The Categorical Imperative Through the Veil of Ignorance: A Solution to Problems in the Contradiction in Will Test
 

I. Introduction

 

Central to Kant’s moral philosophy is the Categorical Imperative (CI), which is an unconditionally binding imperative to act in a moral way. Kant’s moral philosophy contains four formulations of the CI: the Formula of Universal Law (FUL), the Formula of Humanity (FH), the Formula of Autonomy, and the Kingdom of Ends. In Groundwork for the Metaphysics of Morals (Groundwork), Kant defines the FUL as follows: “I ought never to conduct myself except so that I could also will that my maxim become a universal law” (Ak. 4:402).¹ He also defines the FH as follows: “Act so that you use humanity, as much in your own person as in the person of every other, always at the same time as end and never merely as means” (Ak. 4:430). Much work in the area of Kantian moral philosophy has discovered problems with the FUL, specifically that it yields incompatible and incorrect judgments about the permissibility of maxims compared to the other formulations.

​​

I will first explain the FUL as Kant defines it in Groundwork for the Metaphysics of Morals and how Kant derives both perfect and imperfect duties from it. Then I will assess the literature: Korsgaard argues that the FUL can be used to justify lying to a lying murderer who is trying to kill your friend. Herman argues that the maxim corresponding to convenience killing does not create a contradiction in conception (CC) when analyzed with the FUL, which she takes to entail that people have the freedom to sometimes kill for convenience. Kleingeld disagrees with Herman’s interpretation of the FUL and argues that this freedom, what she calls the “latitude problem” does not exist because it rests on a faulty interpretation of what kind of freedom is entailed when a maxim does not present a CC when universalized. Next, I will introduce the theft maxim to demonstrate two key problems with the contradiction in will (CW) test. The first problem is an apparent lack of universality to the CW test and its reliance on personal attributes, which undermines the objective status Kant places on the CI. The second problem is that individuals in special positions can correctly apply the CW test to reason that it is permissible to act on maxims that should be impermissible to act on. I will argue for a solution to these problems by borrowing concepts from John Rawls’s work, A Theory of Justice. I will show that these two problems can be resolved if people using the test to evaluate a maxim imagine themselves in the original position and behind a veil of ignorance in relation to the world of the universalized maxim. This is an expansion of the idea that Rawls suggests in Lectures on The History of Moral Philosophy (175–76). I will show that if Kleingeld is correct that the latitude problem is merely a misinterpretation of the FUL, then the veil of ignorance brings the FUL and the FH into alignment about the judgments they make about the theft maxim. Finally, I will respond to a possible objection that comes from a common critique about the practicality of applying the veil of ignorance thought experiment to real rules. I will argue that this objection does not have merit because maxims are far more general and more predictable in their effects on the hypothetical world in which they will act upon than rules.

​

II. The Categorical Imperative: The Formula of Universal Law and The Formula of Humanity

​

A maxim is a subjective principle of the will, a principle that guides one’s actions. For example, “I will help others when I see them in need” is a maxim that guides one to help others in need. Acting on maxims can be either permissible or impermissible. The various formulations of the CI assist us in determining which maxims are permissible to act on and which maxims are impermissible to act on. 

​

In Groundwork, Kant defines the FUL: “I ought never to conduct myself except so that I could also will that my maxim become a universal law” (Ak. 4:402). The FUL is typically understood as a procedure one follows to evaluate a maxim. First, one imagines a world in which everyone acts upon the maxim in question. Then, one determines if this world is even conceivable. If not, then one concludes that acting on the maxim in question is impermissible and is in violation of a perfect duty. Anytime a perfect duty arises from this CC test, the individual using it must either act or refrain from acting on the maxim in question, without exception. If no CC is found, then one should search the world of the universalized maxim for a CW. A CW occurs when one cannot rationally will to act on the maxim in question in its universalized world. When contradictions of this sort arise, the maxim in question is in violation of an imperfect duty. An imperfect duty is a duty that permits exceptions: individuals are permitted to choose to act or refrain from acting on the maxim in question. If neither of these two contradictions occurs when the maxim is universalized, then the maxim is permissible to act on (Johnson and Cureton).

​

For an example of deriving a perfect duty, consider the maxim: “I will steal from others when I need their wealth.” To analyze this maxim with the FUL test, first one universalizes it. One imagines a world in which people steal from each other any time they need each other’s wealth. If an act is truly stealing, then it must violate someone else’s property rights. In this universalized world where stealing is a completely universal behavior, the concept of property would not even exist to begin with. Therefore, there is a conceptual contradiction in imagining this universalized world, and acting on the maxim “I will steal from others when I need their wealth,” is in violation of a perfect duty. This entails that it is never permissible to steal.

