Reader Comments

Post a new comment on this article

some critical remarks

Posted by kanngies on 25 Jul 2008 at 10:55 GMT

In their article Brosnan et al report costly commodity barter experiments in two groups of chimpanzees – one with exposure to extensive linguistic and cognitive training and one without exposure. Brosnan and colleagues wanted to investigate whether enculturation would facilitate bartering behaviour in chimpanzees as in most bartering contexts chimpanzees do not treat tokens and other goods the same way humans do.

This hypothesis seems surprising given the findings published by the same authors (Brosnan et al, 2007) concerning the endowment effect - the effect to overvalue an item in ones possession over an item one does not yet possess. Using a design very similar to the one used in their current study, in 2007 the authors worked exclusively with the less enculturated of the two groups and found that chimpanzees would not exchange a food item they were endowed with for one that was close in value. These findings were used to argue that chimpanzees show the same biased behaviour as humans do (at least with respect to food) and thus that experience cannot play a significant role in eliciting this effect in humans. Therefore, if enculturation facilitates bartering behaviour, then due to the design chosen to test this hypothesis, one would get the paradoxical finding that in fact enculturation would make chimpanzees less similar to humans in their bartering behaviour.

In addition, Novemsky and Kahneman (2005) point out that a distinction has to be made between goods that are intended for trade and goods that are intended for consumption. Accordingly, they do not predict an endowment effect for goods that are intended for trade because then ‘professional’ sellers would fail to sell their goods. Commodities in general fall into the category of goods intended for trade, because the producer of a commodity usually does not intend to consume the commodity himself but to use it as an exchange good. Given this background it is not quite clear why the authors used a good like food, which functions primarily as a consumption good to chimpanzees, to investigate commodity barter.

Given the authors’ aim to compare two groups of chimpanzees with different rearing histories, one also has to wonder why the authors did not use the same study design for both groups. To give some examples: (1) the number of trading items varied between the groups, (2) preference orders were determined in different ways for the two groups, (3) the number of sessions varied between the groups, and (4) the N varied dramatically between the two groups due to the different number of subjects and the different procedures.

Nevertheless, the authors find in their experiments that chimpanzees, irrespective of their rearing history, find food barter problematic, especially if foods do not differ greatly in value. The authors claim that risk of defection (the risk that a potential ‘buyer’ could defect and run away with the commodity) may be a reason why chimpanzees find commodity barter problematic. A more parsimonious explanation, however, would be that for food items close in value, the perceived benefit of trading the endowed item in comparison to directly consuming the food at hand is very low, so that chimpanzees are averse to losing their endowed food item (loss aversion). In order to decide whether to trade or not, all an individual needs to do is to compare the benefit of the present state of affairs with the benefit of an alternative state - like comparing the benefit of giving up two grapes in order to receive four grapes with the benefit of consuming the two grapes directly. This mechanism can occur independently of social context and does not presuppose expectations about the behaviour of others as e.g. the mechanism of risk of defection does. Risk of defection in contrast, implies that one would have to evoke memories of past interactions with other individuals and so on, in addition to comparing the benefits of the trade itself. It is thus a mechanism that is much more demanding with respect to cognitive resources and cognitive processing than loss aversion.

In addition, there is an alternative explanation for why chimpanzees may find food bartering problematic that the authors neither consider nor control for in their experiments. It may simply be that the costs of trading itself may outweigh the benefits of receiving a good that has a higher value than the one currently in possession. Trading costs can generally be ruled out if one first finds the smallest food item chimpanzees would still be willing to trade for and later offers this item in addition to every item offered for trade. If chimpanzees would then start to trade, one can be confident that their reluctance to trade was caused by transaction costs.


Brosnan, S. F., Jones, O. D., Lambeth, S. P., Mareno, M. C., Richardson, A. S., & Schapiro, S. J. (2007). Endowment effects in chimpanzees., Curr Biol, 17(19), 1704―1707.

Novemsky, N., & Kahneman, D. (2005). The boundaries of loss aversion, Journal of Marketing Research, 42, 119-128.