Posner's Endogenous Beam Is (Still) Not Treisman's Glue.

Abstract

Posner's beam and Treisman's glue are metaphors of visual attention that stimulated research programs on exogenous and endogenous modes of attentional control and feature integration theory. Briand and Klein (1987) asked, "Is Posner's beam the same as Treisman's glue," positing that the orienting of Posner's spatially confined beam (spotlight of attention) could be the mobilization of the same attentional machinery described by Treisman as performing object feature integration. Based on the patterns of interaction between cue condition and the opportunities for illusory conjunctions, they concluded the answer depended upon the mode of control: An interaction suggested a yes answer for exogenous control while additivity suggested no for endogenous control, a difference in the effects of attention suggesting that there may be two independent beams. Kawahara and Miyatani (2001) challenged the lack of interaction between endogenous cues and task type (feature targets vs. conjunction targets) using a different paradigm that emphasized search and contained more items. After noting the importance of presenting all the displayed items at an attended or unattended location, we report two experiments that replicate Briand with two-item arrays and extend this finding to four-item arrays, strongly supporting the claim that Posner's endogenous beam is not Treisman's glue.