The generation of antisaccades: development and dyslexia

B. Fischer

Brain Research Unit, Institute of Biophysics, University of Freiburg, Germany (

Saccade control has at least 3 components: fixation holds the eyes in place, reflexes take the eyes automatically from one stimulus to the next, and voluntary control allows one to look at things by conscious decisions. This contribution concentrates on 4 aspects of the voluntary component as tested by the original antisaccade task introduced by Hallett (1978): antisaccade performance of adults, development with age, antisaccades in dyslexia, effects of daily practice.

With a gap condition (fixation point switched off before stimulus onset) adult healthy subjects cannot perform this task without producing certain numbers of erratic prosaccades (about 15%). Valid spatial precues increase the error rate and half of the errors remain unconscious.

The variables describing antisaccade performance (error rate, reaction time, correction rate, and correction time) varied strongly with age until age 18 y and beyond age 40 - 50 y, while those describing the prosaccades were pretty much developed at age 8 - 9 y (300 subjects).

Saccade performance of 400 dyslexics was compared with that of 100 normally reading children (age 7 to 18 y, subdivided in 4 age groups). Group mean values of the error rate from antisaccade task were significantly different at all ages except for the youngest age group. Up to 50% of the dyslexic children depending on age produced significantly more errors in the antisaccade task than the controls. High error rates may occur because of weak fixation or because of deficits in the voluntary control. Reaction and correction times did not differentiate between groups.

We also tested 300 controls and 370 dyslexics by 3 dynamic orientation detection tasks: one required stationary fixation, the other prosaccades, and the third saccades against the direction of a distracter. Subjects between age 10 y and 40 y performed these tasks easily above 85 - 95% correct. The performance dropped to percent correct values around 65% at age 8 y or at ages above 60 y. On average dyslexic children performed below the level of the age matched controls in particular in the distracter task.

In order to improve saccade control of the dyslexics we gave them a selection of the 3 visual task for daily practice. After 3 to 6 weeks of practice the children were retested (N = 78): most of them had improved their saccadic behaviour. In particular the error rate decreased by 20% and the correction rate increased by 13% on average.

This study shows that the different components of saccade control develop differently with age, voluntary control being fully developed only around age 18 y. A large percentage of dyslexics exhibit developmental deficits of their voluntary not of their reflexive control component. These oculomotor and visual deficits may be considered as a consequence of a deficit in the magnocellular system serving the dorsal route. Daily practice helps the children to overcome the optomotor deficits and hopefully forms the basis for a better learning of reading and spelling skills in the future.