August 31, 2004

Human Utility Functions

Our hardwired taste for revenge:

Boing Boing: Neuroscience of revenge: Researchers at the University of Zurich have shown that revenge is, well, sweet. Their experiment, described in Scientific American, was based on a game where one player was given the opportunity to punish another player for financially screwing him. PET scans revealed that when a player contemplated revenge, his striatum, a "reward center" in the brain, became energized.

This sort of causal relationship may explain why people are willing to discipline a stranger even when there is no immediate gain in it for them. "Emotions play a proactive as well as reactive role," remarks Brian Knutson of Stanford University who penned an accompanying commentary (to a paper about the study in the journal Science). He notes that "passionate" forces may need to be included in economic models because, as this research shows, “people show systematic deviations from rationality."

Posted by DeLong at August 31, 2004 05:38 PM | TrackBack

The desire for retribution most likely is a product of our evolution. We need to punish those who break the norms of our clan so as to make life in a group bearable. So this kind of research lends support to at least one of three reasons why we punish: to deter, to remove the individual from the clan, and finally for retribution. The state has taken these functions away from the individual (and rightly so), but the modern state is falling down on the job by delaying the ultimate punishment so long that it subverts at least two of the benefits of that kind of punishment.

Posted by: A. Zarkov at August 31, 2004 06:02 PM

From a game theory standpoint, revenge is not irrational. It is the tat portion of tit-for-tat, and therefore an essential part of our moral nature. Our willingness to take the risk of trusting, cooperating, and engaging in altruism would make us mere victims were it not backed up by retribution whenever these things are abused. The basic selfishness always is ready to re-emerge.

Posted by: Martin Bento at August 31, 2004 07:25 PM

From art. on same research here - -
"inflicting the punishment (on social norm violators) didn’t cause the players to feel satisfaction. Instead, as they decided to impose the penalty, the players were **anticipating** feeling satisfied."

Which is consistent with the observed* pattern that relatives of murder victims who want the murderer to get the death sentence think they'll feel satisfaction after it's been administered, but don't.
(*sorry, no supporting data, but I'm 90% sure it's real)

Posted by: Anna at August 31, 2004 07:59 PM

Why should the moral theorist (or the economist) find this substantively important? We all recognize that people can be moved to act by desires and emotions, and we all know that the contemplation of revenge is pleasurable. But who believes that these are not elements of a person's motivational set and/or her subjective utilities?

Economists, I would've thought, try to model how we behave and our failures to maximize utilities. Revenge, whatever its biological explanation, seems just another source of utility.

Moral philosophers deal with what should motivate us. It may be that revenge OUGHT not to move one to action, but (with reverence to Mr.Hume) that's logically distinct from what DOES move her.

Or am I missing something?

Posted by: adm at August 31, 2004 08:20 PM

I would like to describe a different game. Two children sat in front of a chess board with a single checker placed near the middle of the board. They were told that if the checker was pushed to the edge of the board closest to their "opponent" they would receive a prize. Each child took a turn moving the checker. American children fruitlessly pushed the checker back and forth between two squares until a draw was declared by the researcher. Children in developing countries had a different response and alternately allowed each other to win by cooperating to move the checker in the same direction until one had achieved victory and then reversing the direction in each succeeding game. This experiment gives me reason to be skeptical about "our hardwired taste for revenge". It just might turn out to be our conditioned taste for revenge.

Posted by: Dubblblind at August 31, 2004 11:36 PM


That's very interesting. Do you have a cite? (I'm asking because I'm geniuinely intrigued).

In any case, what has happened is that the developing children developed a tit-for-tat strategy and the American children did not. The tit-for-tat strategy still requires revenge though. If one child cheats, the other will retaliate, and it is the realization of this that keeps the children honest. I wonder whether the emphasis on law to resolve conflicts actually atrophies the cooperative instincts in developed countries, or if American children just get turned into self-centered ninnies by TV and video games. Hmmm.

Posted by: Martin Bento at September 1, 2004 01:38 AM

Martin Bento:

While it may be true that revenge has a reciprocity-based, outcome-determining role, I think it is also important to note that even in the absence of such a role, there is nothing irrational about revenge, or altruism, or any other way in which our own utilities are affected by what happens to other people.

Rationality is about how we pursue our utilities, not about what they are. If I make myself feel better by making picky comments about other commenters, then I will be rational in making this post.

Posted by: Tom Slee at September 1, 2004 05:42 AM

I read this is Science when it first came out. It was noted there that the center activated by contemplating revenge is the same one that is activated by chocolate.

I've been musing about utility functions since then. Also, is a continuous function the best way to model this?

Posted by: Jonathan Goldberg at September 1, 2004 06:39 AM

Payback Time: Why Revenge Tastes So Sweet

A raised eyebrow was all it took.

She waited until a year after the breakup, until after he had proposed to the other woman - a model, did he mention that? - and the new couple had begun planning the wedding. That's when she ran into a mutual friend who had spent a few days staying with her ex.

