The eye tracks the aesthetic appeal of sentences

Eye-tracking parameters (fixation and pupillary responses) have been shown to be modulated by the aesthetic perception and evaluation of visual and auditory artworks (e.g., paintings, music). The present study investigated whether similar effects can be found in visual text processing. Participants read four groups of short sentences in which a key predictor of aesthetic liking, i.e., familiarity, was systematically modified to four degrees. Across all four groups, the sentences moreover varied with regard to featuring or not featuring meter. During reading, pupil sizes and eye movements were recorded. Aesthetic ratings of all sentences were collected afterwards, and the relationships between the ratings, levels of familiarity, meter, and eye-tracking datasets were tested. The results showed that the rating scores were interactively modulated by both familiarity-driven and meter-driven fluency. Using factor analysis, we extracted two key factors of the aesthetic appeal of the texts: an affective and a cognitive factor. The cognitive factor comprised the rating items "succinctness" and "familiarity," whereas the affective factor reflected the ratings for "beauty" and "liking." A higher cognitive factor predicted shorter dwelling time. Moreover, the two factors modulated the pupillary data antagonistically: A higher affective factor predicted larger pupil dilations, whereas a higher cognitive factor predicted smaller pupil dilations. The study shows a possible application of the eye-tracking method for capturing aesthetically evaluative dimensions of processing sentences.