Of Trolleyology and Pedophilic Princes: The Psychological Inseparability of Acts and Consequences

Moral philosophy has often accused moral psychology of making the naturalistic fallacy, confusing is with ought.  But what about the opposite direction, confusing ought with is?  Here, Chelsea Schein and I suggest that neither deontology nor utilitarianism accurately represent moral cognition because each argues for the separation of acts and consequences. 

We argue that acts and consequences are--psychologically--not only inextricably bound together, but are also often compatible and mutually reinforcing.  Just as we cannot help but see the suffering baby when viewing the venomous snake, we cannot help but see bad consequences arising from immoral acts.  This means that paradigms that artificially separate acts and consequences or make them conflict (i.e., trolley problems) may not reveal general truths about moral cognition.  Read more in our paper in Review of Philosophy and Psychology, co-edited by the fabulous Mark Phelan and Adam Waytz.

For more, see also the excellent recent work of Brittany Liu and Peter Ditto, and Dan Kahan.

p.s. the “pedophilic princes” comes from the thought experiment in our introduction.

Comments

[BioethicsUK]

Looks like some interesting work. There is some recognition of this problem - that ought implies can - in moral philosophy but not nearly enough. And, to make matters worse, certainly not enough in applied (bio)ethics. We spend a lot of time teaching 'ethics' to medical students etc with little understanding that what can (and should) be achieved in clinical practice is different from what can (and should) be achieved in the pages of a journal. Plus there are many contexts that lie between these two extremes. A broader view of the moral and ethical landscape is very much required!

[Clayton Littlejohn]

I have to register a dissenting opinion. I can't seem to see the baby while seeing the snake or see the snake while seeing the baby.

[Kurt Gray]

Hi Clayton,

You make a good point--I should have been clearer in my summary of the figure. The baby and the snake both instantiate the existence of the other, like a figure and a ground.

The fact that you can't see both at the same time, has an analog in morality, and I call it moral typecasting: the basic idea is you either see people as moral agents (snakes) _or_ as moral patients (babies) and generally not both. I've got a paper with Liane Young and Adam Waytz ("Mind Perception is the Essence of Morality") that has a section that provides a quick overview of moral typecasting (and a longer paper with Dan Wegner called "Moral Typecasting"). Of course there are always dissenting opinions and you can read them in the same upcoming issue of RPP (e.g., Adam Arico and colleagues have a paper called "Breaking out of moral typecasting").

Thanks for your thoughts.

[Clayton Littlejohn]

I read this paper last night and again this morning. I have a few questions about the paper. It's all quite interesting, but I'm not quite sure what to take from it as someone whose interests have to do with consequentialist and non-consequentialist theories of right action.

Your research suggests that moral cognition concerns both acts and consequences. You say that the link between wrong and harm is 'inextricable', but are you suggesting that theories that separate the two are literally incredible or beyond belief? I take it that a consequentialist will say (and some deontologists will say) that the phenomenon of dyadic completion is responsible for some relatively robust illusions, illusions that can be corrected by theory. I don't think I understand the modality of "cannot" in your remark, "More narrowly, dyadic completion suggests that people cannot psychologically separate immoral acts from the suffering of victims." On one reading, you might think that people cannot separate Fness from Gness psychologically if they cannot grasp the possibility of a scenario in which an F isn't a G (e.g., they cannot grasp the possibility of something having a length without being extended spatially or cannot grasp the possibility that Zoro is present and Don Diego is absent after they learn about the identity). On another weaker reading, there's a kind of inclination to treat F-involving situations as G-involving, one that can be overcome by conscious reflection. Is the weaker reading all that your after? Some passages later in the paper suggest that this reading is too weak. Is the idea that we can entertain the possibility of harmless wrongs or wrongless harms, but can't integrate this possibility into the way we form beliefs about particular cases? About theories?

Perhaps what worries me the most is found in this passage:
"Recently, this sharp line between is and ought has been blurred as normative theories are advanced to describe how people make moral judgments. In particular, researchers have argued that the two theories of deontology and utilitarianism represent two distinct and competing modes of moral decision-making (e.g., Greene et al. 2001). Evidence for this idea comes from (so-called) trolley problems—moral dilemmas where an unpalatable act (e.g., murder) is pitted against desirable consequences (e.g., saving lives). Because deontology emphasizes acts over consequences, and utilitarianism emphasizes consequences over acts, these two normative theories aptly describe the two conflicting intuitions in these moral dilemmas. The question, however, is whether these competing normative theories describe moral cognition beyond the bounds of trolley problems."

