Research database


24 months (2010)
Principal investigator(s):
Project type:
Competitive calls
Project identification number:


The aim of this project is to advance towards the development of a realistic theory of morphological processing in the human brain. The theory is to be implemented through a biologically inspired computational model and should provide a link between the different levels of description that are involved in cognitive neuroscience. I propose to investigate the information theoretical framework developed by Moscoso del Prado, Kostic & Baayen (in press) to describe the informational complexity of recognising words in visual lexical decision experiments. We will perform 3 lexical decision experiments in English, Arabic and Serbian, investigating how the information theoretical measures affect processing in languages with very different morphological structures. For each language the measures will be calculated by means of statistical analyses on large-scale corpora. We will build a distributed connectionist model of lexical processing for each of the languages, in order to understand how the probabilistic interactions between form and meaning similarities in words give rise to the processing differences that have been observed for languages with different morphological systems. Having a detailed description of the probabilistic issues in morphological processing, I will investigate the spatio-temporal patterns of cortical activity that correlate with the informational loads. This will be achieved by means of 2 neuro-imaging experiments using the EEG and fMRI methodologies. Finally, we will use the results obtained in the previous experiments to constrain a biologically realistic model of lexical processing, based on the neuronal cell assemblies representing words described by Pulvermuller (2003) This training will equip the candidate with expertise in fMRI and EEG, which are complementary to his current background in behavioural, and computational studies on language, and will enable him to become and independent neuroscientist.