Funded by: American Society of Primatologists
PI: Dr. Elsa Addessi co-PI: Fabio Paglieri
Timeframe: 2009-2010
Description: This project aimed to provide a fair comparison of human and nonhuman primates’ ability of delaying gratification and to test a novel hypothesis on the effect of reward types on impulsivity, the Consumption-based Discounting Hypothesis. According to this hypothesis, people may discount future utility taking as reference point the expected time of consumption of a given resource rather than its expected time of acquisition. Thus, the greater impulsivity observed in humans with food than with money might be due to the difference between food and money in terms of expected time of consumption.
In Experiment 1 capuchins were tested in an intertemporal choice task with an adjusting delay procedure and showed a considerable delay tolerance, higher than that of closely related species such as marmosets and tamarins and comparable to that shown by bonobos and chimpanzees. The ability of spontaneously using tools shown by several capuchin populations might explain their comparatively high preference for delayed options in the intertemporal choice task.
In Experiment 2 we presented both capuchins and humans with a series of choices between a small immediate option and a large delayed option with three types of rewards: (i) primary rewards (2 vs. 6 pieces of food, Food delay condition), (ii) secondary rewards (one low-value token, exchangeable for two pieces of food, vs. one high-value token, exchangeable for six pieces of food tokens, Token delay condition), and (iii) “packed” primary rewards (food rewards provided inside a transparent box, which cannot be consumed until the box is opened, Food box condition). Moreover, we presented humans with: (iv) food rewards that could be consumed only at the end of the experiment (Food-to-be-stored condition), and (v) 20 cents available immediately vs. 60 s available after 80 s money (Money delay condition).
On the basis of the Consumption-based Discounting Hypothesis, we expected both humans and capuchins to choose the larger delayed reward more in the Food box or in the Token delay condition than in the Food delay condition, thus proving that consumption times play a key role in determining impulsivity. However, contrary to what we expected, capuchins chose the larger delayed reward more in the Food delay condition than in the Token delay condition, and in humans we observed the same pattern, although the difference was not significant. Moreover, in both populations delayed consumption times (in the Food box condition, or in the Food-to-be-stored condition) did not increase the choice of the larger delayed option.
Nonetheless, although in the opposite direction to that predicted by the Consumption-based Discounting Hypothesis, tokens influenced capuchins’ choices in the intertemporal choice task, in that they significantly reduced the choices for the larger delayed option compared to food rewards. Given that in the intertemporal choice task subjects’ preferences are affected by both the quantity and the delay of the options, it can be hypothesized that the preference for the larger delayed option is partly due to an impulsive preference for the quantity rather than to a high tolerance to delay.
