This marks the 3rd of eleven “meetings” of our virtual reading group on Derek Parfit’s Climbing the Mountain—see here for further details. Next week, we will discuss Chapter 5 of the latest version of the manuscript–the June 7th version–which can be found here.
Recall that, according to Parfit, one of the most practically significant questions in ethics is (1).
(1) Can we often have most, or decisive, reason to act wrongly?
For if it turns out that we often have decisive reason to act wrongly, morality would, according to Parfit, lose much of its practical significance. Answering (1), however, requires answering (2) and (3) for any given decision.
(2) What ought we to do?
(3) What have we most reason to do?
At this point in the manuscript, Parfit has made what he thinks is sufficient progress into answering (3). The remainder of the manuscript is devoted to answering (2). Chapter 4, "Possible Consent," offers and defends one criterion of wrongness. Since, for Parfit, an act is right iff it is not wrong, the criterion doubles as a criterion of rightness.
Parfit’s calls his criterion the "Principle of Possible Rational Consent," or the "Consent Principle" for short. Fully stated, the principle is the following:
Consent Principle: It is wrong to treat people in any way to which they could not rationally consent in the act-affecting sense, if these people knew the relevant facts, and we gave them the power to choose how we treat them (p. 82).
People can "consent in the act-affecting sense," only if they know that those to whom they are giving consent will treat them in some way only if they consent (p. 79). If people know all the relevant facts, they can rationally give such consent to some act if they would have sufficient reason to act that way, i.e., if they would have reasons to act that way which are no weaker than any reasons that they may have to act in any other possible way (p. 83). Parfit thinks that the Consent Principle is the accurate gloss on one part of Kant’s Formula of Humanity: treat all rational beings, or persons, never merely as a means, but always as ends (pp. 77-81).
The Consent Principle is an acceptable criterion of rightness if it is in itself plausible and has plausible implications. (I take it that Parfit thinks the criterion must also not have many implausible implications.) One might think that the Consent Principle is implausible on the grounds that it conflicts with two other plausible theses:
(4) Some desire-based theory of reasons is true;
(5) Our intuitive reaction to the Earthquake thought experiment (p. 83) informs us that we are morally required to save Blue’s life even at the expense of failing to save Grey’s leg.
For, if we assume (4), Grey could not rationally consent to our failing to save her leg, and thus the Consent Principle must be false. However, Parfit has a strong defense from this objection, because he has already rejected (4) (and hence, does not need to reject the Principle) and provided a plausible argument in favor of a wide value-based theory (Chapters 1-3). Moreover, if we adopt a wide value-based theory, and if we accept the plausible "unanimity condition" (6), which implies (7), we have a good argument (pp. 85-86) in favor of the Consent Principle.
(6) There is always at least one possible act to which everyone would have sufficient reasons to consent
(7) Whenever someone could not rationally consent to one act, there must be some facts that give this person decisive reasons to refuse to consent to it. These facts provide moral objections to this act.
(8) These objections must be stronger than the objections to any possible act, or acts, to which everyone could rationally consent.
(9) Whenever there are stronger moral objections to one of two acts, this act is wrong
Therefore,
Consent Principle (abridged): It is wrong to act in any way to which anyone could not rationally consent
Thus, the Consent Principle is in itself plausible. The Principle also has plausible implications, since it accurately predicts our intuitive reactions to some Earthquake and Lifeboat thought experiments (pp. 83-85), and since it prohibits many acts that are most clearly wrong, e.g. "killing, injuring, coercing, deceiving, stealing, and promise-breaking," acts many of which "treat people in ways to which they would not have sufficient reasons to consent" (p. 86).
Parfit spends the remainder of the chapter defending the Consent Principle from four possible objections.
Objection 1: Though the Consent Principle may be a reliable criterion of wrongness (and, hence, rightness), it fails to explain what makes acts wrong and is, therefore, superfluous. According to this objection, what is morally important is not the fact that people could not rationally consent to some act, but the facts that give people decisive reasons for failing to consent. Thus, although the Consent Principle may be a reliable criterion of wrongness, it fails to describe a fact that explains the wrongness of acts. The Principle is therefore superfluous. Parfit concedes that the Consent Principle may not always explain what makes an act wrong, but it often can, especially when a choice must be made between many acts that would affect many people. In such cases, there may often be only one possible act to which everyone could rationally consent, and this fact might give us decisive reason to act this way. If so, this fact would help to explain why the other possible acts would be wrong. Thus, the Consent Principle is not always superfluous.
Objection 2: The Consent Principle, together with plausible principle (10), implies the obviously false Veto Principle; therefore, the Consent Principle is itself false.
Consent Principle: It is wrong to treat people in any way to which they could not rationally consent.
(10) No one could rationally consent to being treated in any way to which they either do or would refuse consent.
Veto Principle: It is wrong to treat people in any way to which they either do or would refuse consent.
Parfit provides a variation of the Earthquake thought experiment (p. 90) to show that (10) is actually false. What is significant in this variation of the thought experiment is that Grey (i) has some reason to give irreversible consent, thereby restricting her future freedom, and (ii) Grey does not later learn facts that might have given her decisive reasons to regret that she gave her irreversible consent. According to Parfit, similar conditions are often met, and so (10) is false: it is possible to give one’s rational consent to being treated in a way to which one either does or would refuse consent. Thus, we have no reason as of yet to reject the Consent Principle.
I don’t think I completely understand the third possible objection. I think it’s something like the following.
Objection 3: There is no adequate explanation of why the Consent Principle does not sometimes require that people perform wrong acts; therefore the Principle ought to be rejected. The Consent Principle supplies only a sufficient condition for an act’s being wrong. Thus, some acts may be wrong for reasons other than that someone could not rationally consent to them. And Parfit argues that the Consent Principle could never require such acts:
(11) The Consent Principle requires some act only when one or more people would not have sufficient reasons to consent to our failing to act in this way.
(12) If some act would be wrong for other reasons, this act’s wrongness would give everyone a sufficient reason to consent to our failing to act in this way
(13) Therefore, the Consent Principle could never require acts that are wrong for other reasons.
Notice, however, that in trying to argue for/explain (13), i.e., in trying to explain why the Consent Principle does not sometimes require that people perform wrong acts, (12) appeals to the wrongness of acts. The explanation is therefore circular. Hence, we do not have an adequate explanation of why the Consent Principle does not sometimes mistakenly require wrong acts, and the Principle should therefore be rejected. Another way to see this point is to imagine the following conversation:
A: Everyone has a sufficient reason to consent to our failing to φ, so φ-ing is wrong.
