Kant's Groundwork for the Metaphysics of Morals

I read this because Alain Badiou, in Ethics: An Essay on the Understanding of Evil, writes:

The explicit reference of [today's 'ethical' orientation], in the corpus of classical philosophy, is Kant. Our contemporary moment is defined by an immense 'return to Kant'. In truth, the variety and the detail of this return are labyrinthine in their complexity; here I will concern myself only with the 'average' version of the doctrine.

What essentially is retained from Kant (or from an image of Kant, or, better still, from theorists of 'natural law') is the idea that there exist formally representable imperative demands that are to be subjected neither to empirical considerations nor to the examination of situations; that these imperatives must be punished by national and international law; that, as a result, governments are obliged to include them in their legislation, and to accept the full legal range of their implications; that if they do not, we are justified in forcing their compliance (the right to humanitarian interference, or to legal interference).

Ethics is conceived here both as an a priori ability to discern Evil (for according to the modern usage of ethics, Evil -- or the negative -- is primary: we presume a consensus regarding what is barbarian), and as the ultimate principle of judgement, in particular political judgement: good is what intervenes visibly against an Evil that is identifiable a priori....

The presuppositions of this cluster of convictions are clear.

1. We posit a general human subject, such that whatever evil befalls him is universally identifiable (even if this universality often goes by the altogether paradoxical name of 'public opinion'), such that this subject is both, on the one hand, a passive, pathetic, or reflexive subject -- he who suffers -- and, on the other, the active, determining subject of judgmement -- he who, in identifying suffering, knows that it must be stopped by all available means....

The heart of the question concerns the presupposition of a universal human Subject, capable of reducing ethical issues to matters of human rights and humanitarian actions.

We have seen that ethics subordinates the identification of this subject to the universal recognition of the evil that is done to him. Ethics thus defines man as a victim. It will be objected: 'No! You are forgetting the active subject, the one that intervenes against barbarism!' So let us be precise: man is the being who is capable of recognizing himself as a victim.

It is this definition that we must proclaim unacceptable -- for three reasons in particular:

1. In the first place, because the status of victim, of suffering beast, of emaciated, dying body, equates man with his animal substructure, it reduces to the level of a living organism pure and simple....

2. In the second place, because the ethical 'consensus' is founded on the recognition of Evil, it follows that every effort to unite people around a positive idea of the Good, let alone to identify Man with projects of this kind, becomes in fact the real source of evil itself....

3. Finally, thanks to its negative and a priori determination of Evil, ethics prevents itself from thinking the singularity of situations as such, which is the obligatory starting point of all properly human action. Thus, for instance, the doctor won over to 'ethical' ideology will ponder, in meetings and commissions, all sorts of considerations regarding 'the sick', conceived of in exactly the same way as the partisan of human rights conceives of the indistinct crowd of victims -- the 'human' totality of subhuman entities.

Badiou's first use of the name 'Kant', above, is accompanied by a footnote to Kant's Groundwork for the Metaphysics of Morals. I decided to read the Groundwork to determine whether Kant actually says the things that Badiou attributes to a 'return to Kant'. My conclusion is that Kant doesn't say any of those things, and that, in fact, what Kant does say is incompatible with the ethical orientation Badiou outlines above.

Kant's project is to find a supreme principle on which all moral deliberation can be founded. He argues that this principle must apply not only to human beings, but also to any other conceivable rational being (he is thinking of God in particular). Therefore, it must be something that can be derived solely from concepts of pure reason, rather than from empirical grounds (which would be contingent on the particular characteristics of human experience).

For Kant, human beings are characterised by their capacity for rationality, and by 'autonomy' of the will, which is (and should be) free to direct itself according to its own best judgement, towards ends which it chooses for itself. Ideally, we can therefore participate in a 'realm of ends', i.e. a community in which our freely chosen rational ends converge harmoniously. These ends include people themselves, because Kant regards each individual rational being as an end in itself (46). He expresses this in another way by saying that while some things have a 'price', other things, such as rational beings, have a 'dignity'; 'Autonomy is the ground of the dignity of the human and of every rational nature' (54).

