The Deepest Error in Kant’s Ethics

21 Replies to “The Deepest Error in Kant’s Ethics”

1. Ralph,
   thanks for clearly describing Kant’s views on goodness. I’m not too sure this:
   “First, I am inclined to believe that the world is full of intrinsic values that are completely independent of and antecedent to the value of the good will. In my view, these primitive intrinsic values include at least the following: the ecological value of flourishing living organisms and ecosystems; the disvalue of physical pain, and the value of freedom from pain; and the value of admirable cognitive achievements of various sorts – including artistic, athletic, intellectual, and scientific, achievements, among others. None of these wonderful things owe their value to their being the object of the good will.”
   is an argument against Kant’s view. Seems to me just more like a statement of an opposing view. Kant of course can accept that all these things have intrinsic value. Yet, the advantage of his view is that he could perhaps give an explanation for the value of these things (additionally, if you accept the constructivist part of his view, this might allow also him to be more parsimonious about the evaluative properties which these things have).
   The points 1-3 seem appropriate too. But, I wonder whether they manage to show that Kant’s theory is fundamentally flawed. One might try to develop his view without referring to natural teleology (perhaps by thinking about the preconditions of rational willing), develop his views on the contradiction, and perhaps ignore the second formulation.
   I’m not a friend of Kant’s formalism about goodness either but I don’t think we’ve yet seen a solid argument against it. Also, these objections are not only objections to Kant but also to various other dispositional accounts of goodness.

2. Hi Ralph,
   I’m with Jussi on this one. I believe Kant (in the spirit of the opening paragraphs of the Groundwork) would have us imagine the “goods” you propose as products of a will which profoundly disregards the dignity or rights of another person. For example, the goodness of “cognitive achievements” (such as, say, coming to understand profoundly interesting facts about human physiology) seems to be negatively affected if we discover that they have beenb produced by putatively immoral means (such as, say, the capturing and dissecting of a human person against their will).
   If Kant is right about this, it follows that such goods are not “intrinsic” in your sense, for their value is partially explained by certain external conditions obtaining (i.e. the presence of a good will).

3. Hi Ralph,
   Can you say more about why Kant is committed to the move from the supreme condition claim, to the explanatory claim:
   (SC) For Kant, the good will is…is “the condition of every other [good]” (G 396)…“the supreme condition of all good” (KpV 62).
   to
   (Expln) Whenever anything is absolutely good, its goodness is explained by the goodness of the good will
   Why not think that the supreme condition claim is only a logical restriction on the things that can count as good (e.g. intrinsically good or absolutely good)? That allows us to avoid burdening Kant with (Expln) and your worries.

4. Ralph,
   I am very much in agreement with you about Kant’s error. I think you’re exactly right that there is a chain of inference from the absolute goodness of only the will, to the formalism of his ethics. Also that this formalism is not carried through effectively and probably could not be. I have toyed with the idea that there is something conceptually confused about it.
   I would not have connected this error with “the Augustinian Christianity of his youth.” (Incidentally, Kant’s ethics bear certain similarities to those of Peter Abelard, and Abelard’s ethics were roundly rejected by all his Augustinian Christian contemporaries.) I’ve always thought it had to do with Kant’s deep anti-moral luck stance; all the good has to be located in the will, because that’s the only thing totally under your control. Still, I guess you might think that opposition to moral luck has something to do with a Christian background.

5. Ralph,
   I am very much in agreement with you about Kant’s error. I think you’re exactly right that there is a chain of inference from the absolute goodness of only the will, to the formalism of his ethics. Also that this formalism is not carried through effectively and probably could not be. I have toyed with the idea that there is something conceptually confused about it.
   I would not have connected this error with “the Augustinian Christianity of his youth.” (Incidentally, Kant’s ethics bear certain similarities to those of Peter Abelard, and Abelard’s ethics were roundly rejected by all his Augustinian Christian contemporaries.) I’ve always thought it had to do with Kant’s deep anti-moral luck stance; all the good has to be located in the will, because that’s the only thing totally under your control. Still, I guess you might think that opposition to moral luck has something to do with a Christian background.

