Against a Kantian Response: Can we feasibly deny the rationality of irrational agents?
In this essay, I argue that denying rational agency to defend Kant’s Formula of Humanity against certain unintuitive moral implications fails. My argument proceeds in four sections. In the first section, I explicate Kant’s Formula of Humanity, focusing on the impermissibility of undermining rational agency. In the second section, I explicate how Kant’s Formula of Humanity leads to unintuitive moral implications – namely that it is impermissible to kill a would-be mass murderer to stop him from murdering millions. In the third section, I offer a response to avoid this, which is to deny the would-be mass murderer of his status as a rational agent (which would make him a would-be mass killer) to make it permissible to kill him. In the fourth section, I conclude that this response is unsuccessful as having irrational ends does not mean that killing the would-be murderer would not undermine any rational agency.
(Those familiar with Kant’s Formula of Humanity, skip to Section 2)
Section 1
Formula of Humanity
In the Formula of Humanity, Kant argues that it is morally impermissible to undermine rational agency.
Kant argues that rational nature is an unconditional end-in-itself (Kant 107). This is because, if there were to be unconditional ends, they could not be conditioned upon any subjective characteristic about anyone. Thus, the only commonality which all rational agents share is that they all have rational nature. This assumes that unconditional ends exist. Kant substantiates his claim about the existence of unconditional ends in claiming that “[Morality must be] … valid not merely under contingent conditions… but with absolute necessity” (Kant 97). Kant argues that morality, being a necessarily inescapable/unconditional thing, would be an unconditional imperative (categorical imperative). It then follows that the existence of an unconditional imperative necessitates the existence of unconditionally necessary actions. As actions are done for achieving their respective ends, the existence of unconditionally necessary actions necessitates the existence of unconditional ends. This establishes that, there are such unconditional ends and that these ends would be rational nature in-itself. This means that all rational agents must necessarily will themselves as unconditional ends-in-themselves.
Kant then argues that, because all rational agents have rational nature, each individual rational agent must necessarily recognize and respect other rational agents as ends-in-themselves. This is because to be rationally consistent is to recognize that all rational agents (oneself and others) are equally ends-in-themselves and that this leads to the conclusion that other rational agents should be respected as ends and not as mere means to one’s ends. It is rationally inconsistent to hold one’s own rational nature as an end while not holding others’ rational natures as ends-in-themselves.
This leads us to the conclusion that one always ought to respect rational agency. What Kant means by respect is to always treat someone as an end and never as a mere means to an end. Kant states “Act in such a way that you treat humanity, whether in your own person or in any other person, always at the same time as an end, never merely as a means” (Kant 107). Kant holds that to treat someone as a mere means is to undermine their rational agency. This culminates into the following deontic constraint: An act is impermissible if and only if it undermines the rational agency of another. To illustrate what Kant means by treating someone as a mere means, I employ the following distinction between treating someone as a mere means in comparison to treating them as a means.
Mere means: Treating someone as a mere means is to treat them as only a means while not considering their rational agency. (i.e. Deceiving someone into helping oneself obtain more money at their expense.) In this case, deceiving someone undermines their ability to act rationally as they are epistemically misled and unable to consider the appropriate knowledge to act rationally.
Means: Treating someone as a means is to treat them in a way that they contribute to one’s ends while one takes into consideration the fact that they, too, have ends. (i.e. Using a cashier as a means to buying items at a store.) In this case, using the cashier considers their personal ends of making money at the institution that they chose to work at. This does not undermine their rational agency despite using them as a means to one’s ends.
Section 2
Objection – Morally unintuitive results
An objection to Kantianism’s deontic constraint – that it is always morally impermissible to undermine rational agency – is that it leads to morally unintuitive results. This essay concerns itself with the following formulation of such an objection. Suppose one has undeniable knowledge that a would-be mass murderer is about to intentionally kill millions of innocent people with the press of a button. Suppose the only way one could prevent this from happening is to kill said person. Moral intuition states that one ought to kill the would-be murderer – or that killing them would at least be permissible. However, taking Kantianism to be true, killing the would-be murderer would be impermissible. Kantianism claims that it is impermissible to undermine a rational agent’s rational agency. Killing a rational agent, regardless of the circumstances, is an instance of undermining rational agency. To kill someone is to absolutely undermine their ability to exercise any sort of rational agency. In a similar vein, killing the would-be mass murderer is also using him as a mere means to one’s end – using his life as a mere means to the end of saving the many.
Section 3
Rejoinder – Making it permissible to kill a would-be mass murderer
Given that the Formula of Humanity claims that one always ought to respect rational agency, one prima facie plausible response to this formulation in Section 2 is to deny that the would-be mass murderer is a rational agent without adjusting the prior formulation of the Formula of Humanity. This disqualifies him from the deontic constraint and makes the act of killing him permissible as one would not be undermining rational agency in doing so. This is because Kantianism is caveated such that it only applies to rational agents – one ought to respect rational agency, which only rational agents have. To deny the rational agency of the would-be mass murderer, one must explicate how murder is irrational. This is prima facie implausible as this is a clear case of the exercise of instrumental rationality. The would-be mass murderer does a particular act (pressing the button which would kill millions of innocents) to bring about a certain set of circumstances (the deaths of millions of innocents).
