Attentional constraints on semantic activation: Evidence from Stroop's paradigm.

It is widely believed that semantic activation from print is automatic in the sense that it is capacity free. Two experiments addressed this issue in the context of the Psychological Refractory Period (PRP) paradigm. Participants identified whether a tone was high or low in pitch in Task 1, and named the color carried by an irrelevant word in Task 2. Tasks 1 and 2 were separated by a short or long SOA. In Experiment 1 incongruent color words and neutral words served as irrelevant distractors, whereas in Experiment 2 the distractors consisted of incongruent color associates (e.g., tomato) and the same set of neutral items. Additionally, the proportion of short and long SOAs between Task 1 and Task 2 varied across blocks, within subjects (e.g., 80:20), so as to determine whether the bottlenecking of semantic activation and response competition reported previously is best construed as structural, or subject to performance optimization. Replicating Miller, Ulrich, and Rolke (2009), SOA Proportion interacted with SOA in both experiments, consistent with performance optimization. In contrast, replicating Besner and Reynolds (2014), SOA and Congruency had additive effects on RT in both experiments, consistent with an account in which both response competition and semantic activation are bottlenecked. The best account to date is that (i) semantic processing and response competition are structurally bottlenecked (require some form of capacity), whereas (ii) other anonymous processes are subject to performance optimization.