To what extent can visual deficits caused by multiple sclerosis be understood in terms of parallel processing?