W. Clark Wolf (Marquette University), "Kant’s Formula of Universal Law as a Test of Causality"
The Philosophical Review, 2023
It’s interesting how spontaneously we tend to put a check on our claims, and even our actions, by putting them through a test of universalization or generalization. An easy way to challenge an act of littering would be to consider: what if everyone did that? We might also determine legal validity in one case by considering how the law could be applied in similar, possibly fictional cases, thus making explicit what it would mean for a law to be applied universally. For example, suppose a conservative Christian baker won’t make a cake for a gay couple – if that’s ok, what if an antisemitic baker wouldn’t bake a cake for Jewish couple? It may be difficult to cash out precisely why this kind of question is relevant in our evaluation of actions, but it seems clear that some sort of universalization test is compelling. Interestingly, we might use a similar universalization test when we examine claims about explanation or causality. Imagine you are sick with Covid. A few days in, you eat an unusually large bunch of broccoli; soon after, the Covid clears up. Was the broccoli your cure? Probably not! But how do we know? It’s plausible that we implicitly rely on an assumption like this: if broccoli could cure my Covid, it is because broccoli in general is a cure of Covid. I should not think that the broccoli cured my Covid unless I could think that it would also cure the sickness of anyone else who is relevantly similar to me. Just as in our practical life, we use some form of generalization test to check on causal reasoning.
Immanuel Kant is perhaps the most famous advocate of the view that morally worthy action depends on a universalization test, specifically the first form of his “categorical imperative” called the “Formula of Universal Law” (FUL). The FUL states that we should “Act only on a maxim that you can at the same time will to be a universal law.” However else it may need to be clarified, Kant expects that this formulation will ring true with our ordinary concept of morality – our own tendency to assess ourselves and others by a similar test, perhaps much like the various “Golden Rules” found throughout many cultures. But the intuitive pull of this formula doesn’t explain why it is supposed to be justified. Why is a universalization test relevant to the moral worth of our actions? In a recent paper, “Kant’s Formula of Universal Law as a Test of Causality” (The Philosophical Review, 2023), I argue that Kant’s formula shares its justification with the kind of causal reasoning I sketched above. In the case of Covid, we want to know if the broccoli really was the cause of the cure; in the case of morally worthy action, we want to know: would my reason or principle really be the cause of my action? In other words, Kant is considering one’s maxim (the reason or principle on which one acts) as the apparent ”ground” or cause of one’s action. However, if Kant assumes that causes must be universalizable even to be causes, we cannot assume by default that each maxim can be the cause of its accompanying action. Instead, just as in other cases of causes, a maxim cannot really be the cause of one’s action (or what Kant calls its “determining ground”) unless there could be some universal, law-like relation between that cause (here, the maxim) and the effect (here, the kind of action).
Before spelling this out a bit more, I should note how this view differs from other familiar readings of Kant’s FUL. A dominant approach is to treat Kant’s FUL test as a test of moral permissibility. On this view, if I cannot treat my maxim as a universal law, then acting on that maxim would be morally wrong; if I can treat my maxim as a universal law, then acting on that maxim would at least be morally OK, or permissible. This reading seems to imply that we could use the FUL to answer moral problems, like ‘Is it every ok to lie?’; ‘Is it really wrong to kill?’; ‘Can I litter if I’m in a hurry?’ Now, it hasn’t been easy to make Kant’s FUL work to answer these questions; and even fans of Kant are not convinced that the test is up to this task. In addition, it’s not quite clear what the supposed connection is between universalizing a maxim and showing it to be permissible. This problem is expressed in the common reaction to ‘What if everyone acted that way?’ scenarios by the blunt response: but they won’t! Even if this is a simplistic dismissal, it is not at all obvious why the universalizability of a maxim is necessary for its permissibility, unless universality is simply standing proxy for another moral value (like fairness or impartiality). But there’s something else odd about the permissibility reading from an interpretive point of view. It seems to be answering a question that Kant does not raise, or at least not in the first instance. Anyone who’s learned a bit of Kant’s ethics knows that he begins the Groundwork by discussing the figure of the “good will” – the will that is good because of the reasons on which it acts, regardless of the consequences of its actions. Kant’s whole emphasis is on something like the motivation of the good person. It is that same motivation that is at issue when Kant talks about acting “from duty.” The shopkeeper who doesn’t cheat his customers surely does the “right” or permissible thing – but he doesn’t act from duty if all he cares about is increasing his profits. Just thinking about this example, one can see that it would be odd if Kant’s FUL – which is meant to fill in how the “good will” acts – would teach us that the maxim to cheat customers is a bad or impermissible one. No kidding! But that would fail to address the question that seems to interest Kant: what motivates the person who decides not to cheat his customers ‘for the right reason’? When Kant speaks about “morality” (in contrast to what he calls “legality”), he is interested in good motivation and not just right action. And speaking of “motivation” is speaking of an action’s cause; in short, the “good” kind of action will be an action with a certain kind of cause. This leads me to argue that Kant’s universalization test is precisely used to sort out the causality of morality worthy action.
