Credit: CORBIS

Related articles

• Morality is a slave to emotion
• Unselfish brain region discovered
• Many hands do not make light work
• Apes show a range of emotions
• Darwin at 200: Saluting the founder of evolutionary theory

OUR INTUITIVE ABILITY to use reasoning to negotiate social exchanges might help us understand situations, but it doesn't necessarily motivate an appropriate response. To do that, you need something extra, such as emotion, to compel us into action.

Consider the following two scenarios and how you feel about them:

"A brother and sister like to kiss each other on the mouth. When nobody is around, they find a secret hiding place and kiss each other on the mouth, passionately."

Or this:

"A family's dog was killed by a car in front of their house. They had heard that dog meat was delicious, so they cut up the dog's body and cooked it and ate it for dinner."

A common response is a feeling something is deeply wrong about both scenarios. Yet consider what is actually wrong about them. Nobody is harmed, and there is no malicious intent. However, we can't help feel both are somehow morally wrong.

In 1993 Jonathan Haidt, a psychologist at the University of Virginia in Charlottesville, USA, presented these two scenarios to people from many different backgrounds and recorded their reactions. He found near universal condemnation of both scenarios, but when pressed, people had difficulty giving reasons for their feelings.

Haidt sees moral intuitions and emotions – disgust being one – as being at the foundation of our moral behaviour. "Emotions are nature's way of making higher animals do things that were good for them," says Haidt. "Emotions involve motivations and rewards, whereas cool reason – simply deciding that one plan is best – has no connection to motivational centres, and hence people can often know what is best yet choose something worse."

That's not to say all emotions are moral, only "those emotions that are linked to the interests or welfare either of society as a whole or at least of persons other than the judge or agent," says Haidt.

In his Handbook of Affective Sciences (2003), Haidt classifies the moral emotions into four broad categories, corresponding to how they flavour interpersonal interactions.

The first is the 'other-condemning' emotions, such as contempt, anger and disgust. Consider how you react when you find you've been robbed, or cheated, or if you see someone shirking their responsibilities. These emotions motivate us primarily towards revenge (in the case of anger) or avoidance (in the case of disgust).

Second are the 'self-conscious emotions', such as shame, embarrassment and guilt. Reflect on how you feel when you've been caught doing something wrong or have broken a social convention. These feelings encourage us to monitor our behaviour and prevent the triggering of the other-condemning actions from others.

The third type of moral emotion is the 'other-suffering' family, best known through empathy, a cornerstone of many moral codes. Think about how you respond to seeing another person get hurt, or the common practice of 'putting yourself in another's shoes'. "Compassion makes people want to help, comfort, or otherwise alleviate the suffering of the other," says Haidt.

The final class is the 'other-praising' family, which are typically associated with the good deeds of other people. We can all recollect feelings of gratitude we've had towards someone else when they've done something that has benefited us.

Together, these moral emotions – present in each and every one of us – are 'pro-social' and work together to encourage cooperative behaviour. They're evolved rules of thumb that we can rapidly apply to situations in which an immediate response is required – a kind of 'judge first, ask questions later' approach.

Steven Pinker, a cognitive psychologist at Harvard University in Boston, is one of many high profile academics who have subscribed to this view. In his 2002 book, The Blank Slate, he encourages us to abandon the notion there is no human nature and instead consider how our behaviour – and morality – may be the product of evolution.

"Many moral emotions – sympathy, gratitude, guilt, shame, trust, righteous anger – can be explained as mechanisms that make adaptive cooperation possible," says Pinker.

But this brings us to a critical juncture. If it's the case that we have some kind of hardwired capacity for social reasoning, steered by a collection of moral emotions that encourage cooperation, how could this capacity have sprung forth from our fundamentally selfish genes?

For the answer to this question, we must turn to economics.

Here's another scenario to consider. Think about what decision you'd make to achieve the best possible outcome.

"The day after you committed a daring bank heist, you and your partner in crime have been arrested under suspicion of robbery – a charge that carries a maximum penalty of 10 years in prison. However, the police have insufficient evidence to convict you immediately, but they could if one of you implicated the other – although to do that, they'd have to confess their own involvement. They interrogate you separately and offer each of you the same deal: if you both confess and implicate each other, you each receive a five-year reduced sentence for cooperating. If you both stay quiet and refuse to implicate each other, you both only get a six-month sentence for the minor charge of carrying a concealed weapon. However, if you confess and implicate your partner, who stays silent, you'll walk free, while your partner will get the full 10-year sentence. However, if you stay silent and your partner implicates you, it's them that will walk free while you get 10 years."