In order to make moral judgements that will be consistent with the realm of ends, we need to consider them from other people's points of view, considering their ends as well as our own. Taken to its logical conclusion, this point leads us the basic principle of morality that Kant proposes, which he calls the categorical imperative: act only on those rules that you can rationally will to be universal laws. (37)

By categorical, Kant means that this principle isn't a means to an end ('you ought to act this way because it will bring about such-and-such'); rather, it is the fundamental justification of all moral ends. Kant makes it clear that, in his view, this is the only categorical imperative, and the only moral principle that can be determined a priori (without empirical grounds). The categorical imperative amounts to a litmus test that can be applied to potential rules of conduct; however, such rules can only be formulated in the context of particular empirical circumstances.

For example, suppose I want to make a promise, knowing that I won't keep it, because keeping it would be inconvenient. (39) If I formulate this as a universal law ('one may always make false promises in order to avoid inconvenience'), it becomes clear that this would be unworkable: nobody would ever be able to trust anyone else's promises. However, it does not follow that false promises are never justifiable: there might be some other rule, based on another (perhaps more specific) set of circumstances, that could justify making a false promise. Far from presenting evil as 'identifiable a priori' (as Badiou claims), Kantian ethics actually gives us no way to determine whether a particular act is wrong. It only helps us determine whether an act would be wrong if performed for a particular reason. In order to determine conclusively that there was never any acceptable reason to make a false promise in any circumstances, one would have to consider the infinite set of all possible reasons, and all possible circumstances -- clearly an impossible task.

The Kantians whose essays are included in this edition of the Groundwork emphasise this point. The Kantian Allen W. Wood, the translator and editor of this edition, sums up the limitations of the categorical imperative:

If the main issue about Kantian ethics were whether the universalizability test for maxims [intentions to act in a certain way for a particular reason, in a particular set of circumstances] is a satisfactory universal algorithm for all moral deliberation, then the right assessment of Kant's ethical theory would be that it is pretty worthless. There are several reasons why the universalizability tests are not up to playing any such role. First, since the tests are suited only to testing individual maxims, one by one, for permissibility, they can never yield anything like a positive moral rule or duty. The test never enables you to conclude that suicide or making false promises is wrong, but only that it is wrong to perform these actions on the specific maxims tested in the Groundwork; for all the universalizability tests might ever show, there might be other, wholly universalizable maxims on which acts of suicide or lying promises could be made. Second, the universalizability tests are notoriously subject to false negatives -- entirely innocent maxims that nevertheless fail the tests. Take the maxim that whenever I arrive at a doorway at the same time as another, I will always yield right of way to the other and will never go through first. There is nothing wrong with that maxim, but it could never be adopted by all agents as a universal law. The problem here is simply that there are many maxims that cannot themselves be made (or willed as) universal laws (or laws of nature) but also do not violate universal moral laws (on any plausible conception of what these might be). Third, the universalizability tests are also threatened with false positives -- maxims that pass the tests buy are impermissible on any reasonable moral view. The most basic reason for this problem is that any action is intentional (or follows 'maxims') on many different descriptions and at many levels of generality or specificity. The universalizability tests by themselves provide no criterion for determining how, or at what level of specificity, the maxim is to be formulated, or (therefore) which intentional features of the action are relevant to its moral evaluation. As Hegel correctly put it, the universalizability test would be fine if we already had determinate moral laws, but the test itself can never generate any such principles. (172-3)

As Wood points out, nothing in Kantian ethics amounts to a 'universal moral decision procedure or algorithm' (172), a sort of 'ethical sausage machine' for 'grinding out what any agent whatever should do under any conceivable set of circumstances' (174).

