An Eye on Numbers: The Processing of Numerical Information in the Context of Visual Perception
The capability of understanding and processing numerical information is a critical skill that allows humans to compare, calculate, judge and remember numbers and numerosities. Without this capability, countless processes in everyday life would be very hard to accomplish. This ranges from simple acti...
|PDF Full Text
No Tags, Be the first to tag this record!
|The capability of understanding and processing numerical information is a critical skill that allows humans to compare, calculate, judge and remember numbers and numerosities. Without this capability, countless processes in everyday life would be very hard to accomplish. This ranges from simple actions like playing dice to the invention of modern techniques, such as personal computers and satellite-based navigation. Hence, it is important to understand the neural processes underlying the (human) perception of numbers and numerosities. As a contribution to this very complex research field I performed three studies using psychophysical methods and electroencephalography (EEG) with the aim to draw general conclusions on human number perception and the processing of numerical information. In the first two studies, I investigated the effect of spatial numerical association of response codes (SNARC). This effect is commonly seen as evidence for the concept of a mental number line (MNL), which is a metaphor for the fact, that the human brain organizes numbers on a mentally conceived line with small numbers on the left and large numbers on the right.
In my first study I showed the effector dependence of the SNARC effect, by measuring the SNARC effect for three different effectors: bimanual finger responses, arm pointing responses and saccadic responses. In my second study, I showed that the concept of the mental number line can be extended to a frontoparallel mental number plane, where small numbers are represented left and down and large numbers are represented right and up. I achieved this result by investigating the SNARC effect for cardinal axes (horizontal and vertical) and for diagonal axes in one and the same subject. This approach allowed me to conclude that the strength of the SNARC effect on the diagonal axes can be expressed as a linear combination of the strength of the SNARC effect along the two cardinal axes.
In this second study I measured the SNARC effect also regarding two sensory modalities (visual presented Arabic digits and spoken number words). The comparison of the SNARC effect elicited by these two modalities revealed that the strength of the SNARC effect depended on the modality of number presentation. Together with the results of the effector dependency of the SNARC effect from my first study this led me to propose the existence of a distributed “SNARC network” in the human brain. Within the framework of this proposal the SNARC effect is elicited in a central number stage (CNS) as a consequence of the interaction between numbers and space in the human brain (e.g. as explicated by the MNL). But in addition, the SNARC effect is further modulated by early, modality dependent processing stages and late, effector dependent processing stages. I hypothesize that these stages modulate the SNARC effect, but not the relationship between numbers and space per se.
My first two studies, explored the SNARC effect, based on abstract numbers represented in the, so-called, approximate number system (ANS). In addition to the number processing in the ANS, it is known that the human brain is capable of perceiving very small magnitudes (up to four) immediately, a phenomenon called subitizing. Previous studies showed that this perception, although very fast, might be influenced by attentional load (Railo et al., 2008; Olivers & Watson, 2008; Anobile et al., 2012). In my third study, I measured the neural basis of the processing of numerical information non-invasively by means of EEG and used the effect of visual mismatch negativity to demonstrate the pre-attentive processing of quantities in the subitizing range. In this experiment, I rapidly pre-sented stimuli, consisting of one, two or three circular patches. To ensure that numerosity was the relevant factor, patches were varied for low-level visual features (luminance vs. individual patch size). While participants were engaged in a difficult visual detection task, changes of the number of patches (standard vs. deviant) were processed pre-attentively. The results of my study provide evidence for the idea that numerosity in this small (subitizing) range is processed pre-attentively.
Taken together, I showed that the mental number line could be extended to a frontoparallel mental number plane and eventually even to a three-dimensional mental number space. I found evidence for the dependence of the SNARC effect on sensory modalities as well as on response effectors, suggesting the existence of a distributed SNARC-brain-network. Finally, I revealed some evidence that number processing of small magnitudes in the subitizing range might be pre-attentive.