Attention demands during reading and the occurrence of brief (express) fixations

Abstract

Eye movements were recorded while subjects read passages of text repeatedly (Experiment 1) and while normal text and strings of homogeneous letters were fixated (Experiment 2). Text repetition decreased fixation durations and increased saccade size, presumably because it decreased attention demands. Irrespective of repetition, however, no distinct distribution of brief (express) fixations emerged. In Experiment 2, fixation durations were shorter and saccades were larger when strings of homogeneous letters were "read," indicating that this condition decreased attention demands. Again, however, no distinct distribution of express fixations emerged. These findings pose problems for the view that attentional processes determine the occurrence of brief (express) fixation durations in reading. Supplementary analyses of Experiments 1 and 2 suggested that visuospatial processing affected fixation durations, irrespective of linguistic processing demands.In overlap and gap tasks, subjects fixate a centrallocation and move the eyes laterally in response to the presentation of a peripheral target. In a typical overlap trial, subjects fixate (attend) a centrally presented visual stimulus while a to-be-fixated peripheral target appears, so that fixated and peripherally available stimuli are concurrently available. In a typical gap trial, the fixated (attended) stimulus is erased prior to target presentation, creating a "gap" between the availability of fixated and to-be-fixated stimuli.Examination of fixation duration distributions on gap trials often reveals a distinct distribution of (express) saccades with latencies between 80 and 130 msec and a mode of approximately 120 msec, at least in some subjects, and additional modes for longer duration latencies. No distinct express saccade peak is obtained in overlap trials (Boch & Fischer, 1986;Fischer, 1986Fischer, , 1987Fischer, , 1989Fischer, , 1993Fischer & Weber, 1988;Kalesnykas & Hallett, 1987;.It is unlikely that express saccades are the result of either more effective visual target processing on gap trials or a higher visual distinctiveness of target stimuli. Gap effects combine additively with visual target characteristics and express saccades occurThe research reported in this article was supported by National Science Foundation Grant BNS-9010067 and by the Center for Cognitive and Psycholinguistic Studies at SUNY-Binghamton, Francoise Vitu was supported by a grant from the French government's Ministere des Affaires Etrangers while she had a postdoctoral position at SUNYBinghamton. Requests for reprints should be sent to A. Inhoff, Department of Psychology, SUNY-Center, Binghamton, NY 13902-6000. on gap trials, even when targets are presented in a nonvisual modality . It is also unlikely that all express saccades are anticipatory movements, as they are present on gap trials even when anticipatory saccades are removed (Kalesnykas & Hallett, 1987). However, anticipations may inflate the frequency of express saccades. Jutner and Wolf (1992) showed that the proportion of e...

Vergence eye movements during fixations in reading

“…That is, mean fixation durations were longer in the word-list than in the prose condition of the Text Type experiment and they were longer in the subvocaliz- ing than in the reading-for-meaning conditions of the Reading Objective and Repetition experiments. These means differ in the same direction as the silent and reading measures reported elsewhere (e.g, Bouma and De Voogd, 1974;Gray, 1969) contrast with previous findings (Hyona and Niemi, 1990;Inhoff et al, 1993), fixa durations were similar for first and second time reading). In order to address the concern that the vergence velocity differences were du differences in fixation duration, vergence velocity was recomputed by taking the in 80 ms of each fixation only (analogous to the prior calculations of vergence velocity, first 10 ms after the end of the saccade were excluded).…”

Eye movements in reading and information processing: 20 years of research.

“…B. Sereno, 1992). Even if there were no controversy surrounding express saccades per se, there are questions about the functional utility of such shortlatency saccades for normal information processing tasks (M. H. Fischer & Rayner, 1993) : Inhoff, Topolski, Vitu, and O'Regan (1993) found no evidence for express saccades or a bimodal distribution of fixation durations during reading.…”

“…However, examination of data revealed such an effect. Carroll & Slowiaczek, 1986;Morris, 1994; S. C. Sereno, 1995; S. C. HyOn~i & Niemi, 1990;Inhoff et al, 1993;Rayner, Raney, & Pollatsek, 1995Beauvillain, 1996Holmes & O'Regan, 1987Hytnlt & Pollatsek, in press;Inhoff, Briihl, & Schwartz, 1996;Lima, 1987Albrecht & Clifton, 1998Blanchard, 1987;Duffy & Rayner, 1990;K. Ehrlich, 1983;K.…”

Against a role for attentional disengagement in the gap effect: A friendly amendment to Tam and Stelmach (1993)

“…Finally, it is possible that subjects were sometimes fixating what was supposed to have been an attended eccentric stimulus. These weaknesses, together with Klein's (1990, 1993a) repeated finding that the gap effect is uninfluenced by the direction of attention, suggest no role for attentional disengagement in the gap effect (see Inhoff, Topolski, Vitu, & O'Regan, 1993, for converging evidence).…”