B: What is this sufficient reason that everyone has?
A: The sufficient reason is that φ-ing is wrong.
B: You have just "explained" that φ-ing is wrong by saying that φ-ing is wrong. This is circular. So, you haven’t really provided an adequate explanation.
Assuming I’ve got the objection right, I think Parfit responds as follows: in saying that the fact that φ-ing is wrong gives everyone a sufficient reason to consent to our failing to φ, I am not saying that the fact that φ-ing is wrong is the reason that φ-ing is wrong—it is wrong for other reasons; the fact that it is wrong is what gives everyone sufficient reason to consent to our failing to φ, though this is not, in this case, what makes φ-ing wrong. Again, I’m not sure I fully get this, so please feel free to help me understand better the material in this passage (section 14).
Objection 4: The Consent Principle is too demanding and, therefore, should be rejected. Parfit replies that the Principle may indeed be very demanding. He’s not sure that it is too demanding, but if it is, he thinks we can just revise it as follows:
Consent Principle 3: It is wrong for us to treat people in any way to which they would not have sufficient reason to consent, except when, to avoid such an act, we would have to bear too great a burden.
It may be difficult at times determining what is "too great a burden." Parfit suggests, I think, that at such times we will have to rely on good judgment (p. 97).
So, to wrap up: Parfit comes into Chapter 4 having made some progress in determining what we have most reason to do for any given decision, and begins, in this chapter, to determine what we ought to do for any given decision. He proposes a sufficient condition for an act’s being wrong, which he calls the Consent Principle. The Consent Principle, Parfit thinks, is a reliable criterion of, and sometimes an explanation of, wrongness. The Principle is plausible in itself, has plausible implications, and can be defended from a number of objections.
Like last week, I’ll let the précis serve as the means of opening up the discussion, rather than providing a long list of questions. Obviously, though, the main question I need answered is whether I’ve accurately understood what is going on in Section 14. I’m looking forward to reading your comments and questions. I’ll be traveling to
Charlotte
,
NC
over the next couple of days, but I’ll make sure I check in and contribute when I have the opportunity.
Suppose I don’t know the relevant facts. It would not be wrong to tell them to me. But if I already knew all of the relevant facts, I would see no reason to consent to someone’s wasting their time telling them to me again.
Is there some quick fix for this obvious counterexample to the consent principle?
Robert: Perhaps, but I’m not sure it’s the case that you *couldn’t* rationally consent to being told the relevant facts again if you already knew them. (It’s that negative formulation that Parfit’s drawn to.)
I have two questions for now on this (fascinating) chapter. First, in discussing the Lifeboat case, Parfit suggests that White, a single person stranded on one rock, could rationally choose that I, a stranger, leave her to die in order to save five people stranded on another rock. But how could it ever be *rational* for someone to choose that she be left to die (if she’s not in great pain, with a terminal illness, etc.)? Surely any conception of rationality must incorporate the minimal economic sense of the term, but then how could such a choice constitute the most efficient means to my ends? (Perhaps I’m still clinging to the desire-satisfaction theory, though…)
Second,Parfit tries to rebut the objection that the Consent Principle is superfluous, that it’s the underlying facts that give people reasons to consent or not that are doing the work here, not the facts about consent itself. His response is rather odd, insofar as I can understand it. He discusses the possibility that one theory about the relevant underlying facts, Act Utilitarianism, might be coextensive with the Consent Principle, insofar as both principles would require all the same acts (87). In this case, it might be thought that AU is more fundamental and that CP is superfluous. But this is false, he maintains: “there are many utilitarian acts to which some people could not rationally consent, and many non-utilitarian acts to which everyone could rationally consent” (88). But just because AU isn’t completely co-extensive with CP, that doesn’t at all mean that there might not be some complete set of more fundamental facts (either capturable by a single moral theory or simply consisting in a patchwork of various moral theories) that are co-extensive with CP, and that thus render CP superfluous. Any thoughts on this puzzling argument?
I did want to remark, though, that I think the real advance of the chapter over previous attempts to defend something like the CP principle is in Parfit’s (correct) insistence that the principle is merely a *sufficient* criterion of wrongness/rightness. This is what enables him to wriggle it out of a number of standard objections to, for instance, Scanlonian contractualism.
David,
sorry don’t have time to write anything on the first two questions but will try do so later. Quick note on the last point though. In a way Scanlon too seems the contractualist principle only as a sufficient but not a necessary requirement for wrongness. It is after all supposed to describe the central, core area of moral wrongness but not the totality of it. So, I think Scanlon can (and in some places does) accept that there are also other grounds for wrongness – that is wrong acts that are not forbidden by principles no-one can reasonably reject.
The difference, Jussi, seems to be that Parfit’s allowing that the Consent Principle is only sufficient *even within the domain of morality of what we owe to each other,* whereas it seems that Scanlon wants that principle to constitute *the* criterion of wrongness for that entire domain.
David, I think to defend his principle as it is, Parfit would have to tell us what reason there could be for someone who possesses all the facts to consent to being reinformed of them. Offhand, I can’t imagine what that reason would be.
What he needs is to reformulate as something like ‘what a fiduciary with the facts would consent to in the person’s behalf’. Would that that fix have an impact on the rationale for the consent principle?
Dave,
You write,
I don’t think that Parfit is trying to offer an account of any technical conception of rationality, such as the economic one. Rather, he is trying to account for our ordinary conception of rationality. Wouldn’t ordinary folks consider giving one’s consent to save the greater number rationally permissible (though not rationally required)? My intuitions tell me that it’s rationally permissible. Moreover, it seems rational to give my consent to save the same number, as where it is my daughter on the other rock. Would you deny that ordinary folks would think that it is rationally permissible to consent to having one’s daughter saved rather than oneself. If so, then you too think that it could sometimes be *rational*, in the ordinary sense of that term, for someone to choose to be left to die even if she is not in great pain, with a terminal illness.
Dave,
Regarding his response to the objection that CP is superfluous, I don’t think that it matters whether or not we find some other moral principle that is completely co-extensive with CP. Parfit’s response to the objection is, I think, best summed up with these words: “If the Consent Principle is true, this principle would be more, I believe, than a reliable criterion of wrongness. Whenever someone could not rationally consent to being treated in some way, this fact would provide an objection to this act, and would be one of the facts that would make this act wrong [emphasis added]” (88).
