An intersection of three literatures (skilled word recognition, spatial attention, and the Stroop effect) is addressed in a series of four experiments, The results, in conjunction with other observations, are taken to suggest that (1) the default value for spatial attention in visual word recognition is distributed across the word, (2) preening a single letter position serves to narrow the focus of spatial attention, and (3) this reduces or prevents activation in the word recognition system, Consequently, the Stroop effect is reduced in magnitude or eliminated, depending on details ofthe context. Contrary to the widespread view that it reflects automatic processing, the Stroop effect is better conceptualized as reflecting the action of default settings in the word recognition and attentional systems. Some relations between consciousness, context, and control are noted.It is clear that mental skill acquired through high levels of practice confers benefits on the performer. For example, skilled performance in a domain such as word recognition is much faster and less error prone than unskilled performance. A more contentious claim is that this speed-up carries a cost; skilled readers are said to be unable to prevent lexical and semantic analyses of words, provided that the stimulus is clear enough, bright enough, large enough, and presented for a sufficiently long duration for the reader to process it. The most famous example, provided in many papers, chapters, and textbooks, in support ofthis assertion is the Stroop effect (Stroop, 1935; see also MacLeod's, 1991, review). When skilled readers are asked to identify the print color of a word and explicitly instructed not to read the word, they are nonetheless typically slower and more error prone when the word references an incongruent color (e.g., the word red printed in green), as compared with the time taken to identify the color when the word references a congruent color (e.g., the word red printed in red) or a noncolor term (e.g., the word pen printed in green). Word recognition, thus, is widely viewed as automatic, in the sense that it occurs without intent and cannot be prevented (see, e.g., Anderson, 1995;Ashcraft, 1994;Crowder & Wagner, 1992;Posner & Snyder, 1975;Rayner & Pollatsek, 1989;Reisberg, 1997).One difficulty with this account of mental skill as being exemplified by the Stroop effect is that it is unwarranted. It is one thing to show that skilled readers sometimes process the word despite being instructed not to but quite another to conclude that word recognition is therefore automatic, in the sense that lexical and semantic analyThis research was supported by Grants A0998. OGP0183905. and EQPO187220 from the Natural Sciences and Engineering Council of Canada. We thank R. S. McCann for comments on a draft of the manuscript and K. Rayner. B. A. Levy. and R. Klein for their reviews. Address correspondence to D. Besner, Department of Psychology. University of Waterloo. Waterloo, ON N2L 3G 1, Canada (e-mail: dbesner@ watarts. uwaterloo .ca).ses of a wo...