6. Ralph,
   I am very much in agreement with you about Kant’s error. I think you’re exactly right that there is a chain of inference from the absolute goodness of only the will, to the formalism of his ethics. Also that this formalism is not carried through effectively and probably could not be. I have toyed with the idea that there is something conceptually confused about it.
   I would not have connected this error with “the Augustinian Christianity of his youth.” (Incidentally, Kant’s ethics bear certain similarities to those of Peter Abelard, and Abelard’s ethics were roundly rejected by all his Augustinian Christian contemporaries.) I’ve always thought it had to do with Kant’s deep anti-moral luck stance; all the good has to be located in the will, because that’s the only thing totally under your control. Still, I guess you might think that opposition to moral luck has something to do with a Christian background.

7. Even as a Kantian, I would say that 1 and 3 are quite right. But they seem fairly easy targets as far as Kant goes. However with a bit of effort, no 2 is not difficult to demonstrate.
   The details need to be hammered out, but here is a rough argument.
   1. We assume that you already have the moral commitment. i.e. the commitment to act according to practical laws (principles that apply to all rational beings)
   2. Given the moral commitment, in order to justify non-beneficence on your part, it must be the case that there is a practical law of non-beneficence.
   3. A necessary precondition of identifying whether there is a practical law of non-beneficence is whether you can will that a maxim of non-beneficence be a universal law of nature.
   4. You cannot will that a maxim of non-benficence be a practical law because you will always desire that someone be beneficent towards you, and in a world where non-beneficence is a universal law of nature, no-one is beneficent to anyone else (including you). Therefore you cannot will a maxim of non-beneficence
   5. Non-beneficence is not a practical law
   6. Therefore, if you are morally committed, you cannot act on a maxim of non-beneficence.
   QED
   Some details may need to be hammered out, but the core of the argument still stands. We have a wide duty of beneficence. (which is kind of consequentialist except that this also avoids aggregation.)

8. Hi Ralph,
   Some more thoughts…
   You are surely right that Kant thinks this: “we cannot analyse the good will as: the will that chooses objects that are good in some appropriate way” This comes out very clearly in the second critique.
   But I do not think his argument for it hinges on the claims about goodness that you mention.
   Why not think, instead, that his argument is roughly as follows:
   (1) The Good will is autonomous
   (2) If the good will were analyzable as the will that chooses objects that are good in some independently identified way, then the good will would not be autonomous.
   Therfore, (C) The good will is not analyzable..
   This does connect with his formalism, which may well be untenable.
   But I think that it is the demand for autonomy, not some Christian influenced denegration of the empirical world or its value, that drives him to his formalism.
   His autonomy argument in the second C. does seem to rest on psychological hedonism about non-moral action, and that, in addition to his assumption about the autonomy of the good will, seems to be a good place to attack him.
   Of course with Kant there are many places to look, so I would be very interested to hear where you see the argument you attribute to him. In this response, I am mostly thinking of the second critique.

9. Ralph,
   On 1.: I don’t think this is Kant’s best argument for his views on suicide or on our duty to develop our natural talents.
   On 3.: I’m puzzled. Kant is very explicit that FH is not supposed to be formal. This is stated most clearly when he says that the Formula of Universal Law is the formal expression of the moral law, FH its material expression, and the Formula of the Kingdom of Ends the union of the formal and material. So I guess I’m not seeing the force of your worry. Perhaps your point is that your not seeing how FH is equivalent to FUL, which is a formal principle? It’s not obvious to me that if FH and FUL are equivalent (not semantically, but coextensive in the obligations they impose), then because FUL is formal, FH must also be formal. If I recall correctly, Mark Timmons has a discussion of this issue in his paper that appeared in Jarbuch für Recht und Ethik.
   More generally, you mostly cite the Groundwork in your post, and it’s key to keep in mind that there he’s trying to discern what principles would govern the maxims of a MORALLY good will. I’m fairly confident that Kant doesn’t mean to deny that there are other kinds of goodness, but the project of the Groundwork is to ascertain what the morally good will would choose. So even if you are correct that the other items you cite are intrinsically valuable (and I actually think Kant could concede this), that doesn’t speak against his claims against the uniqueness of the good will as an unconditioned moral good.
   As an aside: I’ve been thinking lately a good bit about how Kantians might justify direct duties to non-human animals (contra Kant’s own position), and part of my thinking is that there’s space in a Kantian view for holding that animal welfare, while not an unconditioned good, is nevertheless intrinsically and finally good.

