Monday, March 13, 2006

The Conditions of Kindness

A recent paper by Michael Gurven in Current Anthropology (1) explores how choices about whether help is given to others depends on how generosity is returned.

Asking a favour from a mafia don is not without its costs. It might get you out of a tight spot, or enable you to avenge an enemy, but it comes with burdensome strings attached. The time will eventually come when you are called on to return the favour, and you had better not think about reneging on your obligation.

Even among friends, the returning of favours, or reciprocity, looms large. Most people most of the time, of course, do favours for friends and are not motivated by the prospect of a profitable return on the altruistic investment – it simply feels good to help people we like. But when the flow of favours is unidirectional, we normally notice, and it doesn’t feel good. We feel taken advantage of, which prompts feelings of resentment and, taken to the extreme, can cause the breakdown of friendships.

In the early 1970s, Robert Trivers developed the idea of reciprocal altruism (2) to explain some of the puzzles of animal and human cooperation. The basic idea is simple: you scratch my back, and I’ll scratch yours. Lets say I have a surplus of food today, and you’re going hungry. It hurts me less to give you something to eat than it benefits you (that is, although it might cost me 5 ‘health points’ to lose this food, you might get 10 points by receiving it, particularly if I’m relatively stated and you’re desperately hungry). Fast-forward to a time when the tables are turned, and I’m hungry and your larder is full: if you help me out, we’re square, and we’re both better off than we would have been if we had never helped each other (because we gained more benefit by being helped when we needed it than we lost out when we helped). In this way, a self-serving Darwinian creature can profit from entering into cooperative actions, provided it can discriminate cooperators from non-reciprocators.

The most famous (though not necessarily the best) strategy for getting reciprocal altruism off the ground is Tit-For-Tat (TFT). In this strategy, you cooperate on the first encounter with someone, and then do whatever he or she did on the previous round. So if they did not cooperate on the first move, you withold help on the next round. Likewise, if your partner cooperates, you cooperate on the subsequent move. Although TFT does well in reaping the benefits of cooperation, and of withholding help in some circumstances, it can be beaten by a range of other strategies. TFT, and reciprocal altruism in general, have limitations in explaining the long-term nature of human social interactions, and other routes to the evolution of cooperation are no doubt key to explaining human altruism.

In reciprocal altruism, the benefits of cooperation flow directly back to helpers from those they have helped before. But this needn’t be the case. Benefits can flow back to altruists can just as plausibly through indirect routes: A helps B, B helps C, and C helps A (3). If a reputation for being a good collaborator means that you get more opportunities to participate in profitable cooperative ventures, even if this is with individuals that only know of your character indirectly (through hearing of your reputation), then cooperation can pay, even in a selfish world. Such systems of indirect reciprocity are pervasive in human societies, and have even been proposed to constitute the core of moral systems (4).

A crucial feature of systems of reciprocity, and perhaps particularly reciprocal altruism, is that whether or not you give help is determined by what sorts of benefits you are going to get in return. Giving is contingent on subsequently receiving. According to reciprocal altruism, the reason that cooperation can emerge in a world of selfish egoists is that cooperation is not a zero-sum game: my gain is not necessarily your loss – we can both win. In a world of cooperators/reciprocators that shun cheats, it pays to be a cooperator.

The contingency at the heart of reciprocity can take a number of forms. For instance, giving someone some food might be contingent on getting the same quantity of food in return. Or it could depend of receiving the same proportion of the stock we gave away, regardless of the absolute amount returned. Giving might also be contingent on overall levels of exchange between whole families, rather than on an individual-by-individual basis. Alternatively, giving and sharing can depend on the amount of effort that people put into solving problems such as gathering food – it is one thing to do badly despite your greatest efforts, but another to do badly through sheer idleness. So we have here four types of contingency, what Gurven calls, in order, ‘quantity’, ‘standardised quantity’ (percentage), ‘frequency’ (of exchange between families), and ‘value’ (of effort put in or some other factor). Experiments at the interface of economics and psychology have, over recent years, provided support for the role of value in shaping what people think other people deserve out of group efforts, and Gurven’s study adds to this.

A number of theories have been put forward to explain the nature of human altruism, which stands out as an anomaly in the natural world because of the levels of help and cooperation between unrelated people in human societies. It is likely that the different theories explain different aspects of human altruism. However, they do differ in the types of contingency you’d expect to see in certain cooperative and altruistic actions, and so studying them can help determine which processes are operating in which situations.

Unfortunately, little attention has been paid to the different forms of contingency and their roles in regulating altruistic behaviour. So Michael Gurven, an anthropologist at the University of California at Santa Barbara, set out to explore these issues using data previously collected by Gurven and other anthropologists in two populations: among the Ache of Paraguay, and the Hiwi of Venezuela. Through a number of statistical analyses, Gurven demonstrates that contingency does play an important role in food sharing among these populations, and also that different forms of contingency operate in different contexts.

