Spontaneous spatial mapping of learned sequence in chimpanzees: evidence for a SNARC-like effect

Abstract

In the last couple of decades, there has been a growing number of reports on space-based representation of numbers and serial order in humans. In the present study, to explore evolutionary origins of such representations, we examined whether our closest evolutionary relatives, chimpanzees, map an acquired sequence onto space in a similar way to humans. The subjects had been trained to perform a number sequence task in which they touched a sequence of "small" to "large" Arabic numerals presented in random locations on the monitor. This task was presented in sessions that also included test trials consisting of only two numerals (1 and 9) horizontally arranged. On half of the trials 1 was located to the left of 9, whereas on the other half 1 was to the right to 9. The Chimpanzees' performance was systematically influenced by the spatial arrangement of the stimuli; specifically, they responded quicker when 1 was on the left and 9 on the right compared to the other way around. This result suggests that chimpanzees, like humans, spontaneously map a learned sequence onto space.

Figure 1. Schematic diagrams of the number sequence task.

The subject's touch on the white open circle initiated a trial and random number (from 4 to 9) of Arabic numerals (baseline trials) or two numerals, 1 and 9 (test trials) were presented on the monitor. In test trials, 1 and 9 were located on the same row (either 1 on the left or 9 on the left); in baseline trials, stimuli were presented at random locations. The grid arrangement for stimulus presentations were depicted in Stimulus Display. Each grid was 130×130 pixels with a 20-pixel margin to the next grid. 4×4 grids in red were used for stimulus presentation in test trials.

Figure 2. Each subject's performance.

(a) Each dot represents mean latency in each condition for each subject. The dots connected by broken lines represent data from the same individuals. Error bars indicate standard error of the mean. (b) The effect size was calculated and plotted for each subject by dividing the subject's scores in Right-Left condition by that in the Left-Right condition. Any value over 1 suggests faster responding in the Left-Right condition than the Right-Left condition.