10. Thanks for all those comments! Here are some brief replies.
    Jussi — Of course you’re right that I didn’t give much of an argument against Kant’s view that the fundamental sort of goodness is instantiated only by the will itself. (I only said that I would “explain why I believe that this view is false” — I didn’t say that I would give a crushing refutation of it!)
    You ask whether Kant couldn’t work out the implications of the Categorical Imperative without relying on natural teleology. E.g. you suggest that he might “think about the preconditions of rational willing”. But I’m sceptical that there will be any way of giving an account of the “preconditions of rational willing” that won’t have to rely on some idea about the value or goodness of the objects of the will.
    Nicholas Smyth — I don’t agree with your intuitions here. I’m inclined to think that admirable cognitive achievements are still intrinsically good even if they are obtained through immoral means. Of course, this doesn’t mean that those intrinsically good achievements make the immoral means justified after all. On balance, the agent still shouldn’t have acted as he or she did — but there was surely a bit more to be said in favour of the action than if nothing admirable at all was achieved.
    Brad (first comment) — I actually understand Kant’s use of the word ‘condition’ (Bedingung) as having explanatory force. So as I interpret the terminology, there is actually no difference between the principle that you label “SC” and the principle that you label “Expln”. (I may be wrong to read this term in this way, but I seem to remember coming to the conclusion that this was the best way to understand the way in which it’s used e.g. in the Antinomies of the first Critique.)
    Heath — I do think that the distinction between the phenomenal world of appearances and the noumenal world of things-in-themselves lies behind most of Kant’s thought from at least the mid-1770s onwards, and for him, what really has this fundamental kind of goodness is whatever it is, in the noumenal world of things-in-themselves, that underlies and corresponds to the good will. This sort of supreme goodness is not really located in the phenomenal world of appearances.
    Nietzsche’s complaint that this is continuous with Christian “slandering of the world” seems fair to me. It is also deeply connected to the archetypal Protestant insistence that we cannot be saved by our “works” — i.e. our outward deeds — since our outward deeds are utterly trivial compared to the infinite glory of God, but only by our “faith”, which is a matter of our immortal soul’s putting itself into the right relationship with God. Kant was certainly not an orthodox Christian, but he was deeply influenced by his Christian background.

11. Murali — You offer a reconstruction of Kant’s argument for the imperfect duty of beneficence. But I find your reconstruction quite problematic.
    Your reconstruction of the argument crucially relies on the claim: ‘You will always desire that someone be beneficent to you.’ If by ‘desire’ you mean ‘inclination’, then what you say can only be a generalization about human beings. But as we know from the first essay in Religion within the Limits of Mere Reason, there are some frankly terrible — indeed brutely evil — aspects of human psychology. So the fact that we just happen all to have this “desire” cannot be enough to explain why the good will must reject the maxim of non-beneficence.
    So instead, you should probably replace the term ‘desire’ with something like ‘rationally will’ (or ‘rationally prefer’ perhaps). But now Kant seems to be smuggling in some assumptions about rational willing that cannot be explained by his theory. Indeed, surely the reason why it is rational for me prefer that others should be beneficent to me is because I understand that it is in various relevant ways good if others are beneficent to me. But then we are positing a kind of goodness that is antecedent to the good will — contrary to the principles proclaimed at the beginning of the Groundwork.
    Brad (second comment) — Kant uses ‘autonomy’ in a rather special sense, of course. As we find out in Section 3 of the Groundwork (and in the Analytic of Pure Practical Reason), autonomy is the positive power that makes it possible for the will to be determined freely, independently of all “alien causes”. It is true that transcendental freedom is the ultimate explanation (the “ratio essendi”) of the moral law, according to Kant. But as he also explains, it is not our reason for believing (the “ratio cognoscendi”) what we believe about the moral law. (See the Preface to the second Critique, KpV 4, footnote.) I assumed that the claims about the good will that Kant made at the very beginning of the Groundwork (which he claims to be part of common-sense moral thought) represent our reason for believing Kant’s theory.
    Michael — I was making a point about Kant’s derivation of the Categorical Imperative (see especially G 401-2 in Section 1, and 420-1 in Section 2). This derivation requires the Categorical Imperative to be purely “formal”. And indeed he says at several points that the principle of the good will is “formal” (e.g. for a typical use of formal/material contrast, see G 400). So if he then goes on to say that the Formula of Humanity is “material”, we must conclude that it is not after all a formulation of the moral law — i.e. the Formula of Humanity is not a formulation of the principle of the good will! There are clear signs here, I think, that Kant is trying to have his cake and eat it too….