This is the infamous prisoner's dilemma, as articulated by the late mathematician Albert Tucker in 1950. It serves as a paragon example of a 'non-zero-sum' game from within economic game theory, which studies the interactions between two or more agents given certain rules. A 'zero-sum' game is one where there is a fixed reward, and the amount gained by the winner is equivalent to the amount lost by other players. Examples include many board games such as chess, or a presidential election, where there can be only one winner.

A non-zero-sum game is one where the reward varies depending on the players' combined actions. An example – and the reason why economists study it – is two nations trading surplus goods, where both benefit more from cooperation than not.

In the prisoner's dilemma, it's easy to conclude that you should cooperate with each other and stay silent, thus taking the minimal sentence each. Overall, this strategy would yield the best result averaged across both players.

However, if you can anticipate that your partner is thinking this as well, and they're likely to cooperate, then it's in your best interest to 'defect' and implicate them, thus walking free. Then again, your partner may be thinking the same thing, and is thus likely to defect as well, leaving you both with a worse outcome.

The end result is called a Nash equilibrium – named after the mathematician and Nobel laureate John Nash, whose life is dramatised in the film A Beautiful Mind – in which both players default to 'defect' rather than 'cooperate'.

Where the prisoner's dilemma gets particularly interesting is when the game is played repeatedly by the same players – the so-called 'iterated prisoner's dilemma'. In this case you can base your choices on what has transpired in previous turns, such as whether an opponent betrayed you and whether you want to punish them for it. Political scientist Robert Axelrod, of Michigan University in Ann Arbor, explored this in a famous 1981 paper in the U.S. journal, Science.

"My original interest in game theory arose from a concern with international politics and especially the risk of nuclear war," says Axelrod. "The iterated prisoner's dilemma game seemed to me to capture the essence of the tension between doing what is good for the individual (a selfish defection), and what is good for everyone (a cooperative choice). Therefore I was intrigued by the many strategies that had been proposed to play this game effectively."

Axelrod ran a tournament of the iterated prisoner's dilemma and invited participants to submit computer programs with various strategies. One program might always defect, another might be random while another might mimic the opposition's choice on the following turn. What Axelrod observed at the tournament was illuminating.

"The result was that the simplest of all submitted entries won the tournament. This was tit-for-tat: cooperate on the first move, and then cooperate or defect exactly as the other player did on the preceding move." A subsequent larger tournament again yielded a victory for tit-for-tat. Axelrod noticed this tit-for-tat strategy had all the hallmarks of reciprocity and subsequently went on to use the mathematics of game theory to demonstrate how such cooperation could emerge even from a population of selfish agents.

Crucially, fully cooperative – or 'nice' – strategies fared poorly, especially when confronted with a 'nasty' strategy that was unrestrained in defecting. Yet 'nasty' strategies also performed below those that were nice up to a point, yet vigorously punished nasty strategies when they appeared.

This proved to be a revelation to many – but not to Harvard evolutionary biologist Robert Trivers. In 1971 he published a study in The Quarterly Review of Biology called "The evolution of reciprocal altruism", which detailed how altruistic behaviour can arise through natural selection. He showed how humans had both selfish and altruistic tendencies, and were inclined to behave in either 'nice' or 'nasty' ways to establish a balance – effectively a Nash equilibrium – in their ecological environment.

Dominic Johnson, also of Harvard, is one of a new generation of researcher who draws on the insights of Trivers and Axelrod and applies them to diverse fields. "Game theory has revolutionised the understanding of behaviour and strategy across several disciplines, including evolutionary biology, economics, and political science," he says.

"Despite being a game that predicts non-cooperation at face value, when the prisoner's dilemma is repeated over many interactions it predicts some surprising cooperative strategies, such as cooperation with strangers and forgiveness – it is by no means obvious that such apparently complex and fundamental human behaviours should emerge from so simple a game."

Readers' comments