​

For an example of deriving an imperfect duty In Groundwork, Kant analyzes the maxim: I will not help others in need of help. When analyzing this maxim, he finds no CC because a world in which nobody offers aid to each other is perfectly conceivable. However, when he analyzes the world of the universalized maxim for a CW, he reasons that nobody could rationally will such a world: In willing such a world, one would be depriving oneself of the aid of others. Everyone reasoning rationally desires the aid of others. Since this maxim only fails the CW test, not acting on this maxim is in violation of an imperfect duty—the duty to be beneficent. Thus, one must hold this maxim as a principle, but one has the discretion to refrain from acting on it (Ak. 4:423).

​

The humanity formulation (FH) of the CI is preferred over the FUL by modern scholars because it is more straightforward and reliable as a guide to the permissibility of maxims. In Groundwork, Kant defines the FH: “Act so that you use humanity, as much in your own person as in the person of every other, always at the same time as end and never merely as means” (Ak. 4:430). What Kant means is that it is impermissible to completely disregard the humanity of another person, for this would be using them only as a means. Using someone, at least partially, as an end entails respecting their humanity (Johnson and Cureton). We should treat the FH as a benchmark for how maxims should be evaluated, meaning that whenever a conflict arises between the formulations, the FH should be the static one. In this paper, the FUL will be contrasted with the FH, and the specific maxims analyzed to demonstrate a problem with the FUL will be shown to be correctly judged by the FH.

​

III. Issues with the Formula of Universal Law

​

Much has been written about how the CC test fails to judge maxims as being in violation of a perfect duty, even when they should be. In her essay, “The Right to Lie, Kant on Dealing with Evil,” Korsgaard argues that the FUL permits a person to lie to a murderer who is trying to deceive them. This presents a problem for the FUL because lying to another person, who by nature of them being a person has humanity, in any circumstance, should be impermissible according to the FH. Lying does not respect the humanity of the other person because it does not respect them as a rational agent who needs true information to make rational decisions to reach their ends (Korsgaard 329–30).

​

In the essay “Murder and Mayhem: Violence and Kantian Casuistry,” Barbara Herman analyses the action of killing for convenience. She demonstrates that the maxim“To kill whatever that is necessary to get what I want” only fails the CW test and not the CC test. This is a problem for the FUL because acting upon a violent maxim should be in violation of a perfect duty according to the FH, not merely an imperfect one. Killing someone ends their humanity, which is in clear violation of the principle to respect humanity. Individuals should not have the freedom to kill others for convenience under any circumstance (Herman 415–16).

​

In her paper, “A Contradiction of the Right Kind: Convenience Killing and Kant’s Formula of Universal Law,” Kleingeld calls this problem the “latitude problem.” The name is derived from the fact that individuals have the “latitude” to sometimes act on the maxim of convenience killing. Though Kleingeld is responding to Herman, the latitude problem can be used to describe all maxims that should fail the CC test but do not (Kleingeld 66–69).

​

In Kleingeld’s paper, she argues that the latitude problem does not exist because it is not entailed by the fact that convenience killing is an imperfect duty that it is sometimes permissible to kill for convenience. The first argument she makes for this (what she calls the simple solution) is that the definition of the FUL, given by Kant in Groundwork, seems to suggest that the FUL judges that maxims are impermissible merely for failing the CW test. Recall how Kant defines the FUL: “I ought never to conduct myself except so that I could also will that my maxim become a universal law” (Ak. 4:402). Kleingeld argues that the use of the word “never” in the definition seems to imply that individuals do not in fact have the latitude to sometimes act on the maxim corresponding to convenience killing (Kleingeld 71–72).

​

The second argument Kleingeld makes undermines Herman’s interpretation of the analysis of the maxim of non-beneficence that she uses to inform her analysis of the maxim of convenience killing. In Groundwork Kant argues that, since the maxim of non-beneficence (i.e., refusing to help others) only generates a CW, people have the freedom not to act using the maxim of beneficence in certain situations. Kleingeld argues that the proper interpretation of Kant is this: one can only refrain from acting on the maxim of beneficence when they instead act on a stricter maxim, a maxim whose opposite generates a CC. The incorrect interpretation of Kant is that it is sometimes permissible to act on the maxim of non-beneficence. She argues that this is the interpretation error Herman makes when she analogizes the imperfect duty of beneficence with the imperfect duty of convenience killing. Thus, Kleingeld argues that the latitude problem does not exist because the FUL is sufficient to deem it impermissible, under any circumstances, to act on the maxim of convenience killing (Kleingeld 74–76).

