Capuchin monkeys (Cebus apella) treat small and large numbers of items similarly during a relative quantity judgment task

Abstract

A key issue in understanding the evolutionary and developmental emergence of numerical cognition is to learn what mechanism(s) support perception and representation of quantitative information. Two such systems have been proposed, one for dealing with approximate representation of sets of items across an extended numerical range and another for highly precise representation of only small numbers of items. Evidence for the first system is abundant across species and in many tests with human adults and children, whereas the second system is primarily evident in research with children and in some tests with non-human animals. A recent paper (Choo & Franconeri, Psychonomic Bulletin & Review, 21, 93-99, 2014) with adult humans also reported "superprecise" representation of small sets of items in comparison to large sets of items, which would provide more support for the presence of a second system in human adults. We first presented capuchin monkeys with a test similar to that of Choo and Franconeri in which small or large sets with the same ratios had to be discriminated. We then presented the same monkeys with an expanded range of comparisons in the small number range (all comparisons of 1-9 items) and the large number range (all comparisons of 10-90 items in 10-item increments). Capuchin monkeys showed no increased precision for small over large sets in making these discriminations in either experiment. These data indicate a difference in the performance of monkeys to that of adult humans, and specifically that monkeys do not show improved discrimination performance for small sets relative to large sets when the relative numerical differences are held constant.

Keywords: Analog magnitude estimation; Approximate number system; Capuchin monkeys; Enumeration; Object file system; Quantity classification.

Figure 1

Mean accuracy for the monkeys with the small and large magnitude arrays for each trial block. Error bars indicate 95% confidence intervals using the Cousineau (2005) method for calculating confidence intervals for within-subjects designs.

Figure 2

The mean response times for small and large magnitude arrays and each trial block. Error bars indicate 95% confidence intervals using the Cousineau (2005) method for calculating confidence intervals for within-subjects designs.

Figure 3

The group performance of the monkeys with the small and large magnitude arrays shown as a function of the ratio between sets (A) or the difference in dot quantity between sets (B). Error bars indicate 95% confidence intervals using the Cousineau (2005) method for calculating confidence intervals for within-subjects designs.

Figure 4

Data for individual moneys with the small and large magnitude arrays shown as a function of the ratio between sets.