Two moral theories or principles can be co-extensive in their deontic verdicts and it still turn out that only one of them gives us the correct account of what the fundamental wrong-making features of acts are. For instance, rule-consequentialism, contractualism, and Kantianism could conceivably all agree on which acts are right and which acts are wrong, but they would still each give a different explanation of what makes these acts right and wrong. This is why I think that Parfit is careful to distinguish between a principle’s being “a reliable criterion” and a principle’s being “explanatory” on p. 87. If CP does this explanatory work, then it isn’t superfluous even if there are other principles that provide equally reliable criteria of wrongness.
Robert,
I don’t understand your objection to CP. You claim to be giving a counter-example to CP. What is it exactly? What is the act that CP implies is wrong which is obviously not wrong?
Consent Principle: It is wrong to treat people in any way to which they could not rationally consent in the act-affecting sense, if these people knew the relevant facts, and we gave them the power to choose how we treat them (p. 82).
This principle must say something like, “X is wrong iff. (or if) were I to know the facts, I could not rationally choose X”. But this is confusing. Suppose you’re wondering whether you can remove my kidney to save Smith’s life. Now the question is whether I *could* rationally consent to that, knowing the facts. Well, sure I *could* rationally consent to it. I could be an extremely generous person who would consent to your taking my kidney. But then it is (at least) not wrong for you to do so. Yikes! Suppose I could be that sort of generous person, but I am not such a person. That is, I could rationally give consent, but I would not give consent.
Mike,
CP is only a sufficient condition for wrongness. So it doesn’t imply that any act is morally permissible (i.e., not wrong), let alone that removing your kidney without your consent is permissible.
Parfit is clear that CP must be supplemented with other moral principles. Just to illustrate, we could accept what Parfit call the “the Rights Principle: Everyone has rights not to be treated in certain ways without their actual consent” (91). If we assume that people have a right not to have their organs removed without their actual consent, we can account for the wrongness of removing your kidney without your consent by supplementing CP with the rights principle.
Mike,
To be precise, I should have said that CP doesn’t imply that any act is morally permissible except by implying that all other alternatives are wrong.
The counterexample is giving the relevant facts to someone who needs them. That’s not wrong. But if she knew the relevant facts, she would have no reason to consent to being given them and therefore couldn’t rationally consent.
Robert,
I’m sorry, but I still don’t follow. Maybe I’m just not thinking straight this morning. Why couldn’t someone rationally consent to being given the relevant facts that he or she needs to know? You say, “But if she knew the relevant facts, she would have no reason to consent to being given them and therefore couldn’t rationally consent.” But presumably if she “needs” these facts, as you say is the case in your putative counter-example, then she doesn’t know them. So the antecedent of your conditional is false and we can’t infer the consequent: that it couldn’t be rational for her to consent to being given the relevant facts.
Doug,
I don’t think I made that mistake. The counterexample is supposed to show it should be wrong *by CP* that you remove the kidney. But it is not wrong *by CP* that you remove the kidney. I was urging that the principle should replace *could be consented to* with *would be consented to*.
But let me reverse the counterexample. Suppose I say “I have all of the information and I consent to having you remove my kidney”. According to CP, it is still wrong to remove it. Why? Well, because, given that same information, I *could* rationally fail to consent to it. The same goes–I would hope–for any supererogatory action.
I had two main worries about this chapter. First, I worried that “people” in CP could refer to any person, including the agent herself. In that case, CP would have all sorts of counter-intuitive implications. It would, for instance, imply that imprudent acts were morally wrong. To illustrate, suppose that I act so as to avoid minor suffering now, knowing that doing so will result in my suffering much more in the near future. Could I rationally consent to my being treated this way? Clearly, not. So CP would seem to imply that acting imprudently is wrong. But such acts, although stupid, foolish, and imprudent, are not morally wrong. And there are other sorts of cases where CP would have counter-intuitive results. Suppose that I acted so as to sacrifice my arm to save someone else’s finger. It would seem that I could not rationally consent to my treating myself in this way, so such self-sacrificing instances of altruism would be morally wrong. Again, this seems counter-intuitive. Parfit has since informed me, though, that by “people” he only had “others” in mind. So CP doesn’t have these counter-intuitive implications after all.
Second, I worried that Parfit’s move to CP3 is completely ad hoc. The move seems motivated only out of a concern to avoid counter-intuitive implications in cases like the “Aid Agency” on p. 96. But I have a hard time getting a grip on what exactly constitutes an ad hoc move and why it is morally objectionable. If the proper philosophical methodology is wide reflective equilibrium, then what’s wrong with adjusting one’s principles so as to avoid counter-intuitive implications? Isn’t that what the method demands that we do? Besides, I think that an appeal to the cost to the agent in performing an act can justify performing that an act even if it of a type that it is prima facie (pro tanto) morally wrong. So why shouldn’t it be the same with those acts that others couldn’t rationally consent to.
So, in the end, I’m not so worried. I mention these two only on the off chance that others may have shared my worries.
Robert: Along Doug’s lines… Presumably, there is some act X I’m considering performing, but I don’t have all the relevant facts *related to X*. You have them, and there’s *another* act I could permit, namely, your giving me the relevant facts with respect to act X. Call this second act Y. Whether or not you give me the relevant facts makes no different with respect to the wrongness/rightness of my performing X; what I should do there depends on what I’d consent to *if* I had those facts. And with respect to Y, I have reason to consent to your giving me the relevant facts about X, given that if I knew all the relevant facts *with respect to Y* (viz., that I didn’t know the facts with respect to X and you had them), I’d consent to your giving me the relevant facts *with respect to my deliberation regarding X*.
Mike,
So here is your suggestion:
Mike’s Consent Principle (MCP): It is wrong to treat people in any way to which they would not rationally consent in the act-affecting sense, if these people knew the relevant facts, and we gave them the power to choose how we treat them.
I think MCP is very similar to what Parfit calls the “Choice-Giving Principle” and is, therefore, subject to all the same objections. To use Parfit’s example, wouldn’t your principle imply that it would be wrong for us to give our students low grades. They wouldn’t rationally consent to this if they knew how this would affect their chances of, say, getting into law school. There are other counter-examples given on pp. 79-80.
Regarding your case where you reverse the counterexample and where “I *could* rationally fail to consent to it”: It doesn’t matter what *I* could or couldn’t rationally fail to consent to. As I explain in the comment above, Parfit didn’t mean to include oneself as a possible non-consenter in CP, although the way he formulates CP is very misleading in that case.
