Deontology v. Consequentialism
Considering how I would have responded to Mills’s questions about torture reminded me of a few interesting things about moral psychology.
We can divide ethics along these lines: on the one hand, in determining the moral value of an act, consequentialism looks at the consequences of the act, and (in absolute consequentialism) only that. On the other, deontology looks at the act itself as apart from its consequences, and (in absolute deontology) only that. Almost no one is absolutely one or the other: almost all of us will sometimes give typically deontological answers and sometimes typically consequentialist answers. The ratio between them varies, though.
A consequentialist must determine what the good is. A typical answer is the happiness principle: promoting the greatest amount of happiness for the greatest number of people (the good need not be — it can be something else and still be consequentialist). So, considering torture, a consequentialist might reason thus: torture inflicts suffering, and so reduces happiness. But if it promotes more happiness than it causes, then it is justified. So, for a consequentialist, whether or not torture is permissible would depend on how effective it is in producing information that saves lives or generally helps uphold general happiness in the population. Thankfully, torture isn’t very effective when it comes to producing reliable intelligence, but if it were, a consequentialist would have to endorse it. Many of those who oppose torture would have second thoughts if it were proven to be effective, and their arguments are not to the effect that torture is wrong per se, but that torture is wrong because it doesn’t produce good enough results to offset the suffering it produces.
I, on the other hand, take a more deontological approach. To me, torture is wrong as an act. The act of torture, no matter what, is wrong. Consequences (good or bad) cannot justify torture, no matter how good.
The fascinating thing is that the evidence suggests that we all have two competing systems of morality that roughly correspond to consequentialist and deontological ethics. One is an emotional, immediate reaction that condemns certain actions: this is deontological. The other is less emotional, more analytic, and considers the social consequences of the act: this is consequential, and must deliberately override the emotional response. That must not be taken to mean that one is purely emotional and the other is purely rational, of course. The authors report interesting findings regarding the interaction of these two systems of morality. The following moral dilemma was posed to subjects:
Enemy soldiers have taken over your village. They have orders to kill all remaining civilians. You and some of your townspeople have sought refuge in the cellar of a large house. Outside, you hear the voices of soldiers who have come to search the house for valuables. Your baby begins to cry loudly. You cover his mouth to block the sound. If you remove your hand from his mouth, his crying will summon the attention of the soldiers who will kill you, your child, and the others hiding out in the cellar. To save yourself and the others, you must smother your child to death. Is it appropriate for you to smother your child in order to save yourself and the other townspeople?
And they found:
Subjects tend to take a long time to respond to this dilemma, and their judgments tend to split fairly evenly between the characteristically consequentialist judgment (“Smother the baby to save the group”) and the characteristically deontological judgment (“Don’t smother the baby”). (For a discussion of why we consider it legitimate to refer to these judgments as “characteristically deontological” and “characteristically consequentialist,” see Greene (2007)). According to Greene and colleagues’ dual-process theory, the characteristically deontological responses to such cases are driven by prepotent emotional responses that nearly everyone has. If that’s correct, then people who deliver characteristically consequentialist judgments in response to such cases must override their emotional responses. This theory makes two predictions about what we should see in people’s brains as they respond to such dilemmas.
First, we would expect to see increased activity in a part of the brain called the anterior cingulate cortex (ACC), which, in more dorsal subregions, reliably responds to “response conflict” (Botvinick, Braver, Barch, Carter, & Cohen, 2001). The reason is that, according to this theory, these difficult dilemmas elicit an internal conflict between a prepotent emotional response that says “No!” and a consequentialist cost-benefit analysis that says “Yes.” And, indeed, Greene and colleagues found that difficult “personal” dilemmas like crying baby case elicit increased ACC activity, relative to easier “personal” dilemmas, such as whether to kill your boss because you and others don’t like him, in which reaction times are shorter and judgments are more overwhelmingly negative. Second, we would expect to see increased activity in a part of the brain known as the dorsolateral prefrontal cortex (DLPFC). This part of the brain is the seat of “cognitive control” (Miller & Cohen, 2001) and is necessary for overriding impulses and for “executive function” more broadly. Once again, if the characteristically deontological judgment is based on an intuitive emotional response, then giving a characteristically consequentialist response requires overriding that response, a job for the DLPFC. As predicted, Greene and colleagues found that consequentialist judgments in response to difficult “personal” dilemmas (“Smother the baby in the name of the greater good”) are associated with increased activity in the DLPFC relative to the activity associated with trials on whichdeontological judgments were made.
These neuroimaging results support a dual-process theory of moral judgment in which distinct “cognitive” and emotional processes sometimes compete. But neuroimaging data are inherently correlational and can only suggest causal relationships between patterns of neural activity and behavior. To provide more direct evidence for such causal relationships, one must manipulate the processes in question and observe the effects. In a recent study, Greene and colleagues (in press) did this by imposing a “cognitive load” on people responding to difficult “personal” moral dilemmas like the crying baby dilemma. People responded to the moral dilemmas while simultaneously monitoring a string of digits scrolling across the screen. The purpose of this manipulation is to disrupt the kind of controlled cognitive processes that are hypothesized to support consequentialist moral judgments. They found, as predicted, that imposing a cognitive load slowed down characteristically consequentialist judgments, but had no effect on characteristically deontological judgments. (Deontological judgments were in fact slightly, but not significantly, faster under cognitive load.)
We must be very careful here to avoid saying that the more intuitive, deontological system should be disregarded in favor of the more analytic consequentialist system. First of all, as the article states, there are many cases where reasonable people have reasoned themselves into deontological responses. Second, we must not confuse psychology with logic: we can’t psychologize moral discourse to such an extent that we determine the truth or falsehood of a moral statement merely based on their psychological origins. You don’t win moral debates by screaming “HAH! I caught you exercising your emotional, pre-cognitive moral system! My cerebral cost-benefit analysis wins by default!” Both consequentialists and deontologists must argue for their positions, and we must not forget that our trust in the consequentialist analysis’s premise — that the social consequences of an act overall are more important than the act itself — is no less intuitive and pre-cognitive than our trust in our deontological judgments.
The existence of this dual system is very interesting, even if it doesn’t directly solve any ethical dilemmas. More research into how and why one system overrides another would be interesting; in trolley problems, for instance, laymen and philosophers sometimes tend to give typically deontological responses and in slight variations the opposite.
Ethical problems won’t be solved directly by research into the neurological basis of morality, nor by polls, but these results are interesting in their own right and might, possibly, inform future debates. The proper cite for the above article is here, btw:
Cushman, F., Young, L., Greene, J. (in press). Our multi-system moral psychology: Towards a consensus view. In J. Doris, G. Harman, S. Nichols, J. Prinz, W. Sinnott-Armstrong, S. Stich. (eds.), The Oxford Handbook of Moral Psychology. Oxford University Press.
