There’s an interesting argument in Immanuel Kant’s Critique of Practical Reason. Moral theories, he claims, all ultimately rest on three postulates: God, freedom, and immortality of the soul. If one wants to believe in and pursue the moral life, one ought to also accept these postulates as the a priori conditions for doing so.
Why are these specific ideas so critical to morality? It helps to start with some understanding of Kant’s own approach to ethics. In his view, morality is grounded in pure practical reason, which means that an action is morally good or morally bad not in light of its consequences – as utilitarians famously argue – but in light of the maxims on which it is based. A maxim is moral if it can be reasonably endorsed as a universal principle. This is what Kant calls the Categorical Imperative. We should act according to a maxim only if it can be universalized.
The trouble here is that moral actions aim at achieving the highest good, and yet there is no guarantee that living the moral life will bring happiness. Note that happiness for Kant doesn’t mean pleasure, but refers to the kind of eudaimonic view of happiness that sees it more as being about the well-being and harmony of human beings with the world at large. This is something I think Kant is right about, too, since it often does seem that the moral choice is not always the fulfilling choice for us. Some actions even appear to bring about an internal conflict whichever way they’re decided.
Here is where Kant brings God into the picture. If morality is not an illusion, then there must be a harmonious relationship between virtue and happiness. The cause of this relationship can’t be within the natural world because Kant already thinks of reason and the will as part of that world. Thus he locates the cause in the noumenal realm, and identifies it with God. From this stem the other two postulates. Immortality gives us the hope that the highest good is attainable even if it doesn’t seem to be so during this life, and freedom is, in essence, what allows us to exercise our will in the first place.
Kant’s argument is intriguing partly because it tries to accomplish something few other theistic arguments do. Instead of arguing for the existence of God, it argues for the necessity of belief in God, at least concerning morality. Even certain other well known moral arguments, like the one used by William Lane Craig, don’t shoot for this bold conclusion (Craig claims that objective moral values constitute evidence for God’s existence). Kant actually does not have much faith in the traditional philosophical arguments for God, considering them failures that could only deliver inadequate and underdeveloped ideas about first causes and greatest imaginable beings.
With the argument explained more in depth now, we can see two objections that won’t work against Kant’s version of the moral argument. First, there is no implication that atheists and non-theists cannot live moral lives. Kant doesn’t even claim that the highest good is what determines moral action, let alone belief in God. What matters for Kant is duty to the moral law. Those who don’t believe in God are still capable of following the moral law because, as you’ll recall, practical reason is what grounds morality. At best, Kant’s accusation is that atheists and non-theists are simply being logically inconsistent in not recognizing that the moral life should lead one to belief in God.
Another unsuccessful objection, then, would be to raise the Euthyphro dilemma. This can also be expanded to include any objection that generally demands a defense of the foundations or ontology of ethics. While some philosophers have thought that Kant’s moral philosophy is subjectivist, it seems fairly clear that even if this is not the case, he does not ground morality in the commands, will, or nature of God. This can perhaps be backed up by the fact that Kant expresses his view, in many places, that God cannot be known through pure reason, but can only be arrived at via practical reason as a postulate like we’ve been looking at here. One could respond to Kant by adopting a kind of error theory, though Kant’s argument really doesn’t involve a commitment to any particular theory of moral epistemology or moral ontology. The error theorist would only be successful in so far as they’d just reject the pragmatic approach Kant has to ethics.
How could an atheist address this argument if she doesn’t want to give up the moral life and yet doesn’t see the need to believe in God?
One response would be to dispute the way Kant talks about morality having the highest good as its aim. This is too tall of an order, the unbeliever may argue. It’s fine to have that as an ideal, but Kant’s postulates don’t just treat the highest good as an ideal. What he envisions is a world with the greatest possible apportionment of virtue and happiness, and this is such a lofty idea for Kant that he has to come up with immortality to make it more palatable for us. On the other hand, if we are content to think of the moral life as only a duty of striving for the next best possible thing to the highest good, then we might well doubt there’s a need for believing in immortality and God.
In fact, if our starting point is naturalism, we can raise serious doubts against the prospects of attaining a world like that envisioned by Kant. With no help coming from any supreme being, we shouldn’t expect there to be a harmonious relationship between virtue and happiness. Agnostic philosopher Paul Draper has developed a version of the problem of evil that is strikingly similar to this, where he argues that the facts about biological pain and pleasure are so unrelated to moral obligations and the like that this gives us prima facie good reason for believing the universe is not the product of an all-good god, but is instead indifferent to suffering. Of course, Kant would reject these sorts of claims, and likely say that moral duty comes prior to beliefs based in empirical evidence. However, that is a contentious view in its own right.
A stronger response to Kant’s practical moral argument takes issue with his entire project for grounding morality. Alasdair MacIntyre makes this criticism in his book After Virtue. Some, he points out, have proposed to ground ethics in desires, while others have grounded them in religious beliefs. These projects could not establish an objective ground of morality, and neither can Kant’s project. We can conceive of a number of universalizable maxims that may be immoral and yet contain no apparent inconsistency. Kant himself struggled with identifying the inconsistency in the maxim, ‘To kill myself when the prospects of pain outweigh those of happiness.’ This violates an impulse to life in each of us, he argued, venturing into uncharacteristically ambiguous territory.
To be fair, we might revisit Kant’s imperative in other terms that he’s also well known for employing. Are things any better if we think along the lines of treating people, including ourselves, as ends rather than as means? MacIntyre thinks not. There’s debate about whether or not this alternative phrasing really follows from the idea that we should act according to a maxim only if it can be willed as a universal principle. What the new version seems to mean is that we should treat people as people, possessing their own will, instead of treating them as objects or instruments on which we exert our will.
Kant goes into some detail about how his moral philosophy follows from practical reason. To be rational, the moral law must be the same for everyone, and it must be binding on everyone. Intent is what’s important here, too, not the mere ability to obey, and so we have a fairly good idea that finding the rational ground of morality will mean finding the right kind of rational test to tell the difference between maxims that express the moral law and those that don’t. Reason just does work out principles that are universal, internally consistent, and categorical.
Yet MacIntyre observes that we get nothing really like this from Kant in defense of his maxim to treat all human beings as ends in themselves and not as means to an end. Introducing ideas about our internal ‘impulses’ would seem to walk a blurry line bordering on emotivism. And remember that Kant specifies that the maxim of treating human beings as ends must include oneself. Would it be far fetched to suppose this is because the Categorical Imperative taken alone doesn’t provide us any allowance to such things as our own personal intuitions, desires, emotions, etc? Its goal is, after all, universal applicability.
MacIntyre goes so far as to show that the maxim stating to ‘Let everyone except me be treated as a means’ is not inconsistent. Inconvenient, yes, but not inconsistent.
It’s an interesting question why there should need to be a ground of morality to begin with. Even Christian philosophers like Richard Swinburne tend to favor a more ‘first principles’ sort of approach to meta-ethics, where ethics proceeds from certain axioms or starting points, rather than being founded in something outside of it. I don’t think this makes morality illusory any more than it makes the external world illusory. Radical skepticism and the related quest for sure foundations may just be unwarranted.
Provocative as it is, Kant’s moral argument still fails to make the case for the necessity of belief in God to a consistent pursuit of the moral life.