The Psychobiology of Co-operative Behaviour

The phrases 'Survival of the fittest' and 'Nature, red in tooth and claw' are associated with Darwin's theory of evolution through natural selection. They portray a world populated by selfish organisms intent on their own survival. This lecture examines the paradox of altruism. Why do we co-operate with one another? Why do humans and other animals make self-sacrifices that benefit other members of their species?

For example:

• why do bees sting?
• why do animals emit alarm calls that reveal their presence to predators?
• why are some poisonous insects brightly coloured?

These are examples of altruism or self-sacrifice. The existence of altruism posed a significant challenge to Darwin's theory of evolution through natural selection. Recent evidence has highlighted the importance of co-operation to evolution. It turns out that an answer to the altruism-paradox may involve paying more attention to 'selfish genes' rather than selfish people.

Hamilton's kin selection theory

Kin share many common genes. Genes can spread by benefiting other carriers of the same gene. This is called kin selection. Hamilton used kin selection to explain altruism or self-sacrifice. In an altruistic encounter there is:

• an altruist or donor who bears a cost (c)
• a recipient who gets a benefit (b)

The probability that the altruist and the recipient share a gene is called the coefficient of relatedness (r). The diagram shows the extent to which we share genes with our relatives. The value of r varies between 0 and 1. On average we share half of our genes with our brothers, sisters and children (r=0.5), and a quarter of our genes are identical with those of our grandchildren, nephews and nieces (r=0.25)

According to Hamilton's Rule altruism pays off if rb>c. In other words, shared genes will profit if the cost to the altruist is less than benefit to the recipient multiplied by the probability that the recipient shares genes with the donor.

Costs and benefits are expressed in units of fitness or reproductive success with values between 0 and 1.

• A cost of 1 unit of fitness means that the act would reduce the donor's reproductive success by 1 offspring.
• A benefit of 1 unit of fitness means that the recipient would increase their reproductive success by 1 offspring.

For the sake of argument assume you have spare food that you could give to your brother to feed him and his children.

• Assume that the cost to you is 0.1 units of fitness (i.e. if you do this 10 times you will have one less child)
• Assume that the benefit to your brother is 0.25 (i.e. if he receives 4 such donations he will have one more child)

We can test if your altruism would benefit kin selection by putting these values into Hamilton's Rule rb>c where:

• r ( the coefficient of relatedness between you and your brother) = 0.5
• c (the impact on your reproductive success ) = 0.1
• b ( the benefit to your brother's reproductive success) = 0.25
• rb =0.5x0.25=0.125
• c=0.1
• because rb>c (0.125 is greater than 0.1) Hamilton's Rule is satisfied and your altruism would benefit your genes

You might wonder why b and c are not always equal. Why not use the spare food youhave to increase your own reproductive success? Well there is a limit to how much you can eat. If you have an abundance of food and your brother is starving, the cost to you of sharing is small, but it may be a matter of life or death to your brother and his children.

Kin selection theory in action

Bernstein et al (1994) and Petrinovich et al (1993) asked people to imagine how they would react in life-and-death situations in which they could prevent the death of a relative. For example imagine that a train is running out of control down a track that branches. You are in charge of the points at a track junction. If the train goes down the left track it will kill one of your relatives. If you send the train down the other track it will kill two strangers. the choice is yours!
	The results reflect what you may have predicted; we are more likely to help a near relative (e.g. brother or sister (r=0.50) than a stranger (r=0.00) in a life-and-death situation.

But what about the age of the relative involved. Would you have predicted that we are less likely to help a younger relative than one who is slightly older? One explanation for our preference for helping young people is that they have higher reproductive value than children or older people. A person's reproductive value is a measure of the probability that they will have children. Older people have lower reproductive value because fertility declines with age. children have lower reproductive value because they may die before they reach puberty.

Inheritance of wealth

Analysis of bequests (see Cartwright, 2000) made in wills shows that people leave more of their estate:

• to close relatives than distant relatives who have a lower coefficient of relatedness (r)
• to their children than to their siblings (brothers and sisters), even though offspring and siblings have the same coefficient of relatedness (r=0.5) with the deceased. Why do you think this happens?

Trivers' reciprocal altruism theory

Trivers (1971) has suggested another theory of why we co-operate with other people. This works on the principle of "if you scratch my back, I'll scratch yours" This theory depends on important preconditions:

• costs should be small compared to benefits
• the recipient should be able to recognise the donor
• the donor and recipient should regularly exchange roles

The prisoner's dilemma

The prisoner's dilemma game is a way of studying how people behave in a situation with different rewards for co-operating with, and cheating on, another person. For the purposes of this game you are asked to imagine that you and a colleague have committed a crime. You enter into a pact with your colleague-in-crime in which you both agree to say nothing to the police if you are interviewed. Subsequently the police interview both of you in separate rooms. They offer you a 'deal'. They have enough evidence to charge you both with a relatively minor crime for which you will both serve one year in prison. If you both confess to the major crime, you will each receive a five year sentence for co-operating with the authorities. But here's the twist! If you confess and your colleague does not confess, he will receive a ten year sentence and you will not be prosecuted. Likewise if you keep quiet about the major crime, but your friend confesses, you will serve ten years in prison and he will walk free. The choice is yours!. Here are your options:

As you can appreciate there is a great temptation to confess and thereby cheat on your earlier agreement with your colleague. If you were to play the prisoner's dilemma only once this might be your preferred strategy, but what happens if you had to face the same situation again. This variation of the game is called the iterated prisoner's dilemma. In this variation of the game you may be punished by your colleage on the next round of the game.