Doug, thanks.
I’m a little confused. What do you mean here,
“It doesn’t matter what *I* could or couldn’t rationally fail to consent to. As I explain in the comment above, Parfit didn’t mean to include oneself as a possible non-consenter in CP”
I don’t follow you. The principle says this (my emphasis),
CP: It is wrong to treat people in any way to which THEY could not rationally consent in the act-affecting sense. . .
But I am the person from whom you are considering removing a kidney. I am the person you are proposing to “treat” in a certain kidney-removing way. So, presumably, it is my rational consent that matters, right?
This is where I say, “ok, go ahead and take my kidney”. Call that my ACTUAL CONSENT or AC. The problem is that my AC conflicts with my POSSIBLE RATIONAL NON-CONSENT or PRN. Though I do consent in AC, it is nonetheless true that I could rationally not consent to the removal of my kidney. So my AC conflicts with my PRN. And when there is a conflict between my AC and PRN, Parfit seems to think (if I’m reading this right) that my PRN trumps my AC. It is therefore wrong to remove my kidney. I think that conclusion is bizarre.
I’ve some comments on MCP, but later.
David, there’s a lot of literature that pretty well shows that any account, such as Parfit’s, that contains a conditional in which the antecedent includes such things as ‘knowing the all the facts’, will be open to these counterexamples. They’re all related to Chisholm’s refutation of phenomenalism and Robert Shope’s articles from the’70s. So it’s not worth it to rehash the arguments here. I am surprised, however, that Parfit would not have automatically headed this off by getting rid of the conditional or in some other way trying to work around this. Yet another example of why it is treacherous to propose accounts with conditionals. Philosophers seem to love them, but never learn their lesson.
Mike,
Sorry, I misunderstood your case. So that quote of mine doesn’t make sense, as you rightly point out.
Here’s the source of my confusion, though: You say,
CP implies that X is wrong if someone other than the agent could not rationally consent to it, not that X is wrong if someone could rationally not consent to it. So I don’t see why you think that CP implies that it is wrong to remove Smith’s kidney to save Jones’s life; it doesn’t.
The placement of the “not” makes a big difference. Suppose that Mike’s removing Smith’s kidney to save Jones’s life is something that Smith may rationally either consent to or not consent to. In that case, CP does not imply that Mike’s removal of Smith’s kidney is wrong (nor does it imply that it’s permissible) even though it would be rational for Smith to not consent to it.
Robert,
Could it help Parfit that rationally permitting something does not require being rationally required to permit it? My thought is that there is nothing irrational in allowing someone to give you information you already have, though this doesn’t show you have a reason to encourage them to give you the info. So I could rationally permit someone to tell me what I already know, even if I have no reason to.
I’m worried though that there are other examples of this sort which don’t elude the problem even with that sort of fix for this case.
(For those who don’t know, Robert has a very nice paper on this sort of issue in Phil Quarterly in the late 90s with “Conditional Fallacy” in the title.)
Robert: I’m sure there’s a hefty literature on this stuff (as is nearly always the case), but I guess I still don’t see the problem with respect to Parfit’s particular formulation, especially if, as I suggested, we’re talking about two different acts where the scope of what’s a relevant fact is restricted. But I guess there’s no need to rehearse all your arguments on this if I’m just being blockheaded. And Mark, my initial response to Robert was just along the lines you suggest. I was worried, though, that there might still be a problem that could be generated, and that’s why I tried to articulate the different scope of the relevant facts.
Doug: Regarding the rationality of permitting someone to let me die to save the greater number (assuming they’re all strangers): this at least doesn’t strike me as uncontroversially intuitive. Indeed, my own intuitions run counter to it: I believe there’s a threshhold of what’s rationally permissible, and self-sacrifice (for strangers) goes beyond that threshhold. It would be supererogatory, but I’m not convinced that what’s supererogatory is always rationally permissible (imagine someone who gave away all his money and belongings to the poor, who gave his blood everyday, and so forth). Bringing my daughter into the mix changes things, of course, but I’d want to tell a long-drawn-out metaphysical story for why my special relation to her might justify such an action (you can probably guess the details of the story, so I won’t bore you with it here).
Doug: Regarding your reply to the superfluity objection: I figured the criterion/explanation distinction had something important to do with it, but I, like Dan, was having a hard time figuring that bit out. I’ll have to ponder what you say some more, but I think you’re right.
Dave,
I guess that we just have differing intuitions on what’s rational versus irrational. Intuitively, it seems to me that most anything that doesn’t involve significant risk of harm to oneself without a significant chance of compensatory benefit to oneself or others is rational. (I’m borrowing here from Josh Gert’s characterization of our folk intuitions about rationality, although this is much more imprecise than what he says.) Rationally speaking, we can stare at the ceiling for no reason, act self-interestedly, and act altruistically (provided our loss is no greater than others’ gain). About the only thing it is irrational to do is to put oneself at serious risk of harm for no good reason (where benefiting others to a proportional extent would count as a good reason). It seems to me, then, that the verdict “irrational” is reserved for only the most non-sensical of behavior, behavior for which there is no sufficient reason of any sort — i.e., those for which there are decisive reasons to refrain from engaging in.
I can’t help but think that you are using the term ‘irrational’ in some technical sense, as to mean ‘that which is contrary to what there is most reason to do’. But note that there are plenty of actions (watching TV, walking in the park, staring at the ceiling) that are contrary to what the agent has most reason to do, but which we don’t call irrational.
David,
I should have said that my comment was a rephrasing of yours.
But on a more interesting note, do you really think that there has to be a special relationship to rationalize self sacrifice? Suppose I spend some time thinking about the costs to others of my refusal to sacrifice in this case, and decide that I will make it my project/goal/commitment in this case to sacrifice my interests for theirs. It would perhaps, as you suggest, be morally supererogatory since let’s grant it is not morally or rationally required to do this. But what would be irrational about it? It looks like if we think it is that MLK was irrational to continue his work, knowing it’s likely consequences for himself. And that does not seem like the right description of the case to me.
As and aside: I do think that the I am not using the term ‘rational’ in a way that entails the minimal economic sense. I think that people often miss how controversial the supposedly “minimal” sense is, in that people with more robust conceptions of rationality (such as Kantians) pretty well have to reject it insofar as they think acting on desires we should not have is irrational. I’m using it in roughly the sense that I think Gibbard has highlighted in Wise Choices, the sense in which what is rational makes sense. And it is a matter of substantive argument what makes sense. It could turn out that only what the fans of minimal rationality say is rational, but I don’t think that just falls out of using the word ‘rational’.
