What Kantianism Gets Wrong

With regard to moral theory I have two hunches. One is that the wellbeing of one’s fellow humans is an intrinsic moral value. Intrinsic not only in that a moral agent will care about it for its own sake but also in that its value is not derivative from other values, like, say, that of rational agency. So Kantianism is false. The other is that the wellbeing of one’s fellow humans isn’t the only intrinsic moral value.  There are virtues that are independent of  benevolence, and respect, of the sort that makes paternalism wrong, is one of them. So utilitarianism doesn’t work either.

But after decades of Kantian dominance in analytical ethics, some of us have become used to thinking of concern for the wellbeing of others as a somehow coarse, primitive virtue befitting swine and Jeremy Bentham, unless it is somehow mediated by, derived from or explained through something more complicated and refined, like the value of rational agency.

Suppose one is roughly Kantian. Reverence for rational agency is the one basis of morality as far as one is concerned, where rational agency is thought of roughly as the capacity to set ends. What to do with the sense that benevolence is a major part of morality? The answer seems to be “think of benevolence in terms of a duty to adopt and promote other people’s ends”. Now suppose that, as many contemporary Kantians do,  you reject the idea that adopting and promoting a person’s ends is the same thing as protecting her wellbeing – after all, most humans have some ends for which they are willing to sacrifice some wellbeing. In this case, what you say is that at the heart of benevolence we have a duty to adopt and promote people’s ends. We also have a duty to protect human wellbeing because, even though it’s not the only thing people care about, it is a very important end for all agents.

I don’t think this works, though. My argument goes like this:

  1. If the reason protecting a person’s wellbeing is important is purely the fact that her wellbeing is an important end to her and we have a duty to adopt her ends, then it would be of at least equal moral importance to protect any end that is at least equally important to her.
  2. Protecting an agent’s wellbeing is something we are morally called upon to do in some cases where, other things being equal, we would not be called upon to protect her pathway to achieving another equally important (to her!) end.

Therefore, it is false that the reason protecting a person’s wellbeing is important is purely the fact that her wellbeing is an important end to her and we have a duty to adopt her ends.

 Let me talk about 2) and why it’s plausible.

Take a case where an economically comfortable person, let’s call her Mercedes, is asked for help by her desperate acquaintance, Roger. She can, by paying 50 dollars, rescue him from being beaten up. If beaten up, Roger would suffer pain and then have to spend some days in a hospital, but he is not going to be killed. I am trying to stick to 1000 words so let me just promise you I have a half-way-realistic case.  Now imagine an alternative scenario in which a person – call him Leonard – asks Mercedes for 50 dollars because without them, a great opportunity to travel and spread his Mennonite religion will have to be relinquished.  Leonard’s end (spreading his religion) is as at least important to him as Roger’s end (not being beaten up) is important to Roger, and more important to Leonard than Leonard’s own wellbeing – he is willing to suffer for it if needed. For all Mercedes knows, spreading Leonard’s religion is itself strictly morally neutral – she has no particular reason to spread it independently of him.

There is an asymmetry between the cases. In the first scenario, Mercedes would display a lack of benevolence – perhaps of decency! – if she were to refuse to rescue Roger from a beating by giving him $50, given that this would be easy for her, no harm would be caused by it to anyone, etc. In the second scenario there is no such presumption. If Mercedes likes Leonard’s cause, it makes sense for her to make a donation. If she’s indifferent to his cause, no compelling reason to donate is provided by the very fact that Leonard would be ready, if worst comes to worst, to suffer for his cause. Unless she does fear for his wellbeing – fears, for example, that Leonard is a bad shape and will plunge into a horrible depression if she declines – Mercedes is not any less of a good Samaritan, certainly isn’t a sub-decent Samaritan, for not wanting to donate to another’s morally neutral cause, however crucial her donation would be to the cause.

If all that made Roger’s wellbeing matter morally was its importance to him as an end, she would have had as much of a duty to help Leonard.

