Distinct Neural Signatures for Very Small and Very Large Numerosities, Frontiers in Human Neuroscience, (11).

Behavioral studies of numerical cognition have shown that perceptual threshold for numerosity discrimination depends on the range of numerical values to be estimated. Discrimination threshold is constant when comparing very small numerosities via the mechanism called subitizing, while it increases as a function of numerosity for numbers beyond that range governed by subitizing. However, when numerosity gets so large that the individual elements start to form a cluttered ensemble, discrimination threshold increases as a function of the square root of numerosity. These behavioral patterns suggest that our sense of number is not based on a unitary mechanism and is rather based on multiple numerosity processing mechanisms depending on the absolute numerosity to be estimated. In this study, we demonstrate neurophysiological evidence for such multiple mechanisms. Participants electroencephalogram (EEG) was recorded while they viewed arrays containing either very small (1-4) or very large (100-400) number of dots with systematic variations in non-numerical cues. A linear model that tested the effects of numerical and non-numerical cues on the visual-evoked potentials (VEPs) revealed strong neural sensitivity to numerosity around 160-180 ms over right occipito-parietal sites irrespective of the numerical range presented. In contrast, earlier neural responses (~100 ms) showed markedly distinct patterns across the different numerical ranges tested. These results indicate that differences in behavioral response patterns in numerosity estimation across various numerical ranges may arise from the differences in the first stages of visual analysis. Collectively, the findings provide a firmer ground for the idea that there exists a brain system specifically dedicated for numerosity processing, yet they also suggest that multiple early visual cortical mechanisms converge to that numerosity processing stage later in the visual stream.