Perhaps this is just a semantic dispute about what "deontology" and "consequentialism" come to, but I didn't think the deontologists or consequentialists were offering descriptive theories of how people reasoned about morality, but normative theories that provide truth-conditions for moral judgments. Moreover, as many ethicists have stressed, there is an important difference between an account of what makes right acts right and a decision-procedure that guides deliberation. (Sinnott-Armstrong's SEP entry on consequentialism is pretty good on this.) So, I guess I don't see why the deontologist or consequentialist is committed to any descriptive claims about the way that moral cognition operates. Moreover, it's not clear that they are committed to any _prescriptive_ claims about the way that moral cognition ought to operate since their core theses have to do with what makes moral judgments true or false, not what methods or reasoning would reliably lead to true or false judgments.

The worry I have about the commitments of utilitarianism concerns this passage, too:
"Utilitarians may focus primarily on the consequences of moral actions—on the realized or potential suffering of moral patients—but they also acknowledge the importance of agent-based factors leading up to these outcomes. Nevertheless, we suggest this acknowledgement is insufficient, because people care about moral agents even when such concerns are irrelevant—or even antagonistic—to concerns about consequences."

I have a hard time understanding the charge here. What does "sufficiency" come to? Are there important truths about morality that utilitarianism cannot deliver because of their focus on outcomes and neglect of agent-related considerations? If not, I don't see how utilitarianism can fail to provide sufficient acknowledgement as it seems they're just in the business of offering a theoretical explanation of the facts about rightness. The studies that show that people care more about outcomes are interesting, but they are just the sorts of things that my utilitarian friends say explain the irrational attachment we have to deontological theories.

I could imagine someone responding to my worries in this way. Facts about morality _are_ sensitive to facts about the way we actually reason about morality. In particular, the facts about morality have to fit this template of dyadic completion because that's just the way we think.

That's an interesting argument, one that addresses the worries I have about this research. I couldn't defend such an argument, but it at least explains why one might take the sorts of empirical observations contained in your research as relevant to evaluating some consequentialist or non-consequentialist theory. If, however, someone categorically rejects this sort of argument (as I do), then I guess I can't see what warrants the idea that deontological or consequentialist views have any skin in the game when it comes to descriptive models of moral cognition. (And maybe I'm just reading too much into your remarks when I take you to be suggesting that they have some skin in this game.)

[Richard Yetter Chappell]

"neither deontology nor utilitarianism accurately represent moral cognition"

Can you clarify how this is an objection? Normative theories aren't generally intended to "represent [actual] moral cognition" after all. Actual people aren't expected to be morally ideal.

[Chelsea Schein]

Thanks for your thoughts, Clayton and Richard. Clayton, toward the end of your post, you ask why one ought to believe that normative moral theories have any “skin in the game” when describing moral cognition, and Richard, I believe this is your gist as well.

Even if normative theories were never indented to describe moral cognition there has recently been an emergence of research in psychology that applies normative theories to explain moral cognition. In this paper, we challenge this application of philosophy to psychology. As illustrated in our research on dyadic morality as well as research by Dtto and Liu, people generally do not separate utilitarian and deontological concerns. Therefore, research in moral cognition that stems mainly from dilemmas pitting consequentialism and deontology against each other, will fail to fully capture an accurate picture of moral cognition. Whether this empirical finding challenges the truth of normative theories is a question we are happy to leave up to normative ethics.

Do we believe that "theories that separate the two are literally incredible or beyond belief?" A response to this question depends on what one means by beyond belief. When we look at visual illusions, we can know at a theoretical level that our eyes are deceiving us. Nonetheless, this knowledge does not eliminate the illusion. We believe that our perceptions in the moral realm work in a similar way. We may know logically that it is possible that immoral actions can be independent of consequences. In theory, after engaging in a lengthy form of conscious deliberation, one could believe that birth control is immoral even though it leads to great outcomes. Nonetheless, at a perceptual level, one cannot help but see more harm in the world when one believes that an action is immoral.

[Clayton Littlejohn]

Hi Chelsea,

Thanks for your response. I thought that the analogy with visual illusions was helpful, thanks. On the issue of the descriptive/prescriptive, I think Richard and I are probably on the same side here, which is that it would be a mistake to apply philosophical accounts of rightness to explain something about moral cognition. Maybe we're on the same page as you, then.