Some Kantians would reply that what matters here isn’t protecting Roger’s wellbeing but the fact that Roger might lose rational agency. Roger, however, is not in danger of death or brain damage. He might suffer pain, but it takes a truly extreme amount of suffering to deprive someone of basic human rationality. His ability to perform successful actions will be impaired for a few days, but being a rational agent is not about being a successful performer of actions – it is about being responsive to practical reasons. It would be quite wrong to say that anyone with whom the world does not collaborate – because of an injury, or due to being in chains for that matter  – is thereby not a rational agent. Further more, preventing a few days of suffering is more morally urgent than preventing a few days of involuntary deep sleep with no significant harm expected, though involuntary sleep deprives you of agency if anything does.

There is something special about wellbeing.

14 thoughts on “What Kantianism Gets Wrong

  1. Nomy,

    This is an interesting challenge to Kantianism, but I want to push back on what you say here in the spirit of trying out a defense of Kantianism (or, at least, one version of Kantianism).

    Consider a Kantian account that begins from the thought that rational agency is not simply the capacity to set ends, but something like the capacity to set rational ends. On this view, there are certain ends that are *truly* one’s own. We might conceive of these in terms of ends that are rationally necessary for any rational agent or ends that are rationally consistent with such ends. (I’ll leave the details open here. There are familiar enough accounts of how the story is supposed to go.) Granted this, we can explain the asymmetry between the cases you describe, not in terms of the brute value of wellbeing, but rather in terms of it being the case that one’s wellbeing is a necessary end for any rational agent but spreading one’s religion is neither rationally necessary nor consistent with rationally necessary ends. Roger’s end is truly his own; Leonard’s is not.

    Given all this, the Kantian can agree: “There is something special about wellbeing.” But he will disagree about what explains wellbeing’s special status. On your view, wellbeing is an intrinsic value, both in the sense that we ought to care about it for its own sake and in the sense that its value is brute. (Please correct me if I am misunderstanding you at all.) On my Kantian’s view, wellbeing is special because one’s own wellbeing is a necessary end for any rational agent and the fact of being a rational agent’s end is value-conferring. This maintains the Kantian insistence on grounding value in rational agency–wellbeing, on this view, is intrinsic in the first sense but not the second. But (I think) it departs from the flavor of Kantianism you discuss in that it does not simply grant that the status of being a human being’s end is value-conferring. It calls attention to the fact that we are not always rational when adopting ends. Sometimes we adopt and pursue ends that are not rational for us to have, and these ends, on the view I am suggesting here, are not conferred with value for this reason. They are not truly our own.

    I’m sure you have things to say about this. I’d be very interested to hear them.

    Liked by 1 person

    1. Interesting, but 1.do you really want to say that every time I have a non-selfish end that isn’t part of morality that end is incompatible with rationality or not really mine? Is a rational agent not allowed to prefer to make good art at a cost to her wellbeing? Spread the Basque language at a cost to her wellbeing? Work extra to make her child’s party at the expense of the agent’s wellbeing? Are you saying that to be rational, we need to have no other motives except self-interest and morality? 2. Suppose you bit that bullet. If only my wellbeing compatible ends are “really” mine, that means that the only ends of mine that you have to *respect* are the ones conducive to my wellbeing. What a nightmare of permissible paternalism appears to lurk here! What I like best about Kantianism is that according to it, you are not allowed to force my wellbeing on me. You are not allowed to force me to eat vegetables. If you are allowed to disrespect those of my ends that don’t go with my wellbeing, this all collapses. You can use coercion or manipulation to prevent me from getting married to a jerk because my “real” ends are not thereby disrespected….


    2. Correction: it probably doesn’t follow from your view that rational people only care about self and morality, but it seems to follow from it that with the weird exception of morality, having any end that you even occasionally prefer to your wellbeing is “not consistent with with rationally necessary ends”. I can’t agree to that.

      Liked by 1 person

  2. Correction: Instead of “not having a duty to respect my rationally adopted ends” I should have said “not having a duty to respect my *not* rationally adopted ends.”


  3. If you stretch the definition of wellbeing to include everything the agent wants you collapse the distinction between self interest and nonselfish motives. However, if friend who is not rich gives me a $1000 gift out of love, esp a morally optional one, I am the one who is grateful to him. If giving me gifts counted as acting for his well being than if anything he is the one who should be grateful to me for giving him the opportunity! More dramatically,I should not be able to look at people who risked their lives for the Basque language (real people) and say that really they are the same sort as people who are out to make money – after all both are cases of seeking one’s wellbeing! I short, there are self-regarding desires and such that are not. The distinction is important. You flatten it if you extend wellbeing so far. I’m repeating here arguments made by the very Kantian Stephen Darwall in “welfare and rational care”.

