Psychophysical and eye movement measurements showed that processing speed and eye movement patterns are determined both by the visual resolution needed for the safe recognition of the critical details and by the cognitive skill to quickly process the recognised images (Zangemeister et al. 1995, Krischer et al. 1998). Based on such findings a model of reading eye movement control was suggested (Krischer et al. 1998) in which saccadic amplitudes are automatically chosen to allow both safe sensory decoding and cognitive processing during fixations of approximately constant duration. We now present reading eye movement data showing how saccadic frequency can decrease either because sensory performance is degraded by slowing the retinal function or by presenting letters too small to be quickly processed during the (normal) fixations.

In measuring the duration of reading fixations as a function of the quality of visual input we show that fixation duration remained constant at about 300 ms for all display conditions provided retinal function was intact (i.e. for blurred display) but increased to 400 ms if retinal function was slowed by dimming.

In a re-evaluation of classical developmental reading eye movement data by Taylor (1965) a new diagram shows that fixation duration increased from 270 to 333 ms below grade level 3 (reading speed < 50% of the normal value). In Taylor's experiments the letters of normal size were obviously too small for comfortable processing.

Conclusions: The data showed that a prerequisite of smooth visual processing with constant saccadic frequency is appropriate imaging in the centre of the retina (the fovea for visual acuity = 1) having the best cortical representation. Reductions of the saccadic pace are to be expected either for intrinsically slowed retinal performance or when the central information density is higher than can be cognitively processed during fixations of normal duration.

Literature: Krischer, C.C., Zihl, J. and Meissen, R., Reading: Influence of letter size, display quality and anticipation. In "Current oculomotor research: physiological and psychological aspects" Becker, W., Deubel, H. and Mergner T. (eds.), Plenum Press, New York (in press).

Taylor S.E., 1965, Eye movements in reading: facts and fallacies. Am. Ed. Res. J., 2: 187-202

Zangemeister, W.H., Sherman, K. and Stark, L., 1995, Evidence for a global scanpath strategy in viewing abstract compared with realistic images. Neuropsychologia, 33: 1009-1025