12. Ralph,
    I think you are right to worry about the Kantian sort of view you describe here, and it is not clear to me that it isn’t, at least time’s, Kant’s own view. However there is a related view that is also present among the texts you cite, one that you might find less objectionable. Suppose ‘the good will’ that Kant speaks of as the explanation of various values is not, as it were, an individual ‘good person’. So the explanation of value is not the choices of some particular person. Instead, it is the object of an agreement by the Realm of Ends. So, to be good is, on this way of thinking of Kant’s views, to be the object of the appropriate sort of agreement (hypothetical) made between fully rational (in the strong Kantian sense) persons. Then this sort of Kantian can say many, perhaps all of, the things you want to say. A particular person’s choices will be justified (good, admirable, etc.) *because* the thing she chose was a good thing (i.e., would be an object of convergence by rational persons). It is the fact that this thing would be the object of such and such agreement among fully rational agents that makes it good. The world is full of intrinsically valuable things such as you list, ‘prior to’ (in any sense you like) any particular ‘good’ person’s having chosen to pursue, protect, respect, admire, etc., them. Intrinsically good things are just the things that, in virtue of their intrinsic properties say, would be the objects of rational convergence…and so on. So goodness is, on this view, ‘prior to’ the goodness of any particular person’s good choices, and in fact explains the value of that person’s choices.
    You might of course object to the idea that the explanation for why any given thing is ‘good’ is that it is the object of this sort of hypothetical joint agreement or convergence. In particular, you might worry that one needs some sort of ‘prior’ value to explain what makes this convergence of agents on these objects ‘rational’. But that, to my mind anyway, is not as bad as the worries you raise here.

13. If moral philosophy is to have a sound factual basis, it is to be found in the facts about human nature and nowhere else. Nothing else but the sameness of human nature at all times and places, from the beginning of Homo sapiens, can provide the basis for a set of moral values that should be universally accepted. Nothing else will correct the mistaken notion that we should readily accept a pluralism of moral values as we pass from one human group to another or within the same human group. If the basis in human nature for a universal ethic is denied, the only other alternative lies in the extreme rationalism of Immanuel Kant, which proceeds without any consideration of the facts of human life and with no concern for the variety of cases to which moral prescriptions must be applied in a manner that is flexible rather than rigorous and dogmatic.
    Dogma has a place in sacred theology, but in moral philosophy it is truly pernicious and should be avoided.

