1. Rationality as the Basis for Ethics

Deontology is dying. Most deontological systems of ethics, which assert that actions are right or wrong by their own nature rather than by their consequences, rely on the view that humans possess a certain kind of rationality, from which we can derive ethical claims. Kant's system famously depends on humans' status as rational agents, arguing that our rational nature logically implies that we are bound by the moral law, which creates a universal, or categorical, imperative.¹ To Kant, our rationality gives us agency, and allows us to impart value onto ends. Neo-Kantians, most prominently Christine Korsgaard, similarly depend on human rationality to ground the existence of ethical rules. To Korsgaard, our reflective capacities require us to endorse the justification for our actions before performing them.² She argues that as humans we require justification to act, necessarily believing that our actions are universally valuable to all humans through our capacity to legislate value as rational creatures.
But this focus on the rationality capacity of humans as a means of grounding deontological systems is misplaced. Even if we do not possess a rational nature, we can still realize through the definitional structure of the will that some actions are not ethically permissible.

1. The Deontological Approach

The basic approach of deontological ethics involves defining some maxim, and then deriving a contradiction from assuming that maxim's universal permissibility. Deontological systems most famously use such arguments to derive the universal impermissibility of actions such as lying or breaking promises. To determine which maxims are definitionally impermissible, we must determine what about a maxim allows us to derive its impermissibility. First, we can look at a common, albeit flawed, attempt to prove that murder is impermissible.

We will prove that murder is wrong. We shall define murder as having two necessary elements:

1: The intentional killing of another person, and

2: The lack of that person's agreement.

Assume, by way of contradiction, that some agent A maintains that murder is ethically permissible. From P1, to A, some agent B is also ethically permitted to murder agent A. But then, from P1, A would agree to being murdered. Then, B, by definition, could not possibly murder A. Thus, to maintain that "murder is permissible" is to believe a contradiction.

This argument falls short in that it has not shown a contradiction at all. In saying that B could not possibly murder A, we are showing that the impossibility of murder follows from the maxim, not that the maxim implies the impermissibility of murder. And we cannot logically equate impossibility with impermissibility.

To make this maxim impermissible by definition, we must adjust our definition of murder in two ways. We first must replace the second element with the following: the killing is against the will of the person being killed. Second, we then must shift our focus from the acts themselves and instead must formulate the ethical prohibition in terms of intent. We now have the following elements of murder: It includes the intent to both 1) kill another person and 2) violate that person's will.

We now have a standard which relies on an intent to violate one's will. To violate one's will is to act in a way contrary to their objectives, obstructing or preventing them from achieving what they desire. On a model of the will which makes a distinction between first-order desires and desires of a higher order, this will-violation maxim is implicated when an actor intends to clash with what they believe to be the true will of another. In other words, if an actor intends to violate some subject's will X₁, only because they believe the subject possesses a higher-order desire X₂, they do not implicate the will-violation maxim as the motivation on which they act aligns with what they perceive to be the true will of the subject.

1. Will-Violation Maxims and the Universalization Problem

We can now examine the definitional impermissibility of this revised murder maxim. Although this maxim has two elements, we need not refer to any element other than the will-violation maxim, as this alone is sufficient to derive the impermissibility of the maxim.

Using the specification of "the will" above, it is clear that one cannot will for something to occur which contradicts their own true will. Even if one's conscious will differs from their "true will," it must be because they do not recognize the nature of their true will, for if they did, it would become their conscious will. In this way, the will-violation maxim cannot be universalized. When a moral agent endorses such a maxim, they accept that they have no philosophical complaint if someone acts on this maxim, endorsing the very will which they necessarily oppose. Regardless of one's rational status, one necessarily wills what they will. Thus, if one wills that the will-violation maxim is permissible as a universal law, they necessarily will a contradiction: they simultaneously protest and consent to others intending to violate their will. Because the permissibility of will-violation maxims require a contradiction, they can be rejected.

Perhaps we would wish to say that one need not necessarily protest attempts to violate their will. Surely, some of those attempts will be unsuccessful, as some people may not be very good will-violators. But when we act, we assume that we are acting with an effective will. To intend something is to act with the goal of bringing about the object of that intention. By acting, we accept that the action is actually more likely to bring about that result, otherwise we would not act. We accept this as a feature of the action, without it we could not intend at all. Thus, when we accept the will-violation maxim, we maintain not just that acting with the intent to violate a will is permissible as a universal law, but that it is permissible in instances where the action is successfully completed which results in the actual violation of that will. The successful completion of the action definitionally necessitates the harm, and as such we cannot endorse the action without endorsing its effect.

