Holcombe Nguyen Goodbourn (in press). Implied reading direction and prioritization of letter encoding. Journal of Experimental Psychology: General

Abstract

Capacity limits hinder processing of multiple stimuli, contributing to poorer performance for identifying two briefly presented letters than for identifying a single letter. Higher accuracy is typically found for identifying the letter on the left, which has been attributed to a right-hemisphere dominance for selective attention. Here we use rapid serial visual presentation (RSVP) of letters in two locations at once. The letters to be identified are simultaneous and cued by rings. In a first experiment, we manipulated implied reading direction by rotating or mirror-reversing the letters to face to the left rather than to the right. The left-side performance advantage was eliminated. In a second experiment, letters were positioned above and below fixation, oriented such that they appeared to face downwards (90° clockwise rotation) or upwards (90° counter-clockwise rotation). Again consistent with an effect of implied reading direction, performance was better for the top position in the downwards condition, but not in the upwards condition. In both experiments, mixture modeling of participants’ report errors revealed that attentional sampling from the two locations was approximately simultaneous, ruling out the theory that the letter on one side was processed first, followed by a shift of attention to sample the other letter. Thus, the orientation of the letters apparently controls not when the letters are sampled from the scene, but rather the dynamics of a subsequent process, such as tokenization or memory consolidation. Implied reading direction appears to determine the letter prioritized at a high-level processing bottleneck.