Abstract: Number sense means our ability to quickly understand, approximate, and manipulate numerical quantities. The adaptation effect of numerosity refers to the adaptation to the number of stimuli, with an aftereffect that can be revealed by the changes of perceived numerosity in the following numerosity judging tasks, in which numerosity adaptation effects can be quantified by asking subjects whether a test stimulus presented to the adapted region appears more or less numerous than a probe stimulus presented to the un-adapted area. The current debate of numerosity adaptation effect focuses on whether it is really driven from the numerosity of stimuli, or resulted from processing of surface information. In order to testify the independence of the numerosity adaptation effect from texture information, previous studies have concentrated on the influence of related spatial variables, such as dot density. In those experiments, different non-numerical surface characteristics were manipulated, while the perception of numerosity was not taken into consideration, in spite of its important links with surface characteristics of stimuli and adaptation respectively. So, the relationship among numerosity adaptation, numerosity perception and surface characteristics of stimuli was examined in our study. By changing the size of dots or grouping them into chunks, we investigated the influence of surface information on numerosity perception and adaptation. In experiment 1, two stimuli (dot sets) were presented horizontally and simultaneously, and they consisted of dots with different sizes (0.16°×0.16°as the small dimension, 0.16°×0.16°as the large dimension). Participants were asked to compare the numerosity of dots within the two stimuli. The result suggested no size-related influence in the numerosity judgment. In experiment 2, the 405-dot-stimuli were subjectively grouped into 45 chunks according to the spatial distribution and grayness information, each of which was composed of 9 dots. Participants were asked to estimate the numerosity of either chunks or total dots, and the estimation of 45 or 405 discrete dots were set as control. Subjects apparently underestimated the number of total dots, compared to the control condition. In experiment 3, one the hand, we found that the size of dots had no significant influence on numerosity adaptation, as long as the perception of their numerosity was unchanging, despite their apparent differences in surface characteristics. On the other hand, perceptual grouping of dots had a significant effect on numerosity adaptation, and a changed adaptation effect based on the changed perception of numerosity was observed under that condition. In conclusion, the present study provides evidence that adaptation of numerosity is based on perceived numerosity. Surface information does not affect the numerosity adaptation effect unless it affects the perception of numerosity as well. Humans correctly judge the number of dots with different sizes. Grouping has an effect on perceived numerosity, that is, there is a significant underestimation of total number of dots when they can be grouped into chunks. The numerosity adaptation effect keeps constant as long as the varied texture characters had no effect on the perception of numerosity. Otherwise, it will be apparently affected when the perceived numerosity is interfered with the changing of texture. Therefore, number adaptation effect is an adaptation based on perceived numerosity, rather than a co-product driven from the processing of surrogate surface features.

Key words: perception of numerosity, numerosity adaptation effect, perceptual grouping, object representation