From the viewpoint of cognitive science and related disciplines, reading can be conceived as visual information extraction from written text. This process continues at higher levels of cognition by comprehension of text and memory coding of the information provided by the text. Accordingly, one of the basic research topics in reading research has been the relation between reading patterns and cognitive processes. Because cognitive processes are not directly observable, the available research mostly investigates eye movement detection and the framework of the assumptions about cognitive processes.
The primary goal of this project is to investigate this assumption, which underlies the discussion on eye movement control models in reading, by employing Turkish as an agglutinative language application domain. For this, it is proposed to develop a corpus of Turkish reading patterns. This corpus is planned to provide data about fixation location distribution on words, fixation duration and saccadic amplitude, from about 200 participants who normally read about 150 sentences both silently and orally. Those data then are planned to be used for testing the word-length assumption in the recent eye movement control models. The sentences for the corpus will be developed by considering frequency and length of both root form and the inflected form of the words. The predictability scores of the words within sentences will be calculated by the participation of another group of approximately 300 participants. The findings that will be obtained from the recent research proposal will have the potential to contribute to reading research literature and it has the potential to initiate international collaborations. The recorded eye movement patterns of reading in the project may be used for analyzing reading disorders such as dyslexia, in future. Accordingly, the project has the potential to contribute to the society.