​

IV. A Problem with the Contradiction in Will Test

​

In this section, I will introduce a problematic maxim for the CW test which I will call the theft maxim. Consider the maxim—“I will steal money from the poor to give it to the rich.” I will refer to this maxim as the “theft maxim” henceforth. Let us first analyze the maxim—“I will steal for my benefit,” the first maxim that was analyzed in section II. Recall that we concluded that this maxim presents a CC in the world of the universalized maxim and thus a perfect duty is violated when it is acted upon. I do not believe that this same contradiction occurs when the theft maxim is universalized. The concept of property persists in the world in which the theft maxim is universalized because ultra-rich individuals will retain their assets since they will not be stolen from. Since the theft maxim does not present a CC, we should see if it presents a CW. 

​

Now consider the analysis of the theft maxim by two different individuals. If the CW test is used by an ultra-rich person, they will find no CW because they will become richer in the world of the universalized maxim. The slight amount of money stolen from them will be far outweighed by the money they gain from everyone stealing money from the poor on their behalf. On the contrary, if the theft maxim is analyzed by a poor person, they could not possibly wish to exist in a world where they would constantly be stolen from. Therefore, this analysis seems to suggest that the FUL permits acting on the theft maxim for a rich person but not for a poor person.

​

Conversely, the FH does not permit anyone to act on the theft maxim. Stealing from someone violates their humanity by not respecting their human capacity to own. Therefore, there is a conflict in judgments made by the FUL and the FH. Since the FH is our benchmark for determining how maxims should be judged, and because we intuitively know that stealing should not be permissible, our focus should be on reinterpreting or revising the FUL. 

​

I see this example as posing two problems for the CW test. The first problem is that this result is troubling for the objective and universal status that Kant places on the CI. Being the supreme principle of morality, it should reliably guide everyone, independent of their personal attributes, to the same conclusions about maxims. If, when used by different individuals, the CW judges the same maxim differently, then the FUL is partially subjective because the CW test depends on the circumstantial attributes of the individual using it. In the theft maxim analysis, a person’s level of wealth, an apparently arbitrary moral attribute, determines whether or not it is permissible for some-one to act on the theft maxim. The second problem is that the CW can be used to justify acting on maxims that are blatantly impermissible, as judged by the FH. 

​

Our task now is to propose a solution or new interpretation of the FUL to both restore the objectivity of the FUL and bring it into alignment with the plain judgment made by the FH about the theft maxim. In the next section, I will attempt to expand upon a solution that was presented by John Rawls in Lectures on the History of Moral Philosophy, in the section “Two Limits on Information.”²

​

V. A Solution: Original Position and Veil of Ignorance

​

In John Rawls’s book, A Theory of Justice, Rawls attempts to invent a system of justice that is fair. To create the fair principles of justice in this system, Rawls says that we must place ourselves in the original position, in which we are blocked by a veil of ignorance. The original position is a hypothetical time before birth when people are able to reason about the principles of justice that will rule the society that they will live in. The veil of ignorance is a hypothetical barrier placed between the person in the original position and personal details about their life. Rawls argues that reasoning about the principles of justice from this condition and behind the veil of ignorance will generate principles of justice that are fair. While in the original position and behind the veil of ignorance, individuals must consider existing as anyone in society, from the most privileged member to the most disadvantaged. They also have to consider the possible rules that govern society and the effect that these rules have on everyone. Rawls argues that people should adopt the rational decision-making strategy known as maximin, which advises choosing the action with the best outcome among the worst-case scenarios (Schwarz). From behind the veil of ignorance, people will choose rules that will be fair to them even if they happen to exist as the most disadvantaged members of society. These principles of justice will serve everyone in society, not just a group of people in one circumstance (Freeman; Rawls).

​

I argue that reasoning about maxims from this original posi-tion, blocked by the veil of ignorance, will allow everyone applying the FUL test to reason identically about the CWs created by maxims. It will also allow the FUL to judge that acting on the theft maxim is impermissible. The FUL test, supplemented with Rawls’s original position and veil of ignorance, should be used as follows: A person wishing to evaluate the permissibility of a maxim imagines what would happen if their maxim was universalized. If they find no CC, they will search for a CW. First, they will imagine themselves as a disembodied consciousness in the original position relative to the world of the universalized maxim. They will also place themselves behind a veil of ignorance, meaning that they will consider the possibility of existing as any person in any circumstance in the world of the universalized maxim. Finally, they will determine if they can rationally will to act on their maxim, considering that it is possible for them to exist as any person in the world of the universalized maxim. They will do this with a principle of rational decision making, and I believe that the best principle of rational decision making for this application is maximin. This is the one that Rawls proposed in A Theory of Justice. It is critical that everyone use the same principle of rational decision making because different principles will lead to different decisions and thus different judgments about whether or not it is rational to will a maxim’s universalization. We want the judgments made by different people about the same maxim to be the same. 

