Peter DeScioli's Response: The Kantian Denial of the Thomistic Distinction Between Intended and Unintended Killing

Peter DeScioli has sent me some responses to my previous post.  He has given me permission to quote them here:

I just skimmed your article and have just one clarification for now: We did not intend to draw on Kant as an explanation or theory of moral judgment. We use "Kantian" in a purely descriptive sense to mean that people's judgments resist utilitarian tradeoffs in dilemmas like the footbridge trolley problem. Our use of "Kantian" is synonymous with "absolutist", "non-consequentialist", "deontological", "taboo" in Tetlock's sense of resisting tradeoffs, and other synonyms in the moral psychology literature. So just as describing moral judgment as "absolutist" does not appeal to something transcendental, we did not intend "Kantian" to do so. Kant is simply an iconic example of rigid moral judgment focused on prohibited actions and resistant to tradeoffs, and similarly we refer to "categorical imperatives" as a descriptive label of moral psychology, rather than as an appeal to Kant's theories. 

Let me also point you to the passage below from DeScioli 2016 which you cited. It illustrates how I use Kantian as a descriptive label for categorical judgments in contrast to utilitarian judgments which I here label as Vulcan, using Spock from Star Trek to represent utilitarian judgments. I argue that humans are "hybrid Vulcan-Kantians," meaning we have both forms of judgment, as seen in the footbridge and switch versions of the trolley problem. Our view is that the Vulcan judgments come from cognitive systems designed for altruism, while Kantian judgments come from cognitive systems for choosing sides. 

DeScioli has said that "Immanuel Kant would argue . . . that when humans face this dilemma, they should not kill one to save five because there is an inviolable moral rule against killing that cannot be broken regardless of the consequences" (Kurzban et al. 2012, 323).  So, DeScioli suggests, in the footbridge version of the trolley problem, most people show a Kantian morality in refusing to kill the fat man because they believe there is an "inviolable moral rule against killing" regardless of the good "intended outcome" (saving five lives) of the killing.  But in the switch version of the trolley problem, most people show a utilitarian morality that justifies killing one person when the "intended outcome" is good--saving five lives.

DeScioli sees an unresolvable contradiction between two kinds of moral rules--killing is always wrong versus killing is wrong except when the "intended outcome" is good.

But as DeScioli has indicated, John Mikhail's research offers a plausible way to resolve this contradiction by seeing the Thomistic Principle of Double Effect as part of the evolutionary moral psychology of human beings.  By seeing that one action can have two effects--one that is intended and another that is foreseen but unintended--we can see that the outcomes or consequences in the footbridge and switch cases of the trolly problem are not the same.  In both cases, one person is killed to save five people.  But in the footbridge case, both the saving of the five people and the killing of one person are intended.  While in the switch case, the saving of the five is intended, but the killing of one is not.

DeScioli recognizes this Thomistic distinction between intended and unintended killing when he considers the laws of violence.  "Thou shalt not kill . . . proclaims the most unanimous and iconic of moral laws" (DeScioli 2023, 202).  Here then is the Kantian categorical imperative--no killing regardless of the consequences.  "But there are of course exceptions," DeScioli says.  We allow killing to punish a wrongdoer, killing in self-defense, and the killing of enemy combatants in war.  But doesn't that deny the Kantian categorical imperative of no killing regardless of the consequences and with no exceptions?

As Aquinas indicated, these exceptions to the no killing rule depend on the distinction between intended and unintended killing.  One may kill an aggressor to save one's life or the lives of others, but only as long as the killing is not the directly intended effect.  If the aggressor retreats or surrenders, and thus gives up the threat to life, we cannot rightly kill him.  To kill him in that situation would be an intended killing, which is wrong.  This requires that we judge the circumstances and the consequences of our actions.

This Thomistic principle of double effect in cases of killing is clearly illustrated in the laws of just war.  Combatants may kill enemy combatants in war to remove their threat to life.  But once the enemy combatants surrender, we cannot kill them without being charged with intentional murder.  We may directly target enemy combatants for attack, but we may not directly target noncombatants, even though noncombatants are often killed as an unintended side-effect of attacks directed at combatants.  So, for example, we can condemn the directly intended terror bombing of Germany and Japan in World War II by the allies as unjust.

This illustrates DeScioli's primary argument "on the origin of laws by natural selection"--that moral rules and laws are for choosing sides in conflicts.  So when someone threatens to kill someone, and his threatened victim argues that this is morally wrong, we can take the side of the victim by agreeing to moral rules or laws that prohibit unjustified killing.  But strangely DeScioli denies that people taking the same side in a conflict is a form of cooperation.  Equally strange is that DeScioli does not see how the rule of justified killing--such as killing in self-defense or killing to save the lives of others--denies the Kantian absolute rule of no killing with no exceptions.

Consider this scenario at the opening of one of his articles:

"Imagine you are alone on a desert island when a hulking man says to hand over your food supplies or he will kill you.  You protest that his threats are morally wrong.  How much protection could your moral judgment really provide?  If you had to choose, would you defend yourself with a hand axe or moral arguments?  Now imagine instead that there are fifty people on the island.  In this case, morality might actually be the better weapon: a persuasive moral argument could rally dozens of armed defenders to your side" (DeScioli 2016, 23).

Persuading dozens of people to give you armed defense against the threats of a murderer sounds like cooperation to me!  Indeed, this is exactly how social cooperation emerges in Locke's state of nature when people agree to formulate the law of nature for resolving conflicts and to punish the violators of that natural law.  Moreover, recruiting "armed defenders to your side" must violate the Kantian absolute rule of no killing--not even in self defense.

I will say more about this in another post.