    On Kantianism, the reason not to coerce or manipulate is respect for the agent’s ends. If an end doesn’t belong to the agent, there is no reason to respect it, and so no reason to avoid coercing or manipulating the agent into not acting on it. If they are not truly the agent’s ends they don’t count for self-determination either. Surely if you do analytical philosophy you accept that “truly his own” and “his own” should mean the same thing). If it’s ok by you for even a hypothetical all-knowing person to force me not to do something just because it’s bad for me, you might as well be a utilitarian. And, I wouldn’t want to be around you, as I want the right to make my own mistakes! 😉

    There might be “guard rails”, but they involve things like psychosis or maybe really bad addiction. As long as you have the basic, average human capacity to respond to reasons you are part of the kingdom of ends and cannot be coerced or manipulated. If we raise the standards from that to ideal rationality the kingdom will be empty.

    Now, back to the beginning. You still don’t have an explanation as to why Mercedes has reason to help Roger but not Leonard. Stipulating that Leonard is irrational in spreading his religion is very ad hoc.

    Thank you for interesting comments. Why won’t you tell me your real name?


    1. Hi Nomy,

      Sorry for the tardy response. I live and teach a bit North of Houston. It’s been a rough (though not catastrophic where I am) weekend. There was a lull in the storm today, but we’re forecast to get more rain before the week is out.

      I haven’t been meaning to hide my real name. I’ve been abbreviating my long last name: Mitchell-Yellin. (Incidentally, and not that you’d have much reason to remember: we met a few years ago at the Tucson normative ethics workshop. I enjoyed that.)

      I’ll try to speak to your challenges sometime soonish. For now, I should say that I think I agree with your insistence that the distinction between self-regarding and other-regarding commitments is important. I was reluctant to grant that all rational commitments are either moral or to be wrapped up in terms of one’s wellbeing. I granted it for the sake of argument, but your comments help to show why one probably should not go the route I was sketching in my earlier reply.

      I’ll need to think some more, with a clear head, about specifics, including how I want to handle your case. Thanks for your patience.


      1. Hey, no obligation, I will remember the question whenever you feel like getting back to it, if you do. I have friends in Houston – hang in there!


  4. Here’s another defense of Kantianism against some (not all) of the worries you raise.

    1) There are means of action that are needed for a wide variety of human activities. They include a healthy body, broadly useful skills, freedom of movement, the opportunity to own property, and the opportunity to seek others’ help.

    2) Because I cannot predict what I will have reason to do in the future, I have reason to maintain my access to these all-purpose means, e.g., to maintain my health.

    3) Because the opportunity to seek others’ help is an all-purpose means, I should not adopt a maxim of indifference to others. I should be open to helping others. I have special reason to be open to helping others when they need my help to preserve their all-purpose means (e.g. their health). (See Barbara Herman, “Mutual Aid and Respect for Persons.”)

    4) There is no mechanical procedure for deciding which all-purpose means I should prioritize, or whose. I should, however, give higher priority to the preservation and promotion of all-purpose means than I give to satisfying mere whims.

    5) The doctrine of double effect applies. It is wrong intentionally to deprive someone (including oneself) of an all-purpose means of action as a means to an end, except to address a grave threat to someone’s agency. It is sometimes OK to do something that one foresees will lead to someone being deprived of an all-purpose means of action. Indeed, this is sometimes inevitable (e.g. when choosing which drowning person to save).

    This account of the significance of all-purpose means accommodates your intuition about Roger and Leonard. Freedom from serious forms of pain is an important part of having a healthy body. Pain is distracting and interferes with many forms of activity (even when it does not shut down agency altogether). So Mercedes has a stronger reason to help Roger avoid pain than she does to help Leonard with a morally neutral cause.

    The account doesn’t accommodate your intuition about the relative priority of preventing pain and preventing abnormal periods of unconsciousness (e.g. comas). I don’t share the intuition, though. Preventing pain is good. Preventing people from going into comas is good. Is it obvious that if one has a choice between preventing one person from experiencing pain and preventing another person from going into a coma, one must prevent the pain? Without more information about the situation, I think this is a hard decision.

