Numerical Representation Is Not Abstract

The studies of Numerical Cognition have achieved a significant number of results in the past 30 years by psychologists.However,the question of whether the numerical representation is abstract or not abstract is still controversial.The dispute focus on two points: 1)the input form of the nonsymbolic number itself is changeable.Approximate number system theory believes sensory information will not affect the process of the nonsymbolic number,while the sensory-integration system theory does not think so.2)the input forms are different between nonsymbolic number and symbolic number.ANS mapping theory believes that the two share the same type of representations while discrete semantic system theory proposes the opposite view.These disagreements are based on the results of the standard comparison tasks.Researchers often use the comparison distance effect and comparison size effect in standard comparison tasks to embody the process of numerical representation.However,the monotonic connection view doubted the reliability of the comparison distance effect and believed that it not only reflects the numerical representation ability but also the general response-related processes.Hence,the aims of this study are 1)Confirm whether the priming comparison task is a more scientific experimental paradigm than standard comparison task.2)Compare the experimental results in the standard comparison task with the previous.3)Reinvestigate whether the number is an abstract representation from priming distance effect and priming size effect in priming comparison task.The purpose of experiment 1 is to explore the effect of response-related processes on two tasks.In order to avoid the distance effect caused by the numerical representation,the stimulus are the letters that do not contain cardinality,to check the performance of standard numerical comparison task and priming comparison task.The result showed that the comparison distance effect and comparison size effect occurred in the standard comparison task.The priming distance effect and priming size effect did not appear in the priming comparison task.This result supports the monotone connection view.The objective of Experiment 2 is to explore the representation mechanism of nonsymbolic number.The experimental sets dots array by a more advanced method to control for the visual cues and value ranges.Investigate the behavior pattern of standard comparison task and priming comparison task.The result demonstrated that there was a significant effect in comparison distance effect and comparison size effect when the visual cues are entirely random.The effect is also significant in the priming distance effect and priming size effect.However,when controlling the congruency of the size of convex hull and the value,the congruency effect of visual cues is significant.The result supports the Sensory integration system theory.The objective of Experiment 3 is to explore the representation mechanism of the symbolic number.The stimulus are Arabic numerals and Chinese numerals.Compare the representation-related effects in standard numerical comparison task and priming comparison task.The result showed a significant comparison distance effect and comparison size effect,and the two effects do not correlate.Besides,significant priming distance effect and reversed priming size effect is also not related.The result supports the discrete semantic system theory.This study reached the following conclusions through three experiments:1)Compare with the standard comparison task,the priming comparison task avoids the influence of response-related processes on the experimental results,which reflects the individual's numerical representation ability better and is a more reasonable experimental paradigm.2)When processing nonsymbol quantities such as dots,representation-related effects are sensitive to the input form of numerical information.This result supports the idea that number representation is not abstract.3)Representation mechanisms are different in symbolic number and nonsymbolic number.The quantities of different forms also did not show a consistent behavior in the priming comparison task,which supports the view that number representation is not abstract.