Kant's Moral Psychology (II): Consciously Choosing Principles

In my last post I addressed some preliminary issues on the question of whether the findings of empirical psychology can legitimately be used to refute Kant’s moral theory, as Leiter and Knobe try to do. My contention is that they cannot be. This is not to say, of course, that no normative principles can or should be informed by empirical facts. We are natural beings, with natural desires and psychological processes. The specific normative considerations that inform our actions must have something to do with the facts about the beings we are. But this need not imply that normativity is wholly dependent on these natural processes. In this post I will argue that Kant’s claim that we act on maxims that we adopt is not an empirical thesis, and that we cannot take it as such without lobotomizing his moral philosophy.

Let me then begin by summarizing a point I raised in the comments on the last post. L&K are right to say that for Kant, “reason is the source of moral motivation” (2). But what Kant means by this is literal: it is reason, not any actual psychological process of reasoning, that is the source of moral motivation. We can draw an analogy to logic. Given two propositions of a modus ponens, I can provide the right conclusion. There are, of course, psychological and neuro-physical processes involved in my production of the conclusion, and these processes are open to various modes of empirical study. But the story we get through such means will be incomplete: the processes involved can explain why I gave an answer, but they cannot explain why I gave the right answer. And it seems like the logical form of the modus ponens—the way in which a right answer is rationally determined by the premises—has some explanatory role. If many more people are open to the possibility of logical truths that are immune from empirical refutations than to the possibility of similar moral truths, this may show that there are more skeptics about morality than about logic. But skepticism is not a reason to jump ship.

Now, moving on to L&K’s argument, which is roughly that recent studies strongly suggest that the major factors involved in influencing human behavior are heredity and environment, while upbringing and active adoption of principles play extremely minor roles. This, they claim, is good news for Nietzsche, bad news for Aristotle and Kant. Why is this bad news for Kant? Well, briefly, here’s how L&K gloss Kant’s moral psychology:

In the Kantian tradition of moral psychology, moral obligations are grounded in principles that each agent consciously chooses (6)

(1) agents impose moral requirements on themselves, and (2) these self-imposed requirements are motivationally effective. In order for the self-imposition of moral requirements to be genuinely autonomous it must presumably be a conscious process of self-imposition. And for these consciously imposed principles to be motivationally effective it must be the case that conscious moral principles are motivationally effective. (7)

we must presume that these consciously imposed moral “laws” have a substantial impact on behavior… On the Nietzschean view, by contrast, conscious beliefs play no such role in moral (or immoral) agency. People’s behaviors are determined not so much by their conscious beliefs as by certain underlying type-facts. (27)

One reason that I think the research L&K present poses no challenge whatsoever to Kant is just that it poses a challenge only to the Kantian view described here. But Kant held no such view; in fact, this view is highly improbable even before a single study is cracked open. L&K take Kant to be laying out a theory according to which human beings consciously impose moral principles on themselves and then consciously act on them. But keep in mind that the “principles” involved are what Kant calls maxims, which have the form “I will do X in case/in order to Y.” Typical Kantian examples of maxims are: “I will end my life in order to avoid prolonged suffering” and “I will make false promises in cases where I can profit by this.” These examples are well known. I think we need very few studies to realize that we do not, under normal circumstances, formulate such principles consciously before acting. We do have intentions. But cases in which those intentions are preceded by explicit formulations of principles, which we then use to guide our intentions, are incredibly rare. One could, of course, read Kant uncharitably, and simply say: “yes, his moral psychology is just obviously flawed!” But now consider the fact that Kant was much smarter than you (yes, whoever you are reading this, there is a pretty high chance that this is true). So maybe we shouldn’t take Kant to be giving us an obviously flawed thesis about the psychological processes by which people actually make decisions, but rather a thesis about how our actions are to be understood rationally.

