Visual constraints in reading: evidence from non-ocular behaviour

On-line evidence of visual constraints in reading is reported for measures other than eye behaviour. Text was read from a CRT using a moving window between 6 and 19 characters wide, with letters outside the window replaced by blanks, dashes, or visually similar letters. Reading speed was self-adjusted; subjects performed a subsidiary task while reading aloud. Window width and the visual context beyond the window affected dual task performance, and together with fount size, also reading accuracy. The form of text background further affected two measures of eye-voice span. These findings complement ones for ocular behaviour obtained using a CRT window coordinated with a reader's eye fixations.