Roelofs, Piai, and Schriefers (Language and Cognitive Processes) test both the WEAVER++ model of word production and the response