The Ache and Hiwi live in different ecological niches, and collect and consume a range of food types (for instance, the Hiwi lived near a river and therefore had access to fish). Gurven grouped food types together, and analysed the role of contingency in governing whether and how they were shared. The Ache diet was categorised into forest foods (such as meat and honey), ‘cultigens’ (such as sweet manioc, corn and sweet potatoes) and store-bought foods (such as bread and oil); Gurven also looked at contingency in ‘all food types combined’. Contingency among the Hiwi was examined by grouping food as meat, fish, ‘other foods’, which included fruit and roots, and ‘all foods combined’.

Gurven found strong evidence for contingency in sharing meat and fish among the Hiwi, although this wasn’t seen for resources grouped as ‘other foods’ (fruit and roots). On average, for every kilogram of meat given to another family 0.69 kg was given back; for other foods, the return rate drops to 0.08 kg for every kilogram given. Among the Hiwi, the form of contingency called ‘quantity’ was the most prominent in the exchange of meat and when all resources were considered together; ‘value’ had an effect similar in magnitude, though not quite as great. The transfer of fish among the Hiwi seemed to be predominantly contingent on standardised quantity (percentage).

Among the Ache, frequency and value contingency were most important for forest foods and cultigen transfer, and value stood out as an important determinant of giving when all foods combined were considered together.

The lack of contingency in the giving of non-meat (‘other’) foods is interesting – what is it about these resources that makes people share them differently? Roots and fruits, while making up more than 40% of the Hiwi diet, are the least transferred resources. A number of factors explain why giving of these foods is less contingent than for other resources, and why they are not shared much in the first place. First, the existence and location of fruit and roots, unlike animal game, is highly predictable. This means that there is low variability in the amounts of these resources that foragers return with (that is, collecting these resources is less subject to the vagaries of chance). Second, individuals typically gather fruits and roots at the same time, and are therefore usually stocked up or not at the same time. These two factors reduce the need to exchange these foods in the first place: you’re more likely to be without meat or fish than without fruit or roots.

These anthropological results tie in with studies in behavioural economics that reveal that people are motivated by notions of fairness based on labour input into collective actions. The notions of fairness built into human psychology give rise to, and are probably reinforced by, cultural norms that explicitly spell out what is fair and what is not. Gurven suggests that thinking about the types of contingent cooperation seen in his anthropological survey could “begin to bridge the gap between the short-term calculus of reciprocal altruism and the longer-term social relationships emphasized in cultural norms.”

It is important to recognise that although reciprocal altruism and TFT are highly contingent, the finding of contingency in the food sharing of the Hiwi and Ache does not mean they are engaged in a TFT strategy. It seems that the forms of contingency observed, and the motivations driving cooperative behaviour, are the product of psychological systems, buttressed and canonised by cultural norms (and also perhaps in part shaped by them), that promote long-term collaborations in a way that TFT cannot.

The value people attach to the effort other people put into collective actions, and their altruistic intentions, has, according to Gurven, been neglected in past explorations of human cooperation in the anthropological literature. Given the recurring importance of value-based contingency found by Gurven, more attention to value should be paid in future studies. In general, the behavioural outcomes identified by anthropologists and other students of the human social sciences need to be linked up with work on the psychological underpinnings of human cooperation. A problem as complex as human altruism is surely going to require a pluralistic, inter-disciplinary approach to clearly illuminate the multifarious facets of this perennial question.

1. Gurven, M. The evolution of contingent cooperation. Current Anthropology 47, 185-192 (2006).

2. Trivers, R. L. The evolution of reciprocal altruism. In Natural Selection and Social Theory: Selected Papers of Robert Trivers 18–51 (Oxford University Press, 2002).

3. Nowak, M. & Sigmund, K. Evolution of indirect reciprocity by image scoring. Nature 393, 573–577 (1998).

4. Alexander, R. The Biology of Moral Systems (Aldine Transaction, 1987).


Anonymous wmr said...

Are you familiar with Tor Norretranders's The Generous Man: how helping others is the sexiest thing you can do? He is a Danish science journalist and in this book he explores the case for altruism as the human equivalent of the peacock's tail.

10:32 pm GMT  
Blogger Dan Jones said...

I hadn’t heard of that book, but I know the author's name – he wrote a book on consciousness called The User Illusion that I think is supposed to be quite good.

The idea of altruism as sexual display was aired by Geoffrey Miller in The Mating Mind, and also lies behind the ‘costly signalling’ theory of human altruism, and I’d be keen to read a good book-length treatment of the topic. Thanks for the heads up.

11:21 pm GMT  
Anonymous wmr said...

I'm no expert so I can't say whether this is a "good book-length treatment of the topic" and it is definitely a shorter book than User Illusion. I am primarily entertained by the unexpected connections he makes - unexpected by me, at least.

By the way, I came here via your comment at Butterflies and Wheels, which I responded to there.

6:36 am GMT  

Post a Comment

<< Home