    (Shameless self-promotion: I defend this view about all-purpose means in “Imprisonment and the Right to Freedom of Movement,” just published in Chris W. Surprenant, ed., Rethinking Punishment in the Era of Mass Incarceration. Manuscript here: http://www.robertchughes.com/freedom-of-movement-final-manuscript.pdf)


    1. Hi Rob,

      First, I just don’t think everything that isn’t a multi-purpose need is a “mere whim”. This might be a reasonable attitude for a government or charitable organization who has to deal with a large number of people it does not know. As a manager of such an organization I might decide that a budget for providing poor young people with free contraception is a higher priority than a budget for providing poor young people with musical instruments, as sex and relationships that include sex are a good that is important to many, many people’s lives and musical instruments only for some. However, if I am operating as an individual who wants to help a young, poor person I know, it is simply false that all requests for contraception are more urgent than all requests for clarinets. It might be that for an individual person, with particular values, access to a clarinet is important for her wellbeing whereas sex isn’t. I am not going to turn her down because clarinets are only relevant to some lives. “Primary goods” and bads make a lot more sense as a political philosophy thing, so this may not touch views you have about public policy.

      Second, I talked about a sleep and not a coma, and I meant it. A comma usually means that someone has a very severe condition and might not ever wake up, whereas even very bad pain is often temporary by nature or a symptom of a condition that is curable. Other things being equal, 3 days in bed sleeping because someone put Lorazepam in my water supply vs 3 days in bed with pain because someone put a nonlethal but nasty toxin in my water supply? No. Question. There. The badness of pain is not just in the fact that it will interfere with my ability to think philosophy or pay the phone bill on time as Lorazepam-sleep will have exactly the same effect. If we are talking severe pain, it is a lot more morally urgent to help me. Naturally, things are often unequal: toxin on the weekend vs lorazepam on a day I have a job interview, say. But that should not distract us.

      Another piece of evidence to the effect that pain is important above and beyond the fact that it interferes with rational action is that palliative care is important and withholding palliative care is a terrible thing even if the only choice is either a patient prevented from being optimally rational by severe pain or a patient prevented from being optimally rational by morphine.

      Also, I know people who function excellently
      while in pain, literal or emotional or both. Pain does not interfere with their ability to perform rational actions- it ruins their ability to *enjoy* their actions, but that’s not the kind of interference you need. Pain doesn’t cause a reduction in quality of life, the way losing money does, but rather *is* by default a reduction in quality of life. Again, things are not equal – pain can come with gain etc.

      No worries about linking to your own relevant work. Right now if I read about mass incarceration I’ll probably explode – the news is bad enough – but duly noted.


  5. Hi Nomy,

    Thanks very much for your reply. I agree that not all desires for things other than all-purpose means are whims. The desire to be free from pain is not a whim, and as you point out, obtaining freedom from pain does not always improve rational functioning , e.g. when the choice is between being distracted by pain and being equally distracted by morphine. I’m inclined to say that some desires are either so intense or so fixed (not open to revision) that they make demands on our beneficence that compete with the demands of effective rational agency.

    In your Lorazepam v. toxin comparison, it is striking that the comparison case is three days of suffering in bed. If someone is confined to bed, the effectiveness of that person’s agency has been compromised. If the choice is between saving someone from an unwanted three-day sleep and saving someone from three days of pain that is in no way disabling or even distracting, I think one should prevent the incapacitation.

    I’m curious about whether you think of well-being solely in terms of subjective experience. Many people care strongly about what happens to their bodies after they die. If someone dies penniless, don’t people (relatives, friends, perhaps also taxpayers) have a moral reason to pay for a funeral that the person would have regarded as a decent funeral? I would say yes, even though doing so will not affect either the person’s subjective experience or the effectiveness of the person’s agency.

