Print This Post Print This Post

1. I don’t especially mind death, but I’m scared of pain. As Epicurus reminds us, death doesn’t hurt. Epicurus may have been mistaken to think that this was enough to show that death isn’t bad, but it is good to know that death is at least not bad in this one important respect.

There are paradoxical sounding remarks by David Velleman that seem to imply that even pain itself, when it is as its worst, may not be bad in this one respect. In his paper ‘A Right to Self-Termination’, Velleman writes

“Pain is a bad thing, of course, but … what justifies death is the unbearableness of the pain rather than the painfulness. What do we mean in calling pain unbearable? What is it not to bear pain? … Not to bear pain is somehow to fall apart in the face of it, to disintegrate as a person. To find pain unbearable is to find it this destructive not just of one’s well-being but of oneself. But then we make a mistake if we describe the patient in unbearable pain as if he were his rational old self, weighing the harm of pain against the benefits of existence. If his pain is truly unbearable, then he isn’t his rational self any longer: he is falling apart in pain. Even if he enjoys some moments of relief and clarity, he is still falling apart diachronically, a temporally scattered person at best.”

So perhaps when pain is at its worst, it isn’t bad either. It is no longer bad because unendurable pain is like death. And death, we’ve just seen, doesn’t hurt!

This should reassure those who believe that there is a hell, and that hell is unbearable pain without end…

Velleman tells us that when a person’s pain is truly unbearable, this person is no longer a rational self. I don’t know if I’ve ever experienced *truly* unbearable pain, but to me this thesis seems true. The question is whether rational selves is what we truly are—not necessarily in a proper metaphysical sense, but in the sense in which the self is the object of self-interest. In this sense, self and person may come apart. I may misunderstand him, but Velleman writes as if this if what he believes. But if this is so, and the rational self has disintegrated (or is it only disintegrating?), then who is left to feel the pain? Is it some other self, or some other part of the self, and if so, what kind of concern should we, qua rational beings, have for it?

2. No doubt, this paradoxical result isn’t something Velleman had intended or hoped for. His argument seems to lead to it through a series of simple steps:

(1) Pain needs a subject
(2) For pain to be bad for me, that subject needs to be me
(3) I’m a rational self
(4) The rational self disintegrates under unbearable pain
(5) Unbearable pain isn’t bad for me

Since (5) is absurd, and (1) and (2) are unremarkable, (3) or (4) must be false.

But you don’t need to be a devout Kantian to think that what you truly are is a rational self—that this is the object of your concern when you care about yourself in a purely self-interested way. And the argument can be run through, not in terms of rationality, but in terms of psychological continuity. Over time, unbearable pain can also erase psychological continuity—at least in its occurrent form. And if psychological continuity is a necessary condition for your continued existence—for the continued existence of what you care about when you care about yourself—then it again seems that unbearable pain couldn’t be pain that you yourself suffer.

Indeed, there are enough philosophers who already endorse a conclusion parallel to (5) on similar grounds. What I have in mind are views about the relation to us of what would survive of our mentality in the later stages of Alzheimer’s disease. By then, there would certainly be no rational self. And psychological continuity and connectedness would be minimal, at best. On a view such as Velleman’s, or many Neo-Lockean accounts of personal identity, the pain that would be felt at this stage would not be our pain, but that pain of a ‘post-person’ that succeeds us. But even some of those who reject this understanding of identity or what matters in survival, still think that our purely self-interested concern for such future pain ought to be pretty minimal. That’s Jeff McMahan’s view, in The Ethics of Killing. I’ll admit that I find this form of argument persuasive.

3. The challenge then, for those like me who accept some version of this claim about Alzheimer’s disease, is to explain how Alzheimer’s differs from unbearable pain. We need to deny (4).

This claim does sound far too strong. Let us, however, first set aside an alternative that is clearly far too weak. It is not plausible to simply deny that unbearable pain compromises rational agency. Even milder pains and pleasure often explain why we don’t do as we think we ought. And a person facing unbearable pain can choose to do what he knows is morally forbidden or prudentially irrational. He may even choose to relieve unbearable pain knowing that this will lead to far greater pain in a few minutes.

A more plausible claim is rather that although unbearable pain seriously compromises rational agency, the rational self nevertheless lingers on. After all, even when subject to unbearable pain, a person maintains some capacity to think and act, however limited and distorted. And his various rational capacities are still somewhere in there, ready to kick back into play the moment pain is reduced or removed.

No doubt, the consciousness of a tortured human being is not identical to the consciousness of a tortured mouse. But I’m inclined to think that the two are nevertheless not that different. Auden makes this point better than I could in his poem ‘Surgical Ward’:

They are and suffer; that is all they do:
A bandage hides the place where each is living,
His knowledge of the world restricted to
A treatment metal instruments are giving.

