The Mental Representation Of Small And Large Number

The numerical competence is one of the most basic cognitive abilities in human.Over the last decades of years, the study of numerical ability has become a majorresearch field within cognitive science. Converging evidence from behavioral andcognitive neuropsychological studies has led many investigators to conclude that thereare two separable core systems of number representation in many species:approximate number representation and precise number representation. The formermainly reflects an estimating competence of quantities and it is language-independent;the latter may recruit language-related network, especially the verbal counting, exceptthe exact small-number representations of distinct individuals. For example, Careyproposes that linguistic factors play a crucial bootstrapping role in the exact numberrepresentation,exact calculation and the development of number concepts, throughearly experience of number-relevant language. A contrasting view is offered byGelman and Galliste, who propose that precise number representation is alsoindependent of language as approximate number representation. And some peoplesuggest that numerical cognition is ontogenetically independent. So, the question ofwhether there are two different types of number representations or not and theintegration of preverbal and verbal systems in the development of numberrepresentation are the focus of much debate.Therefore, three experiments were conducted in this study, aimed to examine therole of language in exact number representations by testing the forms of numberinformational representation. In experiment 1, we examined whether therepresentations of small and large number were different, through comparing theattentional blink magnitude between four stimuli combinations (T1: small or large quantities; T2: small or large quantities). There were: S-S, L-L, S-L and L-S. Inexperiment 2, we explored detailed mechanisms underlying the small and largenumber representations, through examining if the same T1 letter discrimination taskinterferes for T2 small or high quantities discrimination task. In experiment 3, weexamined two hemisphere's abilities to numerical representations. The small and highquantities were presented in left or right visual field.The results show: (1) The types of the small and large number preciserepresentations were different, having independent resource pathways. (2) The smallnumber concept may be based on spatial representation which is independent oflanguage; the large number concept may be based on semantic representation whichneeds phonological loop for maintaining number words and Arabic symbols bysub-vocal rehearsal. (3) The small and large number concepts may be also stored intwo separable working memories: as spatial representation, small number stored invisuospatial working memory; while semantic large number stored in verbal workingmemory. (4) Results of the study supported the competitive neural model of smallnumber detection: the spatial representation of small number mainly has developedwithin the "where" processing stream, whereas the semantic representation of largenumber associates with "what" processing stream.