Abstract
Humans and many other animal species share the ability to estimate the numerosity of an ensemble of object. How this numerosity estimation mechanism works is still a matter of debate, but recent studies suggest that numerosity is a spontaneous perceptual ability, not mediated by continuous magnitudes, such as the size or density of the ensemble. Here, we tested whether numerosity can be perceived independently from other properties of the stimulus using a novel technique, continuous tracking, where participants are presented with a stimulus varying in some property and are asked to follow these changes by moving a cursor. The temporal similarity between changes in the stimulus and cursor movements provides information about participants’ perceptual abilities. Participants where shown a stimulus varying independently in both numerosity and density and were asked to track either the numerosity of the stimulus or its density, in separate experiments. Results show that participants where able to track the numerosity of the stimulus without being biased by its density, but not the opposite. A second experiment with slower rate of changes (one every 200 ms in the first experiment and one every 800 ms in the second) showed that participants were less biased by numerosity when tracking density, but changes in density were still irrelevant when tracking numerosity. Participants’ results were compared to an ideal observer, a virtual observer with 100% accuracy, showing that participants’ behavior in the numerosity experiments is best reproduced by a virtual observer responding only to the changes in numerosity, while in the density experiments it is best described by an ideal observer responding half to changes in density and half to changes in numerosity. This result shows that numerosity, but not density, is perceived spontaneously, without the need to rely on other perceptual attributes.