    Regarding skills as all-purpose means: there are many widely-useful skills, and we have to make choices about which ones to cultivate. Also, cultivating a concrete skill can help to cultivate a range of other skills that are widely applicable. So giving someone a clarinet can be a way to help enhance that person’s effective agency, even though only some people want to play clarinet. The process of learning to play a musical instrument involves cultivating many abilities in addition to music-making itself: reasoning ability, physical coordination, knowledge of culture and history, aesthetic judgment, patience, time management, cooperating with others in an ensemble…

    Liked by 1 person

    1. I’ll start from the end. Who said anything about skills? A person might just love clarinet music non-instrumentally, and that can result in clarinets being important to her achieving her ends. I am curious. How does your view discriminate between things that are means, all-purpose or otherwise, and things that people value for their own sake – sometimes all people or most people? I haven’t actually read your work, so this is a clarification question rather than an objection.

      But my clarinet example was just meant to make the following point: if everyone needs X and only Jane needs Y, it does not imply that Jane’s need is less urgent (I take that it that not everyone needs clarinets, as there are other ways to gain the skills you are talking about). Thus the reason Leonard’s need fails to have the same normative pull that Roger’s has is not simply the fact that Roger’s need is universal.

      I certainly do not accept the experience requirement on wellbeing! I am just insisting on a difference between desires that are self-regarding and desires that are not. Suppose posthumous good and harm are a real thing. Compare two desires that don’t involve experience: a desire for posthumous fame and a desire that your young friend will do well after you die. The desire for posthumous fame is self-regarding: you want Rob to be successful for Rob’s sake. The desire for your friend to do well need not be: you can want it for your friend’s sake, not your own. It’s a desire for her wellbeing, not for yours. If you stretch the meaning of “wellbeing” to the point that it simply means getting what you want, you lose the distinction between self-regarding wants and other wants.


      1. I think I misunderstood what you were saying about the clarinet example.
        The chapter I linked to argues that it is wrong to deprive people of one of the major all-purpose means of action in order to improve people’s welfare. For example, it is wrong to imprison people (thus depriving them of freedom of movement) in order to deter people from committing “quality of life offenses.” The chapter does not address questions about how one should allocate charitable resources.
        When I posted earlier, I thought that the distinction between all-purpose means and other goods might help to identify higher-priority demands on beneficence. Your examples above persuade me that it won’t do to say simply that people’s interest in access to all-purpose means has priority over other needs and desires. Two possibilities now seem plausible to me (both compatible with Kantianism):
        I) Perhaps two sorts of desires make higher-priority demands on beneficence: desires to preserve or to obtain access to all-purpose means, and desires (like the desire to be free from intense pain) that are powerful, not revisable, and for morally permissible objects.
        II) Perhaps the attempt to rank demands on our beneficence is misguided. The duty of beneficence is imperfect. It can be shaped by other duties, e.g. contractual duties, familial duties, or other duties arising from relationships. When it comes to charity toward strangers, there are no needs or desires (for morally permissible objects) that we are morally required to prioritize over others.

        Liked by 1 person

      2. I don’t think II works, as not all duties to help are imperfect duties (see Barbara Herman on that). If you see a person lying by the side of the road, badly wounded, it is your duty to at least call 911 – to help this particular person, even if you have helped 5 people in the relevant week. I don’t think I works because there are a lot of things that people are committed to the darnest things – be willing to risk death or being imprisoned by a dictator for the sake various morally neutral causes. I don’t think intrinsic desires are ever “revisable” – you can decide no to eat meat, but the desire to eat meat does not thereby go away. But if you want to use I anyway, you don’t sound Kantian anymore. Kantians don’t rate desires according to resistibility but ends according to rationally according to importance.

        Above all, I think your view of imprisonment can easily be sustained without appeal to Kantianism. I criticized Kantians on the subject of benevolence, but it might be that for lawmakers, as for others, *respect* trumps benevolence. If respect is what we care about, it makes sense that we will care about freedom to follow one’s ends and not about wellbeing, and about what’s necessary to pursue ends rather than what’s necessary to pursue wellbeing. Whether or not to imprison a person can be a question about rights, and it’s perfectly possible for Kantians to be basically correct about rights while benevolence comes from a different source. Ultimately, I’m trying for a pluralistic view.

        Also, I won’t be surprised if most of the things you think of as universal means would, in my book, end up things that trigger Duties of Easy Rescue, because universal means are also means to wellbeing.


Comments are closed.