Anthropological Universals – Cultural Differences. Comparative research on differences in cognitive processing of the German alphabetic orthography and the Japanese writing system. An interdisciplinary project between philosophy, psychology, and neurolinguistics.
The aim of the project is to gain insight into the cognitive processing of different writing systems and thereby into the influence of universal cognitive mechanisms and culture specific factors on our symbolic practices.
Our approach is a comparative analysis of an alphabetic writing system (the German orthography) and the Japanese writing system (a complex "mixed" writing system associating characters of Chinese origin, the Kanji, and two syllabaries). We focus on monomorphematic alphabetic nouns and Japanese Kanji. The project combines linguistic theory, an experimental-behavioural approach and connectionist modelling of cognitive processes.
The theoretical description is based on Nelson Goodman's theory of symbols. It provides a language of description that is not biased towards either writing system and permits a comparison of these systems. Furthermore, elements for a typological description of writing systems are proposed. More specifically, the description establishes which units should be considered as equivalent in both writing systems and can be used to create comparable stimulus material for testing.
The experimental approach uses two different methods to determine whether different processes are involved in word recognition in different scripts: 1.) A visual word/character recognition task using the lexical decision paradigm allows inferences about cognitive processes via the analysis of reaction times and error rates. 2.) The same task is used in fMRI scans in order to shed light on the neurological basis of visual word recognition.
The third part of the project combines theoretical findings and the results of lexical decision experiments within the framework of a connectionist model. This cognitive model of word recognition in different scripts is then implemented in a computer simulation of the neural network of word recognition.