​

Now consider the theft maxim. When one analyzes it, assuming no CC, one will use the CW test and imagine themselves as a disembodied consciousness above the world of the universalized theft maxim and behind a veil of ignorance. One must imagine oneself as any person in this world: As someone poor who is always being robbed by those richer, or as someone rich who is being robbed slightly but who is also receiving money from those poorer. Since it is possible that one will be born into the position of someone poorer who is constantly being stolen from, one cannot rationally will that this maxim becomes a universal law. The principle of maximin tells one to avoid the worst possible outcome, and universalizing the theft maxim may result in the worst outcome, ending up as the absolute poorest person. The conclusion one will arrive at is that acting on the theft maxim is impermissible and is in violation of an imperfect duty. 

​

Thus, we have solved the first problem, the CW’s reliance on personal attributes and its subjectivity. Before Rawls’s ideas were added to the CW test, the level of wealth of the person using it could cause them to either accept or reject the theft maxim. Now, however, since the same principle of rational decision making, the maximin principle, will be used by everyone performing the CW test, everyone will come to the same judgment about the permissibility of a maxim that is universalized. We have also solved the second problem, the problem of justifying immoral maxims. The analyses above show that the theft maxim fails the CW test when it is supplemented with the veil of ignorance. This aligns with the judgments of these maxims made by the FH.

​

One might still have the concern that acting on these maxims should be in violation of a perfect duty and should thus present a CC. If the latitude problem is a real problem, then the veil of ignorance cannot be a full solution to the problem presented by the theft maxim. The FH is unequivocal in its judgment that the theft maxim should entail a perfect duty that can never be violated. We would then have to explain why the theft maxim does not present a CC when universalized. This is a valid concern, but one that can be lessened if Kleingeld is right in her conclusion. Recall her interpretation of the FUL that was outlined in section III: The latitude problem does not exist; a maxim presenting a CW is sufficient to deem it impermissible to act on it. If, as Kleingeld argues, the latitude problem is derived from a misinterpretation of the FUL and does not exist, then the CW test is sufficient to bar people from acting on the theft maxim under any circumstance.

​

Finally, there is a practical concern for the veil of ignorance that some level at Rawls’s thought experiment for deriving principles of justice. Some argue that, although the veil of ignorance is a useful thought experiment, it is hard to apply in the real world. It is hard to use it to come to detailed and precise rules to govern a society with the thought experiment because rules have complex effects that are hard to predict. Similar concerns might be leveled at the new CW test. One could argue that the consequences of universalizing a maxim might likewise be hard to predict and thus it might be hard to apply the CW test to real maxims.

​

I do not think that this concern has as much merit when applied to the veil of ignorance in the CW test because maxims are more general in their construction and thus more predictable in their effects on the hypothetical world being considered in the CW test. It is also the case that the effect of any maxim is simple: that everyone will use it to act in the hypothetical world. On the contrary, a law can cause numerous effects on the world. A law could create a new institution or disallow some behavior or any number of things. Therefore, the generality and predictability of maxims make them easier to analyze. 

​

VI. Conclusion

​

The FUL has been heavily criticized because when applied to certain maxims, it both contradicts the remainder of Kant’s moral philosophy or judges blatantly impermissible maxim’s as permissible. I used the FUL to analyze the theft maxim and argued that without the veil of ignorance, this maxim passes both the CC and CW tests. The first problem posed by this result was that maxims pass or fail the CW test based on irrelevant attributes about the person using it, which contradicts the objective status of the FUL. The second problem posed by this result was that it seems like maxims that should be impermissible are permissible for certain people to act on. Supplementing the CW test with the veil of ignorance solves both of these problems with the CW test when applied to the theft maxim. If Kleingeld’s argument, that the latitude problem does not exist and is based on a misinterpretation of the FUL, is correct, then it in conjunction with the CW test using the veil of ignorance can align the judgments that the FUL and FH make about the theft maxim. Finally, I responded to a potential concern with the CW being used with a veil of ignorance, a concern derived from concerns about the veil of ignorance in Rawls’s _Theory of Justice_. I argued that this concern does not have as much merit as the concerns leveled at the veil of ignorance applied to Rawls’s theory of justice because maxims are more general and their effects are more predictable on the hypothetical world under analysis by the person using the CW.