blog.kennypearce.net

Morality as a System of Assertoric Imperatives

I recently read Philippa Foot's paper "Morality as a System of Hypothetical Imperatives" for an ethics class. The paper, as the name suggests, puts forward the view (which Foot has since rejected) that the imperatives of morality are merely hypothetical and not, as Kant had argued, categorical. What this means is that morality tells us how we should act if we want certain things, such as justice and the general happiness of humanity. As Foot recognizes, an untoward consequence of this view is that, if it is true, we can't sensibly tell people that they should want justice or the general happiness if they don't. (Of course, if we want justice and the general happiness then we should try to get others to want those things too, but that doesn't mean we can tell them that from their perspective they should also want those things.)

Interpreters of Hobbes often attribute to him a similar view, namely, that morality tells us what to do if we want to live and flourish. Thus morality, on this reading of Hobbes, is completely indistinguishable from prudence.

Now, this terminology of categorical vs. hypothetical imperatives is due to Kant, and whenever I hear these sorts of discussions, I wonder why people don't bring up another closely related Kantian distinction. Here's some text:

[T]he hypothetical imperative says only that the action is good for some possible or actual purpose. In the first case it is a problematically practical principle, in the second an assertorically practical principle. The categorical imperative, which declares the action to be of itself objectively necessary without reference to some purpose, that is, even apart from any other end, holds as an apodictically practical principle.
...
There is, however, one end that can be presupposed as actual in the case of all rational beings (insofar as imperatives apply to them, namely as dependent beings), and therefore one purpose that they not merely could have, but that we can safely presuppose they all actually do have by a natural necessity, and that purpose is happiness. The hypothetical imperative that represents the practical necessity of an action as a means to the promotion of happiness is assertoric. It may be set forth not merely as necessary to some uncertain, merely possible purpose but to a purpose that can be presupposed surely and a priori in the case of every human being, because it belongs to his essence. Now, skill in the choice of means to one's own greatest well-being can be called prudence in the narrowest sense. Hence the imperative that refer to the choice of means to one's own happiness, that is, the precept of prudence, is still always hypothetical; the action is not commanded absolutely but only as a means to another purpose. (Kant, Groundwork for the Metaphysics of Morals, tr. Mary Gregor, AK 4:414-416)

The problematic/assertoric/apodictic trichotomy is explained in the Critique of Pure Reason:

Problematic judgments are those in which one regards the assertion or denial as merely possible (arbitrary). Assertoric judgments are those in which it is considered actual (true). Apodictic judgments are those in which it is seen as necessary. (tr. Guyer and Wood, A74-75/B100)

So one class of hypothetical imperatives are those which refer to the various ends that rational agents might have. Kant refers to these as imperatives of skill or technical imperatives. Assertoric imperatives are those which refer to the ends that the agents we are concerned with do (contingently) have. The categorical imperative makes no reference to ends at all because it necessarily applies to every rational agent to whom imperatives can apply, regardless of its chosen ends. Now, we usually won't communicate an imperative unless we believe that our interlocutor actually has the relevant end. However, in the cases of technical imperatives, if someone were to simply deny having the relevant end, we would withdraw the imperative. According to Kant, what makes the imperatives of prudence assertoric is not that some people actually have those ends, but rather that they refer "to a purpose that can be presupposed surely and a priori in the case of every human being, because it belongs to his essence." It is not categorical, because there could be rational beings who did not desire their own happiness. However, those beings would not be humans.

Now here's my point: Foot, in her early work, defends this really bad view, according to which someone can get out of morality by simply claiming not to want justice, the general happiness, etc. People can even want injustice and general misery. We'll be able to describe them as unjust or wicked, but we won't be able to tell them they shouldn't be unjust or wicked. There is no motivational force for them.

Hobbes is in better shape, precisely because his imperatives are assertoric. Hobbes attempts to support his theory of morality by what we would now call a theory of psychology, which allows him to say that human beings want certain things (to live and flourish). He then presents a moral theory which consists of hypothetical imperatives conditioned on our want those very things.

Now, to be perfectly clear, I think that both Foot's and Hobbes's proposals are ultimately hopeless. However, Kant's distinction, in my view, represents an easy way to improve Foot's theory. Consider the following view:

All human beings implicitly want to live in a just society. However, every individual unjust action contributes to the injustice of society. So people shouldn't commit acts of injustice.

Now, this seems like a pretty plausible line of reasoning to me. Even if we want to commit individual acts of injustice, we surely don't want to live in a society where such acts are the norm. Similar stories might be told for other moral virtues. If the premises about what we all (implicitly) want could be justified by empirical psychology or other means, then we shouldn't be convinced by claims not to want these things, and so shouldn't let people off the hook for immoral behavior.

People not getting off the hook so easily is one reason why this view is an improvement over Foot's. There is another reason: it is highly implausible to say that morality's authority is derived from the actor's desires. Whether or not I want something just doesn't seem to figure into my moral obligations. However, it is not nearly so implausible to say that morality for humans is grounded in the nature of humanity. (In fact, I think we need something like this, which is why I think that Kant and Aristotle have the two most plausible moral theories with which I am familiar - Aristotle's jumps off from what it is to be human, and Kant's jumps off from what it is to be a rational agent. Of course the theories turn out quite different.) That this is a reasonable analysis of ordinary moral discourse is further supported by our practice of describing especially reprehensible acts as 'inhuman'. According to the theory under consideration, part of what it is to be human is to want certain things (e.g. to live in a just society); ordinary cases of immoral actions suggest that the agent is not very consistent or rational in pursuing these ends, but certain extreme course of action lead us to question whether the agent even has the desires at all, and so to describe him as 'inhuman'.

While I don't ultimately endorse this view, I do think it is far more plausible and interesting than the one Foot defended (I don't know what Foot's new view is), and all due to a little more attention to Kant's distinctions.

Posted by Kenny at April 13, 2010 6:08 PM