Being able to deploy concepts and beg for help doesn’t yet make someone a rational self. And there is not much psychological continuity in a subject who, at every moment, is entirely pre-occupied with the present.

To be sure, the subject under great pain still possesses a full range of rational capacities. The subject reduced by Alzheimer’s no longer does. But it’s not quite clear what difference this difference is supposed to make. The unbearable pain, we are supposing, will just go on. The capacities are there, but there will never be a point in time where the subject will return to fully exercise them. Can this difference really carry so much metaphysical and normative weight? It must, but I’m not sure how.

4. Just two further notes.

(1) The paradoxical conclusion is a claim about unbearable pain that is ongoing and interminable. If we were to extend this claim, even in qualified form, to shorter, transient episodes of unbearable pain, the results would be even less palatable. It would be incredible to claim that a person cannot suffer from torture that is terrible enough, or from surgery without anaesthesia. In surgery, what we want is for consciousness to be switched off, not rationality.

(2) Unbearable pain, if it lasts long enough, leaves little of the rational self. But there is nothing in the argument that required that this effect on rationality should be due to pain. There are other ways to disrupt rationality. Suppose that we could give you a pill that would put you in a state of radical confusion. Under the effect of this pill, you would no longer function as a rational agent. It seems to follow that evenless than unbearable pain suffered in this state would not be pain suffered by you, qua rational self.

I end with the unremarkable conclusion that if we are to resist the paradoxical conclusion, and if we are to assess arguments such as Velleman’s, then we need a far better philosophical understanding of what is meant by unbearable pain.


  1. 1. Posted by S. Matthew Liao | June 11, 2007 11:07 pm

    Guy, thanks for this very interesting post! I had a few questions and comments.

    1. First, you said that by self, you do not mean “metaphysical sense.” Then, in your discussion of the Paradoxical Result, you mention “psychological continuity.” Isn’t the notion of “psychological continuity” typically used in the context of discussing the metaphysical sense of self?

    This is important because one can deny (4) if one uses self in the metaphysical sense. In particular, arguably, the metaphysical self would not disintegrate when facing unbearable pain.

    You did go on to say that “unbearable pain can also erase psychological continuity—at least in its occurrent form.” First of all, “psychological continuity” is not the only available account of the metaphysical self – there is, for example, also the Organism View. Secondly, I don’t see how unbearable pain would erase psychological continuity. Take McMahan’s bare-capacity account as an example, you would still exist even when facing unbearable pain, as your consciousness would still be there. Indeed, if a consciousness was not there, how there be an experience of unbearable pain? Or take a mental-state account of psychological continuity, it seems that your mental state would still be there while you experience unbearable pain. Although it can happen, most people do not lose all their memories after an experience of unbearable pain.

    2. Your “solution” to the Paradoxical Result is also to deny (4) by claiming that the self “lingers on.” You argue that “a person facing unbearable pain can choose to do what he knows is morally forbidden or prudentially irrational.” But is this individual really choosing what he would have chosen under normal circumstances? Or is he just choosing to stop the pain? If so, does the “true” self really linger on?

    3. A comment on Velleman’s argument. Imagine a case in which there is self-disintegration as a result of unbearable pleasure. On Velleman’s account, it would appear that this would be just as bad as a case in which there is self-disintegration as a result of unbearable pain. But arguably, common sense tells us that the latter is worse than the former. If this is right, Velleman’s argument misses a crucial aspect of what is bad about unbearable pain, namely, that pain is bad in and of itself.

  2. 2. Posted by Guy Kahane | June 12, 2007 10:55 am

    Matthew, let me try and answer your queries.

    1. Claims about psychological continuity can go into accounts of metaphysical identity, and they can go into accounts of ‘what matters in survival’, or, as this could be more appropriately put, the object of self-concern. As is well known, the two can come apart. To my mind, if metaphysical identity is not the object of self-concern, then it’s doubtful that it is of much normative interest. In particular, I find it hard to see how an organism could be the object of self-concern. I’ll confess to having at most derivative concern to the human animal I inhabit, considered apart from my psychology.

    Many, including many who reject a psychological account of metaphysical identity, seem to agree. No doubt there will be those who disagree, and they will not find this paradoxical conclusion much of a threat.

    You note that on McMahan’s view, all that’s needed for some prudential concern is (roughly) the continuation of consciousness or the brain parts that sustain it. But note that this is all that’s needed for SOME such concern. On McMahan’s view, absent significant psychological continuity, this would only call for minimal concern, and an only slightly weaker version of the paradoxical conclusion might still follow.

