How Kant’s Flaws Are Misapplied Virtues

As Prof. White notes, much of the suspicion of Kant’s philosophy in libertarian circles is due to Ayn Rand’s severe condemnation. To be sure, Rand’s chief quarrels with Kant concern metaphysics, epistemology, and ethics, not politics; but Rand did think that Kant’s views in those more fundamental areas of philosophy had authoritarian political implications, even if Kant did not fully draw those implications himself. One of Rand’s students even devoted, with Rand’s approval, an entire book to blaming Kantian influence for the rise of Nazism in Germany.[1]

In a nutshell, the links Rand saw were as follows. In metaphysics and epistemology, Kant – according to Rand’s interpretation – denies that our minds can know reality; all we can ever know is a world of appearances constructed by our cognitive apparatus. In ethics, Kant demands that happiness be sacrificed to duty whenever the two conflict (and he does think they often conflict). The first view is an attack on human cognitive efficacy; the second is an attack on human happiness. (Deliberate attacks, in Rand’s view; when Kant goes wrong, it is through malice, not innocent error.) And those who have been taught that they cannot trust their own cognition and have no right to pursue their own happiness are intellectually disarmed against the claims of a totalitarian state.

I said in my initial essay that I thought Rand was partly right and partly wrong here. Let me elaborate.

Peter Strawson argues persuasively that there are two strands in Kant, one “austere” and one “transcendental.” The core of Kant’s “austere” project is to explore the “limits to what we can conceive of, or make intelligible to ourselves, as a possible general structure of experience”; this project Strawson considers legitimate and defensible. But in his “transcendental” moments, Kant mistakenly slides into thinking of those limits as a product of the “ways in which our experience is causally dependent” on “our own cognitive constitution.”[2]

In other words, Kant is on solid ground in seeking to investigate what picture of the world we can make sense of; his error lies in taking the results of that inquiry to be the product of our cognitive apparatus’ imposing certain structures on a reality that, but for that imposition, might have been structured entirely differently. This, as Strawson rightly notes, is incoherent; for if we cannot make sense of the world not having those structures, we cannot intelligibly entertain the possibility of the real world’s not having them, and so cannot intelligibly regard those structures as the mere products if our cognitive apparatus. This would be the mistake that Wittgenstein diagnoses as trying to think both sides of the limits of thought.[3]

The transcendental Kant is the Kant that Rand attacked.[4] And the transcendental Kant does indeed seem to undermine the reality and cognitive efficacy of the individual in a way that is politically troubling. But on Strawson’s reading, which I find persuasive, the transcendental Kant is not the most important Kant; rather the transcendental Kant represents Kant’s misunderstanding of his own project, which properly understood yields the austere Kant. (Now Rand wouldn’t like the austere Kant either; she’d find him too much of an a priori rationalist. But she wouldn’t regard him as an unmitigated disaster; after all, she had kind words for Spinoza.)

Thus the transcendental strand in Kant is a mistake, not a deliberate salvo against the human mind. Moreover, it’s not a stupid mistake or a mistake easily avoided; it’s all too easy to slide from identifying the limits of thought to seeing those limits as impositions on something that might otherwise be different.

With regard to ethics, while it’s true, as Dr. White points out, that on Kant’s view we have duties to ourselves, it’s still the case that Kant believes in the possibility (and the frequent, though not unvarying, actuality) of conflicts between duty and self-interest, and holds that in such cases we must choose duty, with self-interest being given no contrary weight at all. This is why Rand sees Kant as an extreme altruist.

Here, as in the case of his metaphysics and epistemology, I think Kant is sliding from a correct insight into a mistake.

The correct insight is that our motivation to do the right thing should not be conditional on how we happen to feel or what we happen to find strategically convenient. If the reason I respect other people’s rights is that doing so gives me a warm fuzzy feeling, then as soon as I become depressed and no longer get that fuzzy feeling, I can no longer be counted on to act rightly. If the only reason I behave cooperatively toward others is to incentivize them to behave cooperatively toward me, then as soon as I see a way of winning others’ cooperation without being cooperative in turn – say, by deceiving or defrauding them instead – I can no longer be trusted. A genuinely moral person’s commitments cannot be conditional in that way.

This much I think Kant is right about. Moreover, this is one of the parts of Kantian doctrine that is especially congenial to libertarianism of the natural-rights variety. Natural-rights libertarians often criticize consequentialist approaches to liberty for being unreliable in their commitments. As Murray Rothbard writes:

Another problem with the utilitarian is that he will rarely adopt a principle as an absolute and consistent yardstick to apply to the varied concrete situations of the real world. He will only use a principle, at best, as a vague guideline or aspiration, as a tendency which he may choose to override at any time. … To say that a utilitarian cannot be “trusted” to maintain libertarian principle in every specific application may sound harsh, but it puts the case fairly.[5]

But by placing duty at odds with self-interest, and demanding the sacrifice of the latter to the former in case of conflict, Kant arguably does downgrade the importance of human happiness, just as Rand charges. But there is a reason (other than malice) for Kant’s position. Just as, in metaphysics, Kant thinks that there are a priori constraints on what account of the physical world makes sense, so in ethics, Kant maintains that there are a priori constraints on what account of morality makes sense; the unconditional character of moral duty is one of those constraints.

