From quantity to number : Studies on magnitude processing

The ability to work with the symbolic number system, the Arabic numbers, enables humans to perform more complex and accurate calculations than the non-symbolic numerosity system (e.g. arrays of dots). The application of this symbolic system has turned out to be so useful that we have a hard time imagining the world without it. This makes the study of questions about the acquisition and implementation of these processes in the human brain very relevant. In this thesis we therefore investigated the brain processes involved in learning the symbolic number code, and we investigated whether symbolic and non-symbolic numerosities are processed in a similar manner. In addition, we looked at associations between magnitudes and other sensory or cognitive processes in synesthetes. For all these studies we used Stroop-like or priming paradigms, where congruent and incongruent stimuli generally result in different response times, and Event related potentials as a measure for the underlying neural correlates. From the results we can conclude that education plays a major role in the automatisation of symbolic number knowledge. Automatisation means that the child directly and implicitly knows that for instance the number symbol 5 refers to the quantity 5. Children that just learned and automatised the number symbols rely on distinct brain processes than children that have more experience with Arabic numbers. This reliance on different processes is not the result of not yet fully developed cognitive abilities such as attention and working memory instead the brain is presumably not functionally specialized yet. In contrast to the learned symbolic code, humans can already differentiate between non-symbolic numerosities (e.g. groups of dots) at a very young age. Adults rely on the same mechanisms as infants, but are less accurate. This could be due to the larger reliance on the symbolic code with increasing age or to the fact that infants are more prone to numerosity changes (in a task that does not require any responses) than adults are. The non-symbolic numerosity system is suggested to form the basis for the later acquired symbolic system. It has therefore been suggested that these initially distinct numerosities are transcoded into a general magnitude code. Our results underline this notion but also reveal support for an alternative explanation. Similar task related processes such as categorization might also explain this interaction at an early stage of processing. Last, we investigated the interaction between arbitrary associations in grapheme-color synesthetes. Synesthetes experience a color when a grapheme is presented but not the reverse. In our studies we investigated whether colors do activate number processes even though the synesthete is not aware of it (synesthetic number experience). Our results revealed that at the behavioral as well as neuronal level the percept of a color does activate number processes. These processes appeared to be differentially activated for subgroups of synesthetes suggesting that synesthesia is not a heterogeneous phenomenon