To deny the would-be mass murderer’s status as a rational agent, one can claim that although the would-be mass murderer is instrumentally rational, he is rationally inconsistent. Being instrumentally rational is to will a certain end by doing a certain set of actions and have the end and actions be consistent. For example, to achieve end X, one must will Y. It is instrumentally rational to will Y. However, this is conditioned upon the actor having end X, for it would not be instrumentally rational to will Y if one had no ends which would require it. To be rationally consistent, however, is to have consistent ends or beliefs. Conversely, to be rationally inconsistent is to have contradictory ends – like willing X and not-X simultaneously. The would-be mass murderer does demonstrate instrumental rationality – since his goal of murdering millions aligns with his action of pressing the button. However, he is rationally inconsistent. Assuming that the formula of humanity is true, it would be impossible to will with rational consistency the undermining of other rational agents’ rational agency (to kill them). As stated in Section 1, rational nature is an end-in-itself (P) and it is irrational/inconsistent to hold one’s own rational nature as an end-in-itself (P) while not holding others’ rational natures as ends-in-themselves (-P). The would-be mass murderer is hence acting rationally inconsistently in not respecting the rational agency of the millions of innocents that he is about to murder. In denying the would-be mass murderer’s status as a rational agent in the circumstances above, Kantianism could hold that it is then permissible to stop him from murdering millions. In fact, Kantianism could hold that it is then impermissible not to stop him from murdering millions as not stopping him would undermine the rational agency of millions.
Section 4
Rebuttal – Failure to permit the killing of would-be mass murderers
Kantianism cannot permit the killing of the would-be mass murderer and remain consistent. Though the murderer may not be acting irrationally while he is killing a million people, this does not make the act of killing him permissible. As acting irrationally does not remove the existence of other, unrelated rational ends that he has, killing him would undermine the rational agency he has in willing said rational ends. The formula of humanity would still apply to him, as although he may be willing some irrational ends, we still ought to respect the other rational ends that he might hold at that moment. One could argue that he is still an irrational agent due to rational inconsistency but if a single instance of rational inconsistency disqualifies one from being a rational agent, most people in reality would also be disqualified. This is unfeasible to hold for Kantianism as the formula of humanity would then only apply to perfectly rational agents – it would be permissible to kill any imperfectly rational agent.
A possible Kantian reply could be that in a particular circumstance where the would-be mass murderer only holds the end of killing millions in an instance of irrationality, he loses his status as a rational agent. Kantianism could then argue that this makes it impermissible to not kill the would-be mass murderer. This case is very limiting (it is required for the would-be mass murderer to be overwhelmed by the end of murdering millions, losing all other goals) and psychologically questionable (feasibility for someone to hold no other ends is questionable). Furthermore, Kantianism cannot coherently hold this position, even if it were granted psychological feasibility. This is because killing someone does not just undermine their rational agency at that specific moment in time but also undermines all possible rational agency that they would have had in the future. Killing someone absolutely undermines their ability to be a rational agent and uses their future self as a means to an end. Kantianism cannot permit the undermining of future rational agency as it would allow for even more morally unintuitive results. A formulation of one such result is as follows: Suppose an isolated person is in dreamless sleep and is not currently holding any rational ends. That person is not currently rational as their mind is not active (they are not willing any ends) but would be a rational agent in the future if left to wake up. As killing such a person would not undermine any currently existing rational agency, it would be permissible to kill them If Kantianism allows for the undermining of future rational agency. As Kantianism would have to hold even more morally unintuitive results in accepting the undermining of other rational ends/future rational ends, the rejoinder in Section 3 is unsuccessful.
When Denying Rationality could work
In a different case, a similar formulation to that in Section 3 could have worked in a different formulation where one could instead deceive the would-be mass murderer instead of having to kill him. It could permit the undermining of only that specific irrational end of murdering millions as it is not an exercise of rational agency. This could have come in the form of a lie which only affects them in such a way that they stop willing the end of committing mass murder. Unfortunately, this cannot apply to killing due to the absoluteness of killing an individual – the elimination of all their current and future rational agency. Killing is unconditionally impermissible under Kantianism – which is, as explicated, a morally unintuitive result. As the only way to avoid the morally unintuitive result under Kantianism is to deny the would-be mass murderer’s rational agency, Kantianism’s failure to provide a way to resolve this result shows that it is unsound.
Written by,
Joshua Chua (Editor 19/20′)
This essay was written by a university undergraduate, just like you. Have an essay that you’re proud of and would like to share? Have interesting/crackpot ideas/essays you would like to submit? Contact us at our Instagram @nanyangphilosophy.
Works Cited
Cahn & Forcehimes, Principles of Moral Philosophy (Oxford, 2016) – Selections from Kant
The Nanyang Philosophy Review 