So far so good. But Kant seems not to think that there’s much at all in the way of analogous a priori constraints on what happiness might be. Given that he holds to a fancy, highly constrained conception of morality and an un-fancy, rather unconstrained conception of happiness, it’s unsurprising that he thinks nothing can guarantee that the two will go in tandem:

If the doctrine of morals were merely the doctrine of happiness it would be absurd to seek a priori principles for it. … Only experience can teach what brings us joy. … But it is different with the teachings of morality. They command for everyone, without taking account of his inclinations …. He does not derive instruction in its laws from observing himself and his animal nature or from perceiving the ways of the world …. Instead, reason commands how men are to act even though no example of this could be found, and it takes no account of the advantages we can thereby gain, which only experience could teach us.[6]

This is where I wish to contrast Kantian ethics with the approach of the ancient Greek philosophers (in particular, Socrates, Plato, Aristotle, and the Stoics). To begin with, the Greeks draw a distinction between instrumental and constitutive means. As I’ve written elsewhere:

Suppose that I want to play the Moonlight Sonata; and so I save money to buy a piano, and to buy sheet music, and to take piano lessons and so forth, so that I’ll be able to play the Moonlight Sonata. These are all means to the end of playing the Moonlight Sonata; if you ask me why am I saving this money, why am I buying a piano, etc., I would say these are all means to my ultimate goal, which is to play the Moonlight Sonata.

But now suppose you come upon me in the middle of playing the Moonlight Sonata, and I’m hitting a particular note. And you ask me: “Why are you hitting that particular note? Is it just that you find that note valuable in and of itself?” And I would answer: “No, I’m playing that note because I want to play the Moonlight Sonata, and I can’t play the Moonlight Sonata without playing that note at that point.” Well, in a sense, then, playing that note is a means to playing the Moonlight Sonata; but it’s not a means in the other way. It’s not a means that’s external to the end; it’s a means that’s part of the end. …. So there are cases where a means can be a constitutive part of the end rather than being an external means to it.[7]

The Greek philosophers declare each individual’s happiness to be the supreme good and appropriate end for that individual. And morality they declare to be a means to that end. But they take morality to be a constitutive rather than an instrumental (external, strategic) means to happiness – for happiness, on their view, is an objective state of being with objective requirements, not a subjective feeling. (Here, again, I have in mind Socrates, Plato, Aristotle, and the Stoics. The Epicureans are a more complicated case.)[8] Moreover, the objective constituents of happiness are determined by conceptual rather than empirical considerations.

For example, when Aristotle tries to show that “honor” – by which he means being honored, i.e., being respected and admired, or having a good reputation – cannot be the central component of happiness, he doesn’t try to show that honor will leave us subjectively dissatisfied; instead, he argues that honor violates certain a priori constraints on the nature of happiness, such as being an end in itself, or not being excessively outside of our control.[9] Aristotle would thus firmly reject Kant’s dictum that “only experience can teach” what brings us happiness. The Greeks are in effect doing for happiness and morality together what Kant is doing for morality alone.

The advantage of the Greek approach to ethics over the Kantian approach, as I see it, is that it allows morality to have all the unconditional and a priori character with which Kant seeks to invest it, but without severing the tie between happiness and morality. Moreover, the way in which Greek ethics takes determining the contents of morality and happiness to be a matter of mutual adjustment offers libertarians a manner of reconciling rights-based and pragmatic arguments for liberty.[10]

Kant does not seem to understand the Greek view of ethics. For example, he tells us that “[t]he ancients saw perfectly well that happiness alone could not be the greatest good” – with the exception of Epicurus, who “took away the value from virtue, in that he made morality into a means to happiness.”[11] Here Kant is clearly using “happiness” to mean a subjective feeling – which he is entitled to do in describing his own views, but it’s not a position he should be attributing to Plato, Aristotle, or the Stoics, who quite explicitly did declare happiness, objectively understood, to be the greatest good.

Thus Rand is right in championing Aristotle against Kant, and in criticizing the latter for devaluing human happiness. But Kant’s position here was neither a fiendish plot nor a dim-witted blunder; rather, it was the product of a laudable attempt to protect the unconditionality of moral commitment, combined with a failure to see how happiness could be included in the deal. Nor can Rand claim much superiority over Kant on this point, since she herself is notoriously unclear as to whether the relationship between self-interest and morality is instrumental or constitutive, causal or conceptual.[12] Despite her professed admiration of the Greek approach to ethics, Rand is like Kant in not fully finding her way to it.