Some Remarks on the Magnocellular Deficit Theory of Dyslexia

The data bearing on the magnocellular or transient system deficit theory of dyslexia (Lovegrove et al., 1986, 1990; Livingstone et al., 1991) are highly conflicting. While some studies have produced results compatible with this theory (e.g. Lovegrove et al., 1982; Livingstone et al., 1991) other studies have yielded seemingly incompatible results (e.g. Smith et al., 1986; Victor et aZ., 1993; Gross-Glenn et al., 1995). It is therefore commendable that researchers have started to investigate the possible causes for these conflicting results. Borsting et al. (1996) have produced evidence that magnocellular deficits may be confined to only one subgroup of dyslexic individuals. Cornelissen et al. (1995) have presented evidence to indicate that the contrast sensitivity 10SS predicted by the magnocellular/transient system deficit theory can be observed in dyslexic individuals at low but not at high luminance levels. Although the empirical findings of Cornelissen et al. (1995) and Borsting et al. (1996) could account for some of the discrepancies in the data, the magnocellular/transient system deficit theory faces substantial problems which have yet to be resolved. There are two specific issues which seem particularly important for future research to address. The first issue pertains to the many empirical reports of contrast sensitivity loss in dyslexia at high spatialfrequencies. The magnocellular/transient system deficit theory of dyslexia predicts contrast sensitivity deficits associated with dyslexia at low spatial frequencies. The findings of Borsting et al. (1996) that magnocellular deficits may be confined to only one subgroup of dyslexic readers may account for why some studies have found low frequency contrast sensitivity deficits while others have not found such deficits. However, the observations of Borsting et al. (1996) do not seem to be able to account for the many instances of dyslexic readers having reduced sensitivity only to high spatial-frequencies (Hill & Lovegrove, 1993; Gross-Glenn et al., 1995), for general reductions in sensitivity which are most pronounced at high spatial-frequencies (Martin & Lovegrove, 1987, 1988), or for deficits which are strongest at medium frequencies (i.e. at 4 c/deg) and less severe at