Kant gives four different formulations of the categorical imperative (xviii):

First formula:

FUL

The Formula of Universal Law: 'Act only in accordance with that maxim through which you can at the same time will that it become a universal law'

with its variant

FLN

The Formula of the Law of Nature: 'So act as if the maxim of your action were to become through your will a universal law of nature '

Second formula:

FH

The Formula of Humanity as End in Itself: 'Act so that you use humanity, as much in your own person as in the person of every other, always at the same time as end and never merely as means'

Third formula:

FA

Formula of Autonomy: 'the idea of the will of every rational being as a will giving universal law' or 'Not to choose otherwise than so that the maxims of one's choice are at the same time comprehended with it in the same volition as universal law'

with its variant

FRE

The Formula of the Realm of Ends: 'Act in accordance with maxims of a universally legislative member for a merely possible realm of ends'

Wood concludes:

If we rest our theory of duties on FH, as Kant actually does in the Metaphysics of Morals, then we have to admit that the fundamental principle of morality yields no universal decision procedure for all cases. Moral deliberation generates moral rules applicable in particular cases only in a loose fashion, leaving (as Kant himself says) considerable 'play-room' for individual discretion and judgement. It will also depend heavily on what Kant called 'practical anthropology', that is, on our fallible, constantly changing, and always deeply problematic knowledge of what human beings and the human predicament are like (both in general and under specific social and historical conditions). (174)

Let us turn now to Badiou's claim that Kantian ethics defines man as 'he who suffers' and as 'the being who is capable of recognizing himself as a victim'. This is clearly not the case. First of all, Kant does not refer to suffering at all in the Groundwork; nor does he refer to anyone's status as a victim. Far from equating man with 'his animal substructure', or reducing him 'to the level of a living organism pure and simple', Kant paints quite a different picture. For Kant, human being is:

• a creature endowed with an autonomous rationality, who ought to obey only the laws that he creates rationally for himself
• an end in himself, and never merely a means
• a universally legislative member in a possible realm of everyone's rational, freely chosen ends

Some examples of the consequences of this view, for Kant:

...the one who has it in mind to make a lying promise to another will see right away that he wills to make use of another human being merely as means, without the end also being contained in this other. For the one I want to use for my aims through such a promise cannot possibly be in harmony with my way of conducting myself toward him and thus contain in himself the end of this action....

...in regard to the contingent (meritorious) duty toward oneself, it is not enough that the action does not conflict with humanity in our person as end in itself; it must also harmonize with it. Now in humanity there are predispositions to greater perfection, which belong to ends of nature in regard to the humanity in our subject; to neglect these would at most be able to subsist with the preservation of humanity as end in itself, but not with the furthering of this end....

...as to the meritorious duty toward others, the natural end that all human beings have is their own happiness. Now humanity would be able to subsist if no one contributed to the happiness of others yet did not intentionally remove anything from it; only this is only a negative and not a positive agreement with humanity as an end in itself, if everyone does not aspire, as much as he can, to further the needs of others. For regarding the subject which is an end in itself: if that representation is to have its total effect on me, then its ends must as far as possible also be my ends. (47-8)

It is now easy to show that Badiou's other two critiques of what he sees as Kantian ethics are unfounded. Far from attempting to found an ethical consensus on the recognition of evil, as Badiou claims, Kant attempts to found such a consensus on the need to harmonize one's ends with those of others, in order to further every human being's happiness as an end in itself. Badiou claims that Kantian ethics provides no 'positive idea of the Good'; Kant's concept of a 'realm of ends' is surely exactly such an idea. And far from presenting any a priori idea of evil, and therefore neglecting the specificity of situations, Kantian ethics in fact leaves us with no definite means of determining what is evil in general; rather, it proposes a type of moral deliberation which is always forced to consider what Badiou calls 'the singularity of situations as such'.

Badiou claims that his description of ethical orthodoxy is 'what is retained from Kant (or from an image of Kant, or, better still, from theorists of "natural law")'. It should be clear from the foregoing that, at the very least, none of this supposed orthodoxy has been 'retained from Kant'. Moreover, if the essays by Kantians in this volume are any indication, the type of ethics that Badiou disapproves of is not representative of recent Kantianism, either.
to top