Let’s say that you see Freddie promising to pay back a loan when, in fact, you know, and you know that Freddie knows, that he is actually planning to take the money and flee to Mexico, never to be heard from again. Of all the things going on in Freddie’s head, it is unlikely that the explicit maxim stated above is one of them. More likely, he simply wants to get some money so he can make a comfortable getaway, and he thinks that this is a good way to achieve that—and even these thoughts don't need to be explicit for his action to be intentional. But if you—or Freddie—wanted to see his behavior as an action, i.e., not as a process of being entirely pushed around by his motivational states, but rather as somehow rationally structured, then the maxim above would be a good way of formulating it. The maxim is implicit in the action, and to hold Freddie responsible for the action, we need to take him as acting on some sort of rational principle. But this is not the same as saying that this principle was consciously formulated and adopted by Freddie, and that this conscious action on his part was the mechanism involved in producing his behavior.

In fact, it is clear that Kant cannot mean—at least in general—that maxims are consciously chosen principles. This is clear from the fact that Kant believes that human beings choose what Kant calls their disposition (Gesinnung), which is the maxim from which all other maxims are derived; we adopt this maxim through an act of freedom. Since this maxim grounds all our other maxims, its adoption must precede all our uses of agency. And so the grounding maxim must actually be chosen prior to any rational action we might undertake in our lives; as Kant notes, we may represent it as innate. But of course we cannot choose an innate disposition consciously; and Kant is explicit that this disposition is chosen first in the rational order, not in the order of time. It should be obvious by now that Kant is giving us a metaphysical, and not an empirical theory. A theory, then, that can be attacked on metaphysical grounds, or through a rejection of metaphysics (preferably not one that simply encourages us to stick to science), but that it makes no sense to attack empirically.

We can confirm that Kant is giving us a metaphysical, and not an empirical, theory by his constant use of the locution “as if” (als ob). That is, what he is discussing is not how we are, empirically, but how we must see ourselves: we must see ourselves as if we have adopted these principles. This, on Kant’s view, is how we must see ourselves if we are to see ourselves as persons, and not just as complex mechanisms. And yes, this marks Kant as clearly opposed to the naturalism so dominant in contemporary analytic philosophy, and of which Leiter is so fond. But it is question-begging to simply adopt a naturalist position and, from there, argue against Kant. One must first prove that a thoroughgoing naturalism is both superior and justifiable. Anything else is just bad philosophy.

I want to conclude by asking why, if Kant is so very obviously not giving us an empirical moral psychology, do contemporary philosophers (many Kantians included) so often interpret him precisely as suggesting a model on which we consciously choose principles and then act on them? One reason is that the thesis that we can endorse our action, and which in some form Kant really does hold, is easily confused with the thesis I am considering here (I will look at this in the next post). Another lies in the philosophical culture. Rawls, who made Kant respectable in analytic moral philosophy after decades of groundless slander and neglect, explicitly bracketed the issue of moral psychology: he was only looking for a decision procedure for ethics. But his students wanted a Kantian moral philosophy that, first, would give us concrete principles for action and, second, could do so without making any (to them) exorbitant metaphysical claims. Onora O’Neill for example, interpreted maxims as psychological states. Korsgaard, meanwhile, has advanced what many have praised as Kantian ethics without metaphysics. But these projects are much like riding a bicycle without the wheels. Kant’s moral philosophy is metaphysical; the validity of the moral law is intimately bound up with issues of transcendental freedom and noumenal causation. Without these features, one has a radically weakened Kantianism, and one that is not simply vulnerable to empirical refutation, but is almost self-evidently false.

In my next post I want to address a few other features of Kant’s moral psychology, and question L&K’s claim to have provided arguments against Kant. In particular, I will say something about the role of determinism in Kant, as well as his almost wholesale acceptance of Hume’s view that only the passions can determine us to action, two features that are too often overlooked. And I will address the issue of endorsement, which does involve an empirical claim, one that might look very similar to the view that agents consciously adopt principles.