14. Robert —
    Yes, you’re quite right that the “good will” that Kant is referring to in these passages is not the will of any individual person (not even the will of God, or the “Holy One of the Gospel” whom Kant occasionally mentions). The point of these passages is express the abstract Idea of a (possible) good will.
    So, the intrinsic goodness of types of states of affairs, like the happiness of virtuous people, does not depend on any actual person having a good will and choosing the end of making virtuous people happy. The happiness of the virtuous is good because it is a necessary truth that if there ever were a good will, it would adopt the happiness of the virtuous as an end.
    I don’t see, however, how this enables Kant to avoid the objections that I have made. Kant still needs the principle that would guide any genuinely good will to be a purely “formal principle”; and he also needs to claim that the goodness of flourishing ecosystems, freedom from pain, admirable cognitive achievements, etc. etc., is all entirely derivative from the fact that these things would be chosen by any will that was being guided by this purely formal principle. I just find those claims monumentally implausible.
    Max —
    You raise some really interesting points. I hope that you’ll forgive me if I don’t respond to them here, because I actually think that they will lead in quite a different direction from the topics that have been discussed on this thread so far. (One way to see this is that if you’re right, your point would not just tell against a Kantian ethical theory. It would tell equally against a Platonist theory — such as e.g. Jerry Cohen’s theory of justice — according to which the ultimate truths of ethics are necessary truths that would apply just as much to societies of Martians or angels as well as to human beings.) So I think that responding to your comment would lead us too far afield at this point….

15. Hi Ralph,
    In the analytic of practical reason, chapter II, (KrV) Kant indicates awareness that ‘good’ is subject to ambiguity. His example is that German does, but Latin does not, encourage us to distinguish between something being good-for some being and something being good in the sense that interests him (good in-itself). In 5:62, for example, he says we can make appraisals of goodness in reference to well-being OR of goodness “in-itself”.
    This suggests that he does not think that the concept of well-being or good-for is to be reduced to or explained by the concept of a good will. And, by extension, helps assuage, I think, your worry that he is committed to an implausible form of good-reduction.

16. OK Ralph, I understand, is there another time and place for such a discussion?

17. Brad —
    I never claimed that Kant thinks that the concept of well-being can be reduced to the concept of the good will! As I said above, “The only sort of goodness that I am concerned with here is absolute goodness (not mere relative goodness).”
    As it happens, however, I am not quite sure whether I agree with your reading of that passage from the second Critique….
    As I read him, Kant insists that the German word ‘gut‘, when used correctly, cannot express the same concept as the word ‘Wohl‘. (As he says at KpV 58: “Linguistic usage distinguishes the good from the pleasant, and demands that good and evil must always be judged by reason, and hence by concepts that can be universally communicated, and not by mere feeling….”.)
    For Kant, the German word ‘gut‘ can only express the concept of something that is “the necessary object of the faculty of appetition, in accordance with a principle of reason”.
    Of course, there are two kinds of “principles of reason” here: (1) Hypothetical Imperatives, and (2) Categorical Imperatives. Hence there are also two proper uses of the German word ‘gut‘:

    1. The first is mere relative goodness — i.e. being useful or good for something. This meets Kant’s definition of ‘good’, since “the judgment of the relation between means and ends certainly belongs to reason”.
    2. The second is absolute goodness. But as Kant claims here (KpV, 60), “it is only the way of acting, the maxim of the will, and thereby the acting person himself, that can be called good or evil in this way”.

    By contrast, happiness and pleasure can only be judged by feeling, not by reason, and so they cannot properly be called ‘good’ in any proper sense of the German word ‘gut‘. They can only be picked out by the radically different concept ‘Wohl‘ instead.

18. Hi Ralph,
    (1) Sorry about the false attribution to you!
    (2) I am not sure we really disagree about what Kant thinks about ‘gut’. I do not see how anything I said about ‘good’ was at odds with your exegetical claims – which all sound plausible – I was assuming that Kant would think that the english ‘good’ is like the latin ‘bunom’.
    (3) But your response does raise an interesting issue that worries me about your attack on Kant – this relates to my initial comment about autonomy, but I think I can express a question about your view more clearly now.
    As you point out, Kant thinks that ‘gut’ “only express the concept of something that is “the necessary object of the faculty of appetition, in accordance with a principle of reason”,” and is thus not empirical, as is the concept expressed by ‘Wohl’.
    But if this is right, then your original attack on him should really build this into the definition of ‘absolute good’. The english ‘good’ does not obviously have the meaning that Kant assigns to ‘gut’ so your claim that his theory offends intuitions about which things are good (relatively OR absolutely) does not necessarily go through.
    It only goes through if the objects you claim are good absolutely are also ‘gut’ absolutely, so to speak.
    This makes me want to hear more about why you think the relevant objects (e.g. cognitive achievements) are not only non-relatively good but also ‘gut’ in Kant’s sense.

