Eye Movement Control During the Inspection of Words Under Conditions of Pulsating Illumination

Previous work has suggested that eye movement control in reading is disturbed by pulsating illumination such as that characteristic of visual display units, which are periodically refreshed. The average size of the first saccade directed towards a word (the "entry saccade") varies as a function of screen refresh rate. We report the results of an experiment which attempted to discern whether this arises through a direct effect of pulsation on saccade control or as a result of a strategic adaptation to phenomenal image degradation which might be associated with refreshed displays. Subjects examined a prompt word followed by two target words. Depending on the prompt, a decision was required as to whether the two target words were physically identical or had the same meaning. The task was conducted in four conditions of screen pulsation (50, 75, 100 and 125 Hz) and in both positive (i.e. black-on-white) and negative (white-on-black) polarity. Since launch position into the first target word was tightly controlled, saccade extent into this should reflect only direct effects of screen pulsation. In contrast, consideration of saccade extent into the second target word, taken together with launch position in the first, allowed us to distinguish between a direct influence on saccade extent and indirect effects arising from refixations. The results support the contention that pulsation has a direct influence on saccade control. Two kinds of effect can be observed. Following steady fixation of the prompt word, screen pulsation modulates saccade extent towards the first target word primarily in negative polarity, where it is the displayed "figure", rather than the background, which pulsates. The saccade from the first to the second target word (when it may be argued more normal reading dynamics are engaged), is affected to a greater extent in positive polarity. In this case, the eyes must traverse a pulsating field when moving from word to word.