrumental rationality vs. other.
I find these arguments very unpersuasive: the argument against egoism just fails, whereas the argument about subjectivism is good but doesn’t establish anything important. The rational egoist values their own welfare but not that of others. They care about their future pain because it is _their_ pain. We might say that their evaluation of pain is depends on who has it, but not when. And the major argumentative move here is a complete non-sequiter: the felt badness of pain IS dependent on who feels it! When someone else is in pain, I don’t feel anything. So there is a clear asymmetry. What a terrible argument!
The case against subjectivism is less obviously bad, but all it establishes is that, barring some further work, a subjectivist should accept that a moral agent could have preferences that don’t hang together well — are ‘irrational’ in a certain sense. But that doesn’t establish that subjectivism is false or implausible. Subjectivism is an account of what morals are: it isn’t committed to the idea that no possible moral agent could have silly preferences. It’s not hard to offer an account of why actual moral agents would be expected to care about their future selves!
You seem confused. "The felt badness of pain" refers to an intrinsic qualitative property of the pain itself; it's not a claim about *who* is feeling it. (Obviously an egoist doesn't care about the felt badness of pain when it's someone else doing the feeling. But that's just to say that they aren't responding to the intrinsic badness of pain, but only to their *relation* to it. As you say, "When someone else is in pain, I don’t feel anything." But it's equally true that when your future self is in pain, your momentary self doesn't feel anything. So the egoist has no non-question-begging basis for criticizing the presentist—who cares only about their *present* self—as irrational.)
Note that the argument is NOT that any person who has the concerns of the egoist must change their concerns. It's that the thesis of rational egoism—according to which egoism is the ONLY rationally permissible pattern of concern—is not a principled, well-supported thesis about the scope of rationality. A more principled view would either be more broadly permissive (incl. permitting narrowly presentist concerns) or else require broader concern (as in forms of rational altruism).
N.B. "Subjectivism" (like "internalism") can mean different things in different contexts. Here I am not talking about the "account of what morals are", but "the view that we have reason to do whatever will fulfill our ultimate (non-instrumental) desires".
Ah, that’s interesting that the argument isn’t actually that egoism is irrational, only that the view which says that egoism is obligatory is irrational. I didn’t realize the scope was intended to be so limited.
It seems to me that the argument fails with even the limited scope. If, as you say, the egoist is asserting that one ought to care about one’s own pain (that is only the intrinsic bads they will have a personal relation to), there is still also an asymmetry between the later time version of their self and the later time version of a different person. The asymmetry is that they will have the relation they care (experiencing the pain) about to the future pain of themself, but not to the other person’s pain. It’s just false to say that the egoist has no question begging basis for criticizing the presentist: their basis is exactly that they can rationally anticipate that they WILL feel their own future pains and WON’T feel others’. There is nothing in rational egoism that prevents them anticipating future events. In the case of another person’s future pain, they know they will never experience, which is what they care about. In the case of their own future pain, they know they will. Therefore the fact that they care about the pain they experience gives them a clear rational basis for caring about their own future pains but not those of others.
Just for fun:
P1: (Egoism) I have reason to care about all and only the pains I experience.
P2: I will experience my future pains.
C1: I have reason to care about my future pains.
P3: I will not experience the future pains of other people.
C2: I have no reason to care about the future pains of other people.
What am I missing?
I'm enjoying this. In a somewhat-unrelated note, I use your IEP article on utilitarianism with my business ethics students. One thing that I've always wondered about is the four-part definition you give at the start. It seems to preclude sub-types of utilitarianism, which I also think fit to teach (rule, cooperative, etc.) Can you peel back the curtain on that particular choice?
We can criticize desires without declaring them irrational, can’t we? It is one thing to want something, another to act. I can refrain from acting without having to declare my non-enacted impulses irrational.
Excited to read this, though I was already convinced by Parfit's arguments. Is the pre-print's text and page numbering identical to the published version?
Pagination is very different, and there might be some slight wording tweaks from the copyediting. If you need access to the published version for citation purposes, feel free to shoot me an email.