So what I am saying about MLK above is that his actions seem to me to make perfectly good sense, though they required great sacrifice of him.
Mark (and David, as a second)
The issue, as I see it, is that it is possible to rationally consent to something only if there is a reason to. And there is no reason to consent to someone’s giving me information that I already possess. It’s a waste of everyone’s precious time. You can assume that there’s no cost to giving information, I suppose. But often enough there is a cost. So given there is *any* cost to giving information, there is no reason to get it again. And so no rational consent.
Robert, you say this,
“The issue, as I see it, is that it is possible to rationally consent to something only if there is a reason to. And there is no reason to consent to someone’s giving me information that I already possess. It’s a waste of everyone’s precious time”
Do you mean that there are no circumstances in which I might have such a reason? Why isn’t there reason to hear information that confirms what I have? Suppose that I will hold the information with greater stability after hearing it from you. Or suppose everything I believe becomes less and less credible to me over time. Over time I wonder whether I heard that right, or I wonder whether I am emphasizing some part of it unjustifiably. Couldn’t there be reasons like this?
Robert,
So, is the following your putative counter-example to CP: CP implies (implausibly?) that it is wrong for you to “waste everyone’s precious time” by “giving me information that I already possess,” because I cannot rationally consent to your giving me information that I already possess? But why is this implausible? Insofar as I see your giving me information that I already possess as doing something bad for no good reason, your wasting my time with useless information for no good reason does seem wrong. Insofar as I see your giving me information that I already possess as possibly good, as Mike is suggesting, it seems that it wouldn’t be wrong, but then in that case I would have sufficient reason to consent to your doing so and so CP wouldn’t imply that it is wrong.
Earlier you had said that something else was the counter-example: “The counterexample is giving the relevant facts to someone who needs them. That’s not wrong.” Dave and I spelled out our reasons for thinking this can’t be a counter-example to CP, but your reply was that there is a whole lot of literature demonstrating “that any account, such as Parfit’s, that contains a conditional in which the antecedent includes such things as ‘knowing the all the facts’, will be open to these counterexamples.” I don’t want you to rehash those arguments, but I would like to know what “these counterexamples” are. I’m still confused on what you take the counterexample to CP to be.
Doug, contrary to popular belief, Robert does not have all relevant information. Is it morally wrong to give him some or all of the relevant information?
If he had all relevant information, he would have no reason to consent to be told the relevant information, so he would not rationally consent. So the conditional analysis entails that it would be wrong to give Robert, as he actually is, the relevant information.
Like many philosophers, you have been misled by the fact that Robert always seems to have all relevant information. I happen to know that most of the time he’s just bluffing.
Jamie, it was worth reading this entire thread just to get to your comment!!
OK, last try. Perhaps I’ve been just too cryptic, since there is no question that Parfit’s CP is open to this objection.
CP says (at least) that if P is aware of the relevant facts (etc.) and cannot rationally consent to S’s doing y, then y is wrong.
Assume y = ”making P aware of the relevant facts”. If P is aware of those relevant facts, then P cannot rationally consent to being given them.So making P aware of those facts is wrong.
If P is already aware of the facts, P cannot rationally consent to being made aware of them because P cannot have a reason to consent to it. P cannot have a reason to consent to it because there is a better use of S’s and P’s time than rehearsing facts already known and appreciated by P.
The fact that, as David suggests, there is some *other* action that is not wrong that would result in the ignorant P’s getting the relevant facts isn’t relevant to explaining why giving the ignorant P the facts is not wrong, yet CP says it is wrong.
The only way to deal with this problem is to alter the account such that we get two different versions of P, the P who possesses all of the facts and consents and the P who does not. Call the first ‘P’ a fiduciary. Then we may give ‘ignorant P’ the facts because if the ‘P who is aware of the facts’ would indeed consent to giving them to ‘P who is ignorant of the facts’ (if the former is, indeed, a ficuciary).
My original question had to do with this last, inevitable, change in the principle. Does the fact that, in the account of wrongness, the relevant consent will have to be in the hands of a fiduciary have an impact on the overall theory? If I can’t get us to see the problem, and the inevitable solution that Parfit will probably have to accept, well, then I guess there’s no point to asking that question.
Shit, Jamie; should have waited to post. Absolutely perfect.
I think that the controversial part of Robert’s argument is this paragraph:
There’s a way, though, to see why he’s right here, but it takes a bit of background about Parfit’s principle to see what one has to say about it first.
So here’s the conditional: if X has all relevant information and we give X the power to choose how we treat them, then they could not rationally consent to being treated in way W.
And the principle is that if the conditional is true, then treating X is way W is wrong.
So X=Robert, who does not have all the relevant information, and W=informing Robert of the relevant information. Robert claims the conditional is true, and hence that W is wrong.
But maybe it can be rational for Robert to hear information he knows repeated. After all, his grandfather was a Southern Baptist preacher, and people like this revel in hearing the same stories repeated!
But I don’t think this can help Parfit. To see why, consider some other objections to the principle first. So think about causing another gratuitous pain. The conditional might be true if X is a masochist, so the conditional had better not be thought of as a necessary truth. Plus, you better be willing to think that the principle shouldn’t be required to rule out causing gratuitous pain to masochists.
But now consider me, who isn’t a masochist. But I could be and if I were, I’d rationally consent. So not only does the principle not rule out causing gratuitous pain to masochists, it doesn’t rule out causing pain to non-masochists who might be masochists.
Surely, however, this is precisely the sort of case that a consent principle ought to rule out. To get it to do so, we would have change the conditional to include in the antecedent a fixing of your nature, dispositions, interests, etc., before asking whether rational consent is possible. Or else Parfit better be happy with a principle with, as far as I can tell, hardly any implications for the wrongness of anything! And once the fixing takes place, the conditional has no resources to avoid Robert’s example. Part of what gets fixed may include a strong aversion to wasting people’s time, including a strong belief that it is wrong to do so. Then concluding that you can’t rationally consent is pretty obvious.
Ok, you do have to go through Jon’s considerations to nail down the objection completely. But it still looks like the fiduciary move is available to Parfit, or, at least I don’t yet see why he can’t adopt that fix to the conditional. Once you fix the nature and knowledge of the fiduciary, it looks as if he can say that it is wrong to treat anyone in a way that that a fiduciary with the relevant facts and nature would not consent to that person being treated.
