Most reading theories assume that readers aim at word centers for optimal information processing. During reading, saccade targeting turns out to be imprecise: Saccades’ initial landing positions often miss the word centers and have high variance, with an additional systematic error that is modulated by the distance from the launch site to the center of the target word. The performance of the oculomotor system, as reflected in the statistics of within-word landing positions, turns out to be very robust and mostly affected by the spatial information during reading. Hence, it is assumed that the saccade generation is highly automated.
The main goal of this thesis is to explore the performance of the oculomotor system under various reading conditions where orthographic information and the reading direction were manipulated. Additionally, the challenges in understanding the eye movement data to represent the oculomotor process during reading are addressed.
Two experimental studies and one simulation study were conducted for this thesis, which resulted in the following main findings:
(i) Reading texts with orthographic manipulations leads to specific changes in the eye movement patterns, both in temporal and spatial measures. The findings indicate that the oculomotor control of eye movements during reading is dependent on reading conditions (Chapter 2 & 3).
(ii) Saccades’ accuracy and precision can be simultaneously modulated under reversed reading condition, supporting the assumption that the random and systematic oculomotor errors are not independent. By assuming that readers increase the precision of sensory observation while maintaining the learned prior knowledge when reading direction was reversed, a process-oriented Bayesian model for saccade targeting can account for the simultaneous reduction of oculomotor errors (Chapter 2).
(iii) Plausible parameter values serving as proxies for the intended within-word landing positions can be estimated by using the maximum a posteriori estimator from Bayesian inference. Using the mean value of all observations as proxies is insufficient for studies focusing on the launch-site effect because the method exhibits the strongest bias when estimating the size of the effect. Mislocated fixations remain a challenge for the currently known estimation methods, especially when the systematic oculomotor error is large (Chapter 4).
The results reported in this thesis highlight the role of the oculomotor system, together with underlying cognitive processes, in eye movements during reading. The modulation of oculomotor control can be captured through a precise analysis of landing positions.
Berühren Lesen
(2021)
Berühren changiert zwischen Buchstäblichkeit und Metaphorik. Gegenüber dem Distanzsinn des Sehens wird mit dem Berühren eine größere Unmittelbarkeit assoziiert. Doch die Möglichkeit des Kontaktes ist von Beginn an prekär. Das Berühren kann sich selbst nicht berühren. In das Berühren schiebt sich ein Dazwischen, das den Entzug dieser ambivalenten Figur bedingt. Diese aporetische Bestimmung des Berührens begründet das Unternehmen des Bandes. Jeder Eintrag wiederholt eine Bewegung des Berührens: In einzelnen Text- oder Bildlektüren werden Spuren verfolgt, die das Berühren im stetigen Sich-Entziehen in seinen mannigfaltigen Nachbarschaften hinterlässt. Die einzelnen Einträge generieren sich aus diesen Lektüren. Dabei spielt die Nachbarschaft der Einträge selbst eine tragende Rolle. So wird das Berühren zum produktiven Prinzip von Philologie als einer kollektiven
Lektüre- und Schreibform.