    Also notice that Velleman is talking here about unbearable pain that continues until death, not about periods of such pain. So even if the subject could still be said to have a range of beliefs, memories, or desires — of such dispositional states — they would virtually never manifest themselves. Arguably, their potential presence is not sufficient for psychological continuity, though it’s not as if there is some very definite agreed account of what this must involve.

    2. I agree with your remarks about rationality under ‘duress’. In the post I did not think that such observations are sufficient to show that the self lingers on. Nevertheless, a subject with rational capacities that is suffering unbearable pain is interestingly different from an animal and not only in potential. It’s hard to say in exactly what way.

    3. I did not mean in this post to directly criticise Velleman’s argument — I think your point hits the target, and both Kamm and McMahan have raised similar objections. (Though remember that Velleman does not deny that pain is bad, only that it is ever bad enough to in itself justify death.)

    But I’d like to say something about ‘unbearable pleasure’ — something I’ve written something about somewhere. Strictly speaking, this is itself a paradoxical notion. Would pleasure that is unbearable be, in some sense, UNPLEASANT? If we wanted the pleasant mirror of unbearable pain, we might rather consider ALL ABSORBING pleasure. If unbearable pain is pain such that one would do virtually anything to end it, all absorbing pleasure would have to be self-effacing pleasure that one would do virtually anything to keep going.

  3. 3. Posted by Nick Shackel | June 20, 2007 1:56 am

    I’m quite puzzled by what you say in this paragraph:
    “Being able to deploy concepts and beg for help …. ”

    We’re concerned only to deny that the rational self disintegrates under unbearable pain. I don’t know what there is to being a rational self, in the relevant sense, but for there to be intact rational capacities and sentience. The unbearable pain is not disintegrating this self but confining it (and that is part of what is bad about it).

    You then go on to say ‘It is not plausible to simply deny that unbearable pain compromises rational agency’. I think you’re failing to distinguish the difference between rationality contrasted with arationality, which latter is what premiss 4 says is the upshot of unbearable pain but is not true, from rationality contrasted with irrationality, which latter is what unbearable pain may indeed bring about. So maybe the argument turns on an equivocation. Unbearable pain may make you an irrational agent, but what the argument needs is that it makes you no longer an agent.

  4. 4. Posted by Guy Kahane | June 20, 2007 2:26 pm

    Thanks Nick. This is a very good point. It advances the discussion, but I don’t think that it offers a complete solution to the problem.

    The problem was to find a normatively relevant difference between a person in interminable unbearable pain (for short, a person in pain), and an animal or demented person, where rational capacities are simply lacking.

    The first point to make is that your suggestion addresses this version of the problem, but not the version of the problem that is stated in terms of psychological continuity. But let’s set this aside and focus on rationality, as Velleman does.

    The simple answer is to point out that the person in pain has the intrinsic potential to exercise rational capacities and the demented person (or ‘post-person’) doesn’t. I raised the doubt about this that merely having the potential isn’t enough if one is never in a position to actualize it. People in a permanent coma also have such a potential. And I also doubted that this could be the whole difference. The rational capacities of a person under unbearable pain are somehow present, if irretrievably blocked.

    You offer a solution. A demented person is arational, but a person under unbearable pain is irrational, which implies he is failing to exercise rational powers which is in possession, and not only in potential.

    The word ‘irrational’, however, is not very precise. It is used to refer to very different things. Weakness of the will is a paradigm form of practical irrationality. Here, irrationality typically implies rational culpability. The person can be held to blame for his error. But consider even a brief episode of unbearable pain. Structurally, this is similar to weakness. But I don’t think we hold people rationally (or morally) blameworthy for their mistaken choices in such conditions. In this respect they are closer to arational subjects.

    Moreover, what we are considering is not a person’s irrational choice when undergoing a short episode of unbearable pain, but a person who will continue to remain in such a state. So attribution of irrationality to choices here is not done against a background of rational or even minimally rational choice. The person, if you wish, is in a permanent state of irrationality, a state that is not so easy to distinguish from arationality. And further, as you note, his rational capacities, even if active, making possible attribution of irrationality, are exercised in the narrowest, pain and present focused way. So there is some case for describing a person in such a state as akin to an animal.

    Imagine an animal whose sole ends are the satisfaction of present impulses and avoidance of present pain, but which, unlike existing animals, can exercise fairly sophisticated cognitive capacities in pursuit of these narrow ends. Would we think of such an animal as a rational being?

  5. 5. Posted by John Alexander | July 3, 2007 1:40 pm

    You write: “A more plausible claim is rather that although unbearable pain seriously compromises rational agency, the rational self nevertheless lingers on. After all, even when subject to unbearable pain, a person maintains some capacity to think and act, however limited and distorted. And his various rational capacities are still somewhere in there, ready to kick back into play the moment pain is reduced or removed.” Earlier you maintained that you thought that being in unbearable pain limited a person’s rational self thereby, I assume, disallowing that person to make decision regarding the permissible outcomes.

