ABSTRACT
This chapter focuses on a research which investigates the effects of picture context on eye movement patterns and, by inference, on cognitive processes underlying visual exploration. Biederman has attempted to specify the characteristics of a contextually coherent "well-formed" scene that distinguish it from an array of unrelated objects in terms of relationships among objects. The well-formed scene is limited by two types of constraints on object relationships, syntactic and semantic. The 23 pictures used in this research, taken from Antes and Metzger, consisted of line drawings of varying themes, and contained a number of readily identifiable objects. The subjects were 21 undergraduates at the University of North Dakota who reported normal vision without eye glasses or contact lenses. Eye movements were recorded by means of a Mackworth corneal reflection instrument from which a permanent data record was obtained on 16 mm film.