There is a great deal of interest in the prisoner's dilemma game because it is a model for real-world situations in which people and animals co-operate and compete with each other. For example the proliferation of nuclear arms during the Cold War can be analysed in terms of prisoner's dilemma strategies. Likewise the problem of persuading people to accept the inconvenience associated with abandoning private cars in favour of public transport, or separating their domestic rubbish for recycling are variations on the theme of the 'tragedy of the commons'.

What is the best strategy to adopt in the iterated prisoner's dilemma game. There are various possibilities:

• always co-operate - the 'dove' or meek and mild stategy
• always cheat - the 'hawk' strategy
• always retaliate - the tit for tat strategy
• cheat on every third game
• co-operate unless your colleague cheats twice in a row
• etc.

It doesn't take much imagination to realize that finding the best strategy would be very important for foreign policy advisers to leaders of world superpowers.

It turns out a good strategy is to play tit for tat :

• co-operate on the first game
• never be the first to cheat
• retaliate on the next game if the other player cheats, then co-operate if the other player responds by co-operation

Soldiers in trenches on the Western front in World War 1 engaged in tit for tat shooting. The commanding generals stamped out this behaviour by court-martialling some soldiers.

Picture: "Humanity - Stretcher-Bearer Post, 9th Field Ambulance" by Gilbert Rogers (Official War Artist). From Swedish University Network SUNET Archive ftp.sunet.se

But tit for tat can run into problems if you make an error. For example you might loose concentration, and imagine that the other player has cheated when in fact they made a co-operative move. A slight modification of the strategy overcomes this problem - only retaliate if the other player cheats twice in a row - the generous tit for tat strategy. Cartwright (2000) provides a clear explanation of the effectiveness of these strategies and variations such as the Pavlov 'win stay, lose shift' strategy.

Requesting and giving help

Gaulin & McBurney (2001) review the conditions that encourage us to ask for, and give help.

We are more likely to give assistance if:

• the cost is low
• the need is urgent
• the person we help will be able to help us in future i.e. they can reciprocate our behaviour
• our repution is at stake or our behaviour will be observed by others

We are more likely to ask for help if:

• we feel we will be in a position to reciprocate the helper in the future

Ultimatum games indicate that people are more altruistic than mathematical game theory would predict.

Imagine I have £10 which you can split with someone else. If the other player accepts the split then you each get the amount you specified. If your offer is rejected, then you both get nothing. How much would you offer to share?

Game theory predicts that the best strategy is for you to offer £1 and keep £9 for yourself. But the majority of people offer the other player £5 and keep £5 for themselves.

Maybe you are worried that you will be punished for being mean. The other player might reject £1 in order to punish your selfishness.

Would you change your offer if the rules of the game are changed so that the other player cannot reject your offer?

This manipulation does have an impact on what people offer, but only 20% of subjects give away £1 and keep the remaining £9. And more than 20% of players split the money equally with the other player.

One explanation for people's niceness in ultimatum games is that they feel their reputation would be damaged if they behaved selfishly. However even under 'double-blind' conditions - in which anonymity is preserved - some people still fail to act selfishly. Humans are exquisitely sensitive to cheating. We scan our environment to detect people who cheat: we remember the faces of people who have been introduced as cheaters more effectively than pictures of non-cheaters we dislike 'free-riders' we are interested in gossip about the exploits (often sexual) of people we know we take steps to preserve our reputations (Gaulin & McBurney, 2001) we feel guilt after cheating or failing to reciprocate a kindness Do you know who these men are, and why they are famous?

Points to ponder

'Ethics is an illusion fobbed on us by our genes to get us to co-operate'

'Our belief in morality is merely an adaptation put in place to further our reproductive ends'

'Morality is a means whereby individual human beings attempt to induce altruism in others in their own self interest'

'I am just going outside, and I may be away some time'

References

• Burnstein et al, Journal of Personality and Social Psychology, 67, 1994
• Cartwright, Evolution and Human Behaviour, Macmillan Press, Basingstoke, 2000
• Gaulin & McBurney, Psychology: An Evolutionary Approach, Prentice Hall, New Jersey, 2001
• Petrinovich et al, Journal of Personality and Social Psychology, 64, 467-478, 1993
• Trivers Quarterly Review of Biology, 46, 35-57, 1971