19. If by ‘desire’ you mean ‘inclination’, then what you say can only be a generalization about human beings. But as we know from the first essay in Religion within the Limits of Mere Reason, there are some frankly terrible — indeed brutely evil — aspects of human psychology. So the fact that we just happen all to have this “desire” cannot be enough to explain why the good will must reject the maxim of non-beneficence.
    A few things:
    1. The good will, according to Kant, is one that has the moral commitment.
    2. Now, it may very well be the case that w have downright evil inclinations, but kant seems to be saying that the one thing all limited beings desire is their own happiness. (There might be some kind of hedonism going on here, but we should at least take it tentatively that whatever else people may have an inclination towards, they have an inclination towards their own happiness). What Kant is saying is that given that we are acting on principles that are applicable to all rational beings, what kind of limitations can we place on our desires?
    So, applying the FUL basically takes a maxim M: phi if you desire X and asks whether or not we would desire a world where people invariably followed maxim M. If, in such a world, your desire for X was even more frustrated than a world without the maxim, or a desire that all limited agents/wills (holy wills have no sensuous natures) had, then it follows that you could not desire that the maxim be a universal law. If you could not desire that it be a universal law, and given that you are committed to acting only on practical laws (i.e. you have a good will) there would be a contradiction if you acted on the maxim anyway. So, it is the case that the mere fact that we all have this desire is sufficient to establish the wide duty of beneficence.
    That being the case, it does not preclude that happiness itself is independently important. (it may be valuable in some complicated non-aggregative way for instance). I agree that Kant doesnt adequately make the case that things other than the good will are not good in themselves. But, the goodness of the good will doesnt rely on the goodness of other things. The goodness of the good will is more closely related to it legislating ends for itself that are universally applicable. (or something like that. I’ll have to re-read my kant)

20. Ralph,
    Thanks for the post. I want to take up a point made much earlier by Nicholas about the conditional nature of the value of all external objects and ends-to-be-effected (as opposed to us, who are ends-in-ourselves). Take the example of the intrinsic, unconditional disvalue of pain. I think that this is mistaken, and that we can see this by considering cases where pain itself is the thing that is valued. The point of the following example is to show that pain is sometimes, itself, valuable, and therefore that it is not inherently disvaluable.
    Think of going to the gym to exercise: for many, the pain of exercise is sought. If you don’t feel a little pain you’re not doing it right. One way to think of this phenomena is that the pain is still disvaluable, but it is valued as a means to health. However, I think that sometimes we value the pain itself, not simply as a means, but insofar as the pain is a signal (of effort, or effect, or whatever). It is the pain that is taken as valuable, because of the kind of thing it is (that is, taken as intrinsically valuable). A related case is someone who, as a result of a guilty conscience, engages in a punishing exercise routine. The pain is valued as a punishment. The experience of pain is a good thing, because it is unpleasant, and this unpleasantness is (from the point of view of the agent) deserved.
    If these examples work, they show that pain is not unconditionally disvaluable, but rather its value or disvalue depends on the ends of some agent. Then, of course, we get Kant’s argument that the good will is unconditionally valuable, ending a regress of merely conditionally valuable things and grounding a world that can then contain objective value.

21. Pete, in addition, you might also want to look at cases where you eat spicy food (with lots of chilli/pepper) Lots of people like to eat very spicy food (not the mild stuff with jalepenos, but the really fiery stuff) Speaking from experience, indian curry will not taste as good without the chilli. However the spiciness of chilli is painful. In fact, the without the sharp stabbing pain followed by the slow burn, the food wouldn’t taste as good. i.e. we are not eating it in spite of the pain (the chilli doesnt seem to have any other flavour other than pain), we seem to be eating it because the chilli gives us pain. It seems that at least to this extent, a lot of us are masochists.