"And you were, uh, comfortable staying there?" she said to the friend.

What are you talking about? he said.

And then the eyebrow arched, and voilŕ, suspicions about her former boyfriend's sexual orientation were loosed.

"Yes, I'm a Scorpio, so I'm un peu vindictive," said the woman, who swore certain payback if her name appeared in this newspaper.

Vindictive, perhaps, but also fundamentally protective. Revenge may be frowned upon, viewed as morally destitute, papered over with platitudes about living well. But the urge to extract a pound of flesh, researchers find, is primed in the genes.

Acts of personal vengeance reflect a biologically rooted sense of justice, they say, that functions in the brain something like appetite. Alternately voracious and manageable, it can inspire socially beneficial acts of retaliation and punishment as well as damaging ones. The emerging picture helps explain why many people who think they are above taking revenge find themselves doing nasty, despicable things, and how unconscious biases pervert what is at bottom a socially functional instinct.

"The best way to understand revenge is not as some disease or moral failing or crime but as a deeply human and sometimes very functional behavior," said Dr. Michael McCullough, a psychologist at the University of Miami. "Revenge can be a very good deterrent to bad behavior, and bring feelings of completeness and fulfillment."

Retaliatory acts, anthropologists have long argued, help keep people in line where formal laws or enforcement do not exist. Before Clint Eastwood and Arnold Schwarzenegger, there was Alexander Hamilton, whose fatal duel with Aaron Burr was commemorated this month on the banks of the Hudson River. Recent research has shown that stable communities depend on people who have "an intrinsic taste for punishing others who violate a community's norms," said Dr. Joseph Henrich, an anthropologist at Emory University in Atlanta.

In one experimental investing game involving four players, for example, people pay to punish others who contribute meager amounts to the shared investment pool. In another, a one-on-one exercise in sharing a sum of money, people often reject any offer from a partner that is not split 50-50 or close to it, denying both players a payoff. The participants are typically strangers who will not see each other again, Dr. Henrich said, so they are not penalizing others to develop an equitable relationship in the future. They are retaliating to enforce the rules that hold the game - and, theoretically, the community - together.

Using brain-wave technology, Dr. Eddie Harmon-Jones, a neuroscientist at the University of Wisconsin, has found that when people are insulted, they show a burst of activity in the left prefrontal cortex, a part of the brain that is also active when people prepare to satisfy hunger and some cravings. This increased activity, Dr. Harmon-Jones said, seems to reflect not the sensation of being angry so much as the preparation to express it, the readiness to hit back.

The expression itself is all pleasure. In one recent experiment, psychologists demonstrated that students who were ridiculed were far less likely to avenge themselves on an offensive peer if they had been given a bogus "mood-freezing pill," which they were told blocked the experience of pleasure.

"We've shown many times that expressing anger often escalates and leads to more aggression," said Dr. Brad Bushman, a psychologist at the University of Michigan who conducted the study, "but people express it for the same reason they eat chocolate."

Posted by: anne at September 1, 2004 08:47 AM

Sorry Martin, I came across this information some time ago and the source, sadly, has disappeared into the haze.

Anne, your mention of retaliatory acts, formal laws or enforcement, and Clint Eastwood reminds me of one of my favorite Dirty Harry lines: "I don't mind a little shooting as long as the right people get shot."

Posted by: Dubblblind at September 1, 2004 09:33 AM

Dr. Michael McCullough >>> "Revenge can be a very good deterrent to bad behavior, and bring feelings of completeness and fulfillment."<<<

Funny, revenge has always brought me a sense of guilt accompanied by feelings of powerlessness and having come up with an inferior response. As to revenge being a very good deterrent to bad behavior, one only need look at the Israeli Palestinian conflict, for example, to know that this doctor doesn't know his ass from a hole in the ground.

Posted by: Dubblblind at September 1, 2004 09:45 AM

"American children fruitlessly pushed the checker back and forth between two squares until a draw was declared by the researcher. Children in developing countries had a different response and alternately allowed each other to win by cooperating to move the checker in the same direction until one had achieved victory and then reversing the direction in each succeeding game."

I'd like a cote for this. I remember that in another gaming study, the most equitable sharers were American & W.European students.

Posted by: Tom at September 1, 2004 10:02 AM

"Developing countries" is not an actual place. The data may be good, but ascribing them to "developing countries" is no good.

Stephen Jay Gould, IIRC, told the story of poisonous seasnakes which, if eaten by a fish, would kill the fish by stinging it from inside. No benefit for the individual snake, but a lot for the species.

Economic "rationality" doesn't seem either descriptive or normative. It only exists in order to be plugged into formal systems. I find Amartya Sen's conclusions on the subject interesting, but the rationality he describes seems very common-sensical and incapable of being plugged into any equation.

Posted by: zizka / John Emerson at September 1, 2004 10:55 PM