The team used rules to exclude effects that might have resulted from sex, hierarchy, age, and friendship: "Everything I could come up with," Gomes says. She concludes that the only way to explain the symmetry of grooming exchanges between pairs over time is through reciprocity. "If you don't have a set price, then you're susceptible to being cheated and cooperation would probably break down."

So far, so good, and so classical, but Gomes goes on to say that the accuracy of the exchanges is more likely to be driven by an emotional agenda than a cognitive social calculus. "It does not necessarily have to be a cognitive process," she says, "it could be emotional." Gomes hypothesizes that chimpanzees - and by extension, humans - use fine adjustments to levels of endorphins to associate particular levels of generosity or meanness with individuals, and it's then hormonal motivation that causes the tit-for-tat behaviour.

While not excluding a major affective element in relationships, which indisputably exists, this view seems to deny the very adaptive advance which made human group social life possible, that being the evolution of a cognitively and cortically based social calculus. The vast enlargement of the neo-cortex in primates and then in humans cannot be adequately accounted for other than by the need to store enormous volumes of data about the historical relationships between members of the group, and as much between other members of the group as about one's own relationships. Think soap operas. Your endorphins may help you assess the state of things between you and your mates, but they won't help you much when it comes to understanding why Joe dumped Jane in favour of Chardonnay.

As well as enabling the matrix of extended group relationships, cognitive reciprocity led to exchange in goods, ie trade, the other key building block, after the group itself, of the society of which we are all members. The notions of fairness, value and trust that originally developed and expressed themselves through behaviours such as grooming and food-sharing underlie all types of trading activity: it seems intuitively implausible, even unworkable, that they could be successfully underpinned by a hormonal mechanism rather than a cognitive one.