1. Distinguishing Belief Maxims

Relying on the definitional contradictory implications of such maxims leads to conclusions which diverge from Korsgaard's. For instance, consider the case of a driver who is late to work, and runs over a pedestrian without concern for whether they harm that person. Such an action does not violate the will-violation maxim, as it lacks the intent to harm.

Acting with the belief that we may cause harm does not lead to a contradiction. When acting with a background belief that we might cause harm, we do not assume the correctness of that belief, unlike with issues of intent, where we do assume the efficacy of our will when acting. The correctness of their belief is not a feature of their action or a condition of its success, and thus a completion of that action does not involve the violation of another's will. In this way, it is logically possible that we can will that someone act while holding a background belief that they are causing us harm. One who knows they are harming another is, or at least may be, acting consistently.

I must note that accepting this conclusion does not imply that we cannot construct laws to protect the pedestrian. Laws exist not only to criminalize provably unethical action, but because they seek to prevent harm. A society would strive to mitigate the impacts of earthquakes and hurricanes without condemning them as immoral. However it does suggest that we should not be criminalizing such actions on retributive grounds, nor do we have a basis to condemn the actions as objectively wrong. People who have committed manslaughter did not act in order to harm someone else, they acted with priorities inconsistent with the state's-they prioritized their own interests over someone else's. But because these interests are subjective by their nature, we use government to respond with measures of containment rather than punishment.

The conclusions which arise from distinguishing intent and belief in this way become more intuitive with less extreme cases. Imagine that one enjoyed playing the piano, but they also were aware that their neighbor did not like the sound of them practicing. We would likely say that that person would be permitted to keep playing the piano, even knowing that they might be harming their neighbor. To believe otherwise would be to require people to essentially never act, because they know that virtually every action carries some risk of harm. There is no logical reason that such actions, even if they are likely to cause clear harm, are necessarily impermissible.

There of course exists a further debate about when one intends an outcome. In many instances, the violation of another's will is a means in producing the intended result. One could act because of a proxy indicator of pain, for instance someone who ties the bull before a bull riding show. They need the bull to act in a way that results from a violation of its will to not feel pain, but that pain is not the end itself. Although this pain is almost certainly causally involved in the intended result, it is not logically necessary to the action, thus we cannot say that the harm is intended. Similarly, if one plays a game of chess, they are likely frustrating their opponents' will when pursuing the end of checkmating them, but they do not act in order to frustrate the true will of their opponent. Surely, if their opponent actually was making moves at random, with no will to win, one would still defeat their opponent for other reasons, such as the satisfaction of winning or achieving a favorable result if we were playing in a tournament. Even if I would prefer that my opponent genuinely attempt to win the game, I can endorse their intent to win the game, as even though it contradicts my objective in the game, it does not necessarily conflict with my true will outside of the game, which may be concerned with, for instance, promoting an environment where chess players genuinely seek to win their games.

This definition of intent does not however merely render prohibitions of the will-violation maxim only true vacuously. Admittedly, technically speaking, someone else's suffering itself is never what directly motivates us to act as we are always motivated by some other end tied to our own emotional state; we are motivated instead by our perception of their emotional state. But to say that we intend to harm that person, implicating the will-violation maxim, is simply to say that our satisfaction from performing the action is logically dependent on our belief that that person is actually suffering. We intend to cause them harm because we value the outcome of causing what we believe to be harm. The chess player and the bullrider do not achieve their intended result because they believe that they have successfully violated the will of another. If the chess player was only playing to make her opponent suffer, she would be acting unethically. The problem arises when the action which we are endorsing is one aimed at actually producing that harm, as the belief that the actual harm will occur is necessary to form the intent and thus motivate the action.

This definition of intent aligns with our intuitions when considering issues like meat consumption. We object when someone abuses animals for some sadistic desire to make the animal suffer. We would likely socially relegate such a person and feel moral outrage toward their actions. But simultaneously we, generally, do not feel such outrage when someone purchases a steak, even when an animal's death is necessary to produce that meat. Because they do not purchase that steak in order to produce suffering, it does not produce the same moral reaction.

1. Reflections and Intuitions

In these ways, this system simultaneously is stronger and weaker than our intuitions would predict. But this feature creates an epistemic foundation that is logically distanced from commitments to humans' capacity for rational agency. Deontology, particularly when relying on definitional contradictions in maxims, sets forth to derive ethical truths with epistemic certainty rather than to holistically explain or defend the values which we hold more broadly. Doing otherwise leaves room for the skeptic to challenge value systems and metaphysical claims thereby undermining deontological ethics. Distinguishing what can be grounded in the logic of the maxim itself epistemically insulates a deontological foundation, preserving its ethical claims in the face of changing cultural values, an evolving understanding of metaphysics, and neuroscientific developments.