The argument works in three basic steps. I first show that Kant’s requirement that causes or “real grounds” must have a law-like relationship to their effects is universal in its application: it applies to grounds whether they are “phenomenal” or not. Secondly, I show that Kant does indeed treat the will as the ability to have reasons as causes. Combined with the first step, this shows that the reasons of the will must stand under laws, but more specifically, causal laws. Finally, I work out the causal structure of the will in more detail, including the role of representation therein. I aim to show that the FUL takes the shape it does precisely due to constitutive requirements on maxims being causes at all.
If this is correct, then Kant is not relying on any special assumptions about moral or practical principles to justify his “universalization requirement.” His assumptions depend only on what it would mean to have a will in general – to be able to act on the basis of reasons at all. In the paper, I call the constitutive principle of a will in general the categorical declarative. It goes something like this:
A person’s maxim is the cause of her action only if there is a lawlike relation between the maxim and the action, and the subject represents that law-like relation as the cause of her action.
I try to show that the categorical declarative stems from two sources: first, the general requirement that causal relations be universal (no different here than in the case of broccoli and Covid); second, the special addition in the case of rational beings, who need to represent their reasons as causes for them really to be causes.
Importantly, the categorical declarative is not itself a normative moral principle; it just tells us what it means to act for a reason. In effect, Kant’s FUL (the categorical imperative) turns the ‘descriptive’ principle of the categorical declarative into a moral or normative one. But, for my purposes, what is important about this is that the requirement that we should ‘universalize’ our maxims doesn’t need a special moral justification. It falls out of something required for our reasons really to be causes at all. This implies that if one puts a maxim into Kant’s FUL, and it proves to be impossible, this shows not just that it would be wrong to act on that maxim, but that such a maxim could not even be the cause of that action.
Kant makes this point quite clearly: “If we now attend to ourselves in any transgression of a duty, we find that we do not really will that our maxim should become a universal law, since that is impossible for us, but that the opposite of our maxim should instead remain a universal law, only we take the liberty of making an exception to it for ourselves (for just this once) to the advantage of our inclination” (GMS 424, underlined). In a passage like this, it seems clear that the FUL isn’t merely sorting actions morally into the good and the bad; it is making a judgment on what could really be moving us at all. If our maxim cannot pass the test, the maxim can’t be playing the role of determining cause; there must be some other causal explanation of our action. The way Kant often puts it is that, in the “bad” cases, our maxims are really “conditioned” or “mediated” by some other source of causality – typically our inclination. Here, the inclination provides the “ground” of action, so we should not look to the law-like relation of our maxims to our actions to explain them. Since for human beings, this difference in our motivation for acting is always possible, the FUL functions as a causal test. Serving this role, I argue in the paper, is not equivalent to sorting actions into the right and wrong.
This might make Kant’s view seem less interesting for deciding hard ethical questions, where our question often is whether some action is right or wrong. Perhaps so. (Though I think many of these questions would fall under the Doctrine of Right for Kant, and so concern not so much our morality in a strict sense but instead our legal obligations to others in a free society). But what I find promising about this reading is that it gives us something like a way of expressing the nature of morally worthy action in a non-normative vocabulary: if Kant is right, we can distinguish the morally worthy from the rest by appealing to a difference in the way the actions are caused (or, indeed, the way we cause them). This looks like the basis for a “metaphysics of morals,” which is what Kant advertises after all.
I hope this gives you some glimpse of the paper and its basic argument in a simplified form. If you are interested in the topic (or perhaps looking for a new way to teach Kant’s puzzling moves in the Groundwork), I hope you’ll check out the full paper (here is a penultimate version).