    I am wondering if we can make the following claim regarding determining the rational ability of a person : In order to determine if a person is rational we need to determine their compentency to make decisions. There exists questions that can test for competency, for example, 1)does the person understand the situation she is in, 2) does the person understand the consequences of taking certain actions, 3) can the person give supporting reasons why she wants to choose one outcome over another?

    It would seem permissible that if a person were able to successfully answer these questions then she should be able to choose the outcome that best serves her deired objective regardless of whether she is in ‘unbearable’ pain or not.

  6. 6. Posted by Guy Kahane | July 3, 2007 9:56 pm

    Thanks John. You seem to be suggesting certain criteria for rational agency. Let see how they apply to the case in question. Imagine someone being tortured. The pain is unbearable. Presumably, they don’t lose their ability to know what’s going on around them, though, because of their focus on their pain, they may not be taking in all that much of what is going on. To some extent they also understand the consequences of their choices, though again possibly in a greatly constrained way. The main problem is with the third condition. A person in this state can, notoriously, be made to choose anything that would relieve the pain, even if that thing is the very opposite of their deepest commitments or values. Is it sufficient, for her to count as a rational agent, that she would still have grasp of the normative reasons for not doing what she will in fact inevitably go on to do? But then her rational agency is utterly passive, playing no role in shaping action. And, on the other hand, she could of course always cite the terrible pain as her motivating reason for what she did. But that doesn’t see sufficient for rational agency either.

    So it seems to me right to say that under these conditions a person’s rational agency is deeply compromised. We need a substantive view of rational agency to be able to say more. We need to decide, for example, whether rational agency might be a matter of degree, and even if so, if it needs to meet certain a minimal threshold or minimal criteria. And the initial worry was that a person under long enough unbearable pain would fail to meet such criteria, and thus, on certain accounts of what matters in survival, would no longer preserve enough of the self.

  7. 7. Posted by John Alexander | July 4, 2007 4:58 am

    Thanks Guy for responding
    I agree that in your example she may be compromised. As you write: “A person in this state can, notoriously, be made to choose anything that would relieve the pain, even if that thing is the very opposite of their deepest commitments or values.” The problem with this example is that her rationality and agency is being subverted by being ‘made to choose’ by some third party that is acting ourside her control. However, even in this situation she might want to die if the ony other option given her is that the torturing is going to go on without end. I am not convinced that this might not be a rational decision given her options; die or continue to suffer. She might very well satisfy the three conditions I mentioned earlier. If she is being tortured she might not be able to act on her decision, but that is another issue. (I should note that there may be more conditions that need to be meet, but the three given seem the minimum.)

    One might suggest that what make her decisions rationally compromised, if it is compromised, in the ‘torture case’ is similar to an example where a person is suffering pain from terminal cancer. But, I think that it is certainly conceivable, and has in fact happened, that a person in this type of situation where she is going to die in some period of time, and that she will continue to suffer during that period of time, might very well reason that shortening that time frame is a good thing for her to do. I am reminded of Rachels’ argument here.

    It seems rational to make it a conditon of a life plan that one might want to choose to die if certain conditions prevail. Using Kant, this would be part of a hypothetical imperative wherein one chooses to live a life within certain parameters and acts accordingly within the parameters of the categorical imperative (CI) to meet that objective or goal. It seems consistent with the CI that one of these parameters might include some physical/mental issues arising that might compromise her abilty to act on her own accord and deciding that if these conditons prevail she will choose to end her life. I think that this can be universalized into a normative principle.

    Can rational people disagree on what can be universalized or what are the parameters wherein people can choose to live? It seems plausible to maintain, based on empirical evidence, that people do disagree and can support their positions with sound reasons. Where disagreement exists the choice is left up to the person(s) directly affected. If condition x (terminal cancer for example) exists, one person may choose to die while another may choose to live. I think that each can be rationally defended.

    A final point; the criteria for determining competency are important in order to avoid to problems discussed in the lasst paragraph of your reply. In orde to know if a person fails to meet certan criteria one needs to test that person utilizing some firm criteria. I should note that the criteria I mentioned where derived from those developed by Roth and Meisel to help health care professions deal with questions of a patient’s ability to make important decisions. They indicated 7 criteria, I gave only 3. Maybe more are needed. I think that a crucialpart of understanding the situation one is in is understanding disclosed information from others concerning that situation and available outcomes.
    Guy; thanks for an enjoyable exchange of ideas.

Post a comment

Name: (required)

Email Address: (required) (will not be published)