Robert,
Thanks for your patience. I don’t expect you to continue to try to enlighten me. I’ll probably need to read the relevant literature.
What’s giving me so much trouble, though, is that I can’t imagine any concrete instance of your proposed counter-example schema. I understand the schema. But in CP the “relevant facts” are, I take it, supposed to be the facts pertaining to S’s doing y. Thus CP says that if P is aware of the facts pertaining to S’s doing y and cannot rationally consent to S’s doing y, then y is wrong.
But when y just equals making P aware of the relevant facts (the facts pertaining to S’s doing y), things get very weird. In that case, I’m supposed to imagine a case where y equals making P aware of the facts pertaining to S’s making P aware of the facts pertaining to S’s making P aware of the fact pertaining to…
And I just can’t imagine such an example.
Let y be explaining the details of quantum electrodynamics to P. The relevant facts, then, would include which propositions would be expressed in the course of y and in what order.
Another example: Let y be throwing a surprise party forP.
Mark: On the issue of the rationality of self-sacrifice, a few remarks. First, I don’t see something like the MLK case as analogous to the case Parfit points to in his Lifeboat case, where it’s stipulated that the five on the other rock are all strangers to me, those with whom I bear no special relationship or any significant ties. Presumably, MLK took himself to have a special relationship with those on whose behalf he was working and sacrificing. And in general, a cause to which one has devoted one’s life could easily rationalize sacrifice. But my intuitions draw the line at such sacrifice for the sake of neither a cause, nor a special relationship, but simply for the greater number.
Further, while the Gibbardian gloss on rational as “makes sense” makes sense at first, I wonder how far it can be pushed. For example, how does such a notion help to make sense of the distinction between what’s rationally permissible and what’s rationally required? I can understand it as helping with the scalar notions of more and less rational, but not the former distinction.
What this exchange brings out, though, is a lacuna in Parfit’s account, namely, a specification of what he means by “rational.” In other words, while we get a criterion of rationality, we don’t get a definition of it.
Yes, a surprise party, and let ‘P’ be me!
Here’s a variation on Robert’s counterexample.
Imagine there’s a kind of training that irrational people can undergo in order to become rational (philosophy classes, perhaps). And suppose that George, an irrational person, requests such training. Would it be wrong to give it to him? It seems not. But Parfit’s consent principle implies that it would be wrong. If George were rational, he would not consent to rationality training, because he wouldn’t need it. So it’s not possible for him to rationally consent to such training.
“If George were rational, he would not consent to rationality training, because he wouldn’t need it”
I don’t know. There seem to be all sorts of reasons to deny this. A rational being couldn’t use any help with rationality? A rational person could not have reason to improve or sustain or sharpen his rational edge? Mathematicians cannot consent to mathematical training? I can’t see why.
Similarly in Robert’s example. I can’t see why I could not rationally consent to be offered information that I possess, for all sorts of reasons. Carnap once said that metaphysics might be a good (I think it was) as a kind of music. Maybe I’d like to hear the information that I possess because it has a nice sound to me. But there are just dozens of reasons, it seems to me, apart from this that I could rationally consent to be given this information.
Campbell, exactly right.
So we can assume that CP as it is is completely doomed. Parfit must fix it to avoid the conditional fallacy. The question is how he might do this. My suggestion is that he do this in terms of a fiduciary:
It is wrong to treat P in any way to which an informed version of her, P+, to whom we gave the power to choose how we treat P, could not rationally consent in the act-affecting sense.
The idea is that it is not P’s consent that matters; it’s P+’s consent. If P+ would not consent to our treating P in some way, then it would be wrong to do so.
That solves all of the problems of giving information or improving P’s rational condition. But it raises new questions.
One is that the actual consent of P has dropped out of the picture. Often it seems as if we should have the consent, however ill-informed, of the person in front of us.
Mike,
I think yours are variations on cases that Jon discussed. In the end, Parfit will just have to throw in the towel on this formulation. Think of the objection as a version of the objection to full information accounts of the good. Is it good to have dinner with a friend? (Gibbard’s example.) Well, if I were fully informed, I’d know all about his digestion. Yuk! Fully informed, I’d probably not want to have dinner with him. But it might be good for me to.
Robert, I doubt the fiduciary version really solves all the ‘conditional fallacy’ problems. What are P+’s criteria supposed to be in making her choice?
I think Jamie is exactly right (as usual…?). It’s amazing to me that in spite of the appearance of Shope’s article more than 25 years ago, it is so widely ignored. I don’t think Robert was defending the fiduciary proposal though–I believe he knows that it will fail for related reasons. It helps mask this fact, though, if you refer to the fiduciary as “one’s better self”. That automatically makes the criteria to be used sound like just the right ones to have!
True, the account probably won’t be able to survive. Jon and I are united in thinking conditional analyses don’t work pretty much anywhere for this reason. But this is a shot at survival, and I’m just wondering how the whole thing would play out. So suppose P+’s criteria includes her caring about P’s welfare. Plus all the other stuff. P+ is not consenting to ways of treating P+. P+ is just consenting to ways of treating P, in whose care she is entrusted. Why not?
Robert,
The examples I gave at (June 16, 2006 at 01:15 PM) allude to my post above (June 15, 2006 at 06:13 PM). Between there are several posts that (I hope) I’m not reiterating.
But one favor. Explain to me why would P’ be able to rationally consent to rational instruction on behalf of (rational) P whereas (rational) P would not be able to rationally consent to this on behalf of himself.
My examples are designed to show that P would in fact be able to consent to this on behalf of himself. That he allegedly cannot do this is what I doubt.
Let me offer one other example. In reading Lewis’s ‘Mad Pain and Martian Pain’ (first as a graduate student and then after) I was struck (as many are) by the closing sentence ‘far from ignoring questions of how states feel in the odd cases we have been considering, I have been considering nothing else!’. I know that sentence very well, as I’m sure you do. But even after getting a pretty firm grip on what he had in mind, I still reconsider it. Here’s the point. I have no trouble actually reading again a sentence that I know. I don’t think it is irrational to do so. I think I could perfectly rationally have someone esle read that sentence to me again. I mean I (i.e. P) could rationally do so (not P’). Why an I mistaken about that?
You’re mistaken because Parfit is not interested in the consent of someone who hasn’t appreciated the facts. You haven’t fully appreciated Lewis’ sentence yet.
You can tell bright philosophy students who say they want to go to law school that writing legal briefs is going to be really boring. They say ‘yeah, yeah, I know’. But they don’t really get it. Until they go to law school. Then they’ll admit ‘I hadn’t imagined how boring this is!’ Their decision would’ve been better had they imaginatively engaged with that future boredom a little bit more.
I don’t think that Mike’s wrong that sometimes it might be perfectly rational to consent to hear the things you know one more time. But the problem is that the account needs that to be true whenever it is not wrong to tell you something. And that is harder to believe.
Dave, your answer to my question makes your view much clearer and I see the point, though I do wonder how long it takes to make other people’s well being something you are committed enough to to rationalize sacrifice. FWIW, it matters a lot to my intuitions about cases whether the things I’m committed to seem intrinsically worthwhile or not. Saving lots of people seems quite worthwhile so it seems to me that it is relatively easy to make that one’s project in such a way that it rationalizes sacrifice.
Shame that I missed so much of this interesting debate. But talking about fixes to the CP due to the problems with conditional, doesn’t Parfit’s own CP2 do the job?
CP2: It is wrong knowingly to treat people in any way to which they would not have sufficient reason to consent in the act-affecting sense, if we gave them the power to choose how we treat them.
Here being rational is not under the conditional but only whether we give others the power to choose. The rest is done with reasons to consent which one can have or lack irrespective of whether one is aware of them, rational or irrational. So the agent who is consenting is under the conditional as she is actually – no paradoxes should follow. But maybe I’ve missed something.
Jussi, not entirely ‘as she is actually’. Anything that shows up in the antecedent is going to generate a problem. For instance, CP2, as you state it, contains ‘giving a person the power to choose how we treat them’. But that is surely something we might do, too. Is it wrong to give someone that power? Well, if we had already given them that power, as the antecent says, we could not give it to them. So there would not be a reason to consent. And so it would, according to CP2, be wrong to do so.
Mark,
It won’t be perfectly rational to consent to hear the things you know one more time if you already understand and apprecciate those things as fully as you need to in order for your consent to have the normative status it is supposed to have on Parfit’s account. There would need to be some reason. And once we fix facts about your condition — as, for instance, Jon suggests — there just isn’t any available reason to hear those things again.
Robert,
I wonder. Even if I have been given power to consent to a way someone is going to treat me, have expressed my consent, and am consenting (whatever that is), I don’t see that in this case my reasons for continuing to be consenting would go anywhere.
Robert, this is interesting,
“It won’t be perfectly rational to consent to hear the things you know one more time if you already understand and apprecciate those things as fully as you need to in order for your consent to have the normative status it is supposed to have on Parfit’s account.”
I think it might be perfectly rational because a “fully informed” person, I take it, need not be an “omni-informed” person. If I were omni-informed relative to the brief sentence from Lewis, I would know all of the implications of what he say in that excerpt. But I don’t. So I want to say that (i) I understand the sentence, (ii) I appreciate it as I need to and (iii) I am not omni-informed relative to that sentence. It is because of (iii) that I might rationally consent to review or hear the sentence again.
On the other hand, if Parfit demands that my consent be omni-informed, then there are going to be very few people who meet that standard relative to any sentence.
Mark: You’re right that there are surely threshhold worries on my view, and I confess I haven’t really spent any time thinking about them. My real resistance here, though, is to these sorts of me-or-5-strangers cases, where the *only* reason-giving consideration is supposed to be that there are five of them and only one of me. When it comes to moral deliberation where I’m one of the affected parties (as opposed to third-party consideration of 1-vs.-5 cases), I find it hard to make sense of self-sacrifice *for that reason alone*. You’re right, it might become part of my commitment-set fairly quickly, and when it does it will stem from something I judge to be worthwhile about those five individuals (somehow, from my lonely rock a fair distance away), likely some story I can tell myself about my relation to them (qua fellow humans in need?) but that will render the relevant reason as something significantly more than “well, there’s five of them and only one of me.”
Dear all,
I’ve enjoyed reading the posts in this and previous ‘meetings’. I wonder if people still read this thread but here’s just a thought concerning Robert’s challenge.
Consider a maxim of roughly the following kind ”tell relevant facts to someone who doesn’t know them [i.e. relevant facts]”. It seems everyone could consent to being treated according to this maxim, so acting in accordance with this maxim seems to pass the test of the Consent Principle, and thus is not wrong.
[Consent Principle: It is wrong to treat people in any way to which they could not rationally consent in the act-affecting sense, if these people knew the relevant facts, and we gave them the power to choose how we treat them (p. 82).]
Arto,
that’s not quite the solution. Given how CP is formulated no-one could rationally consent to be treated in accordance to your maxim. Whether one could consent to be treated in accordance to a maxim is tested with counterfactuals where you know all the relevant facts. So, imagine now Arto+ who knew all the relevant facts. Could he consent in that possible situation to be treated with the maxim ‘tell all the relevant facts to someone who does not already know them’? No. Because, by definition, he could not be even treated with this maxim as he knows all the relevant facts. And, if he cannot be even treated with a given maxim, then he cannot possibly consent to be treated with that maxim. As others suggested one way to avoid this problem is to move from the example model to the advice model but even this move is not without its problems.
Jussi,
Good, there may be something funny in the idea that someone is here and now being “treated” in the light of a maxim which says nothing about her in that situation. But let’s see.
Let’s take first a maxim of a slightly different form, “if a person is x, then φ; if a person is not x, then do not φ”. Clearly even people who lack feature x can be treated according to that maxim. Someone could for example resist a temptation to φ, because she realizes that the person in question is not x. Accordingly, people who know all the relevant facts can be treated in the light of the maxim “tell all the relevant facts to people who don’t know them, and don’t bother telling if they already know”. (In their cases, keen informers may have to resist their temptation to recite facts.)
It seems to me that people who lack feature x could also consent to being treated in the light of a maxim of the form “if a person is x, then φ”. Others’ conditional maxims may matter to me, even when I do not meet the condition. That is, I may meaningfully consent to being treated according to the maxim “if a person is x, then φ” even if I happen to lack the feature x. It may matter to me here and now how someone would treat me if I were drowning, even though I am not drowning here and now. And it might matter to me here and now how someone would treat me if I were (born) female, even though I am not (I’m not sure though what to make of that if I’m not (born) female in any possible world – and I have similar doubts about people who by definition know all relevant facts).
So, I’m not (yet) convinced that people cannot consent to being treated in the light of maxims that say nothing about their current situation. The sense in which they are here and now being treated according to that maxim is simply that they are at all times treated in accordance with that maxim, and it seems they could rationally consent to that.
Arto,
you may be right. Not sure I follow all the way. The point though is that it does not save CP that there are many instances where it gives the correct answer. All we need in Parfitian methodology is one counterexample – a false positive where the principle says one thing and our moral intuitions something else. Now, I take it that if we take an agent who goes around indiscriminately telling everyone relevant facts we might think that she’s lost it and doing something very silly but we would not want to think that what she is doing is *morally wrong*. But CP as it stands seem to tell us that what she does on occasion is wrong – she treats others, those who have the relevant facts, in a way they could not rationally consent. What reason would they have to hear the facts again…
Jussi,
OK, good, so one objection would be that the CP gives “thumbs down” to acts that intuitively we’d give “thumbs up” (roughly, giving relevant info to people who need it) – I had that in mind.
Another objection would be that the CP classifies as moral wrongs acts that intuitively we’d give “thumbs down” but would not call moral wrongs, but just silly or crazy etc. There might well be cases like that (I think they would make a good objection), but why not say that it is actually morally wrong to be a nuisance to others (like the indiscriminate guy in your example who does not care whether people need more info or not)?
Why not? I don’t know – maybe it’s my good nature. I know some people who are quite a nuisance to me but I just cannot get myself to say that what they do to annoy me is morally wrong : ) And sometimes even giving me info I already have is involved… You think the first years are wronging you when you mark their exams? I know there is a temptation to answer ‘Yes’ to this question but still in honesty…
Hi Robert. Sorry to have gotten to this great discussion so late. You write,
I, for one, was not aware of the “Conditional Fallacy” nor of the literature on it. So, thank you for bringing it my attention. (Now that I think about it, didn’t you raise this kind of objection to virtue theory a short while back in one of your articles?) At any rate, though I take your main point, I wanted to point out that what you have written here seems too strong. P could certainly have a reason to consent to being given all the relevant information, if consenting to being given all the relevant information would, in some way, benefit the informant to give such information. Parents and teachers, for example, could rationally consent to being given all the relevant information about something (even about quantum electrodynamics) by their kids or students, even if the parents or teachers already knew all the relevant information. For example, perhaps it would please the child to give the information to the parent, or it would benefit the student’s learning to give the information to the teacher. In such cases, giving such information wouldn’t be wrong. So, I think we’d have to revise the objection slightly.
Your main question, though, was whether Parfit could somehow retreat to a principle that referred to some kind fiduciary:
(I take it that, in the last sentence, you meant ‘could’ not consent.) This is an interesting suggestion. If some version of this were to work, I think we’d have to restrict the fiduciary to some kind of ‘fiduciary in rational decision-making only’, since we wouldn’t want P+ to have any special, care-involving relationship with P. Even with this change, however, you would still worry about cases in which it seems important to have P’s actual consent. Since P+ is an informed version of P, could we build in to our description of P+ that P+ is informed about whether P would actually consent?
Thanks for raising these point Robert (and Jamie, and Jon).
Mike A wrote:
Doug responded:
Mike, I think Doug’s response is the right one. The wide-scope of ‘not’ is important in Parfit’s formulation of CP. To stick with Doug’s example, CP implies that it would be wrong for Mike to take Smith’s kidney if Smith could *not* [rationally consent] to Mike’s taking the kidney. It does not imply that it would be wrong for Mike to take Smith’s kidney if Smith could [rationally *not* consent] to it.
Doug, a clarificatory question. You say,
Do you really have the inuition that it can be rational to stare at the ceiling for no reason? I don’t have the same intuition as you here. But, then, I also don’t have the intuition that staring at the ceiling for no reason is irrational. I think it is just one of those “acts” that is neither rational nor irrational. The only way I can make sense of how such an “act” could be rational is by having the intuition that any act is rational just because it is not irrational. Bernie Gert has such a view. For him, the fundamental notion is not rational, but irrational, and an act is rational if and only if it is not irrational. I don’t know whether this is Josh Gert’s view. Is it yours?
In the original post, I said this:
This is wrong. Since CP provides only a sufficient condition for wrongness, CP, by itself, does not double as a criterion of rightness. That is, even if CP did not imply that a certain act would be wrong, that does not mean the act would be right, since it could be wrong according to some other criterion.
Dan,
Yes, I treat ‘rational’ and ‘irrational’ as logical contradictories on a par with ‘morally permissible’ and ‘morally impermissible’. So, by ‘rational’, I mean ‘not irrational’. In that case, then, it would seem that we share the same ‘intuitions’ and differ only in terminology.
I’ve been assuming that Parfit has been using ‘rational’ and ‘irrational’ as logical contradictories, as I do. Do you think that he means to be using them as logical contraries, as you do?
Hi Doug. Well, we share the same intuition that staring at the ceiling for no reason is not irrational. We don’t share the same intuition that doing so would thereby be rational. I don’t really have any sense of where Parfit would stand on this issue. I’ll ask him when I write to him shortly. I’d love to hear what everyone else’s intuitions are about this case. Does anyone else have the intuition that staring at the ceiling for no reason would be rational just because it is not irrational, or do you have the intuition that doing so would be neither rational nor irrational?
I don’t think Parfit would need to say that staring at the wall if rational. Didn’t he say something like that a desire to do something (and doing it too at the same token) is rational if you have a belief which, in the case it was true, was such that the believed fact would be a reason for the given desire or act. In this case, my belief that there is a wall or that the wall is of such and such colour would not be a belief about a fact that was providing me a reason to desire to stare the wall or to do so. And, I don’t think this would need to make him say that it is irrational either. Maybe for the desire to be irrational, the fact corresponding to the belief would need to be a consideration counting against desiring the thing or acting in some way. I’m sure there is something about this in Chap. 2 but my memory fails me here. I’ll try to check this.
Hi Jussi. You’re right that, for Parfit, a desire, and act based on that desire, is rational iff you have beliefs whose truth would give reasons to desire or act that way (for desires or acts based only on nonnormative beliefs). So, I think you’re right that Parfit would not have to say that staring at the ceiling for no reason is rational just because it is not irrational. I’m curious, though, about what you think? Doug has the intuition that staring at the ceiling for no reason is rational. Do you?