Abstract
Enumeration of very small quantities is a common task that we perform everyday. Much research has highlighted that in these conditions humans display fast, near errorless performance, a phenomenon dubbed subitizing. It has been suggested that this regime has a pivotal role in numerosity perception. Here we asked if this system can process multiple sets of items in parallel. At odds with what happens for moderate numerosities, we found a strong impairment caused already by the introduction of a second group of items marked by a different color. Adding shape as a cue provided no benefit. The only case in which subitizing was possible was when the target and distractor group were held constant through the experimental block. These results show the surprising fact that whilst being rapid and errorless, subitizing does not have the capability to disentangle multiple groups of items and deals only with coarse stimulus statistics.
Highlights
Enumeration of very small quantities is a common task that we perform everyday
Significant difference occurs in three-color group, t(12) = 3.79, Q = 0.003, d = 1.05, BF10 = 17.89, indicating that subjects could not enumerate effectively the number of items of the target color when the target is told after the presentation of stimulus, as opposed to the probe-before condition
Halberda et al.[8] pointed out that as to most subjects, visual attention is efficient for selecting and enumerating the number of three sets in parallel: two color subsets and the superset, which is compatible with the threeitem limits of object-based attention[17]
Summary
Enumeration of very small quantities is a common task that we perform everyday. Much research has highlighted that in these conditions humans display fast, near errorless performance, a phenomenon dubbed subitizing. One of the most salient traits of a perceptual systems is whether it can operate multiple estimates at the same time with little or no cost at all as this indicates an abundance of neural resources which operate in p arallel[12,13] This property has been tested for numerosities processed by the Approximate Number System (ANS) and it was found that even if two groups of dots are interspersed and subjects are told only in retrospective which group to enumerate, they pay very little cost for doing so[8]. This indicates that approximate numerosity proceeds in parallel, suggesting localized computational modules across the visual scene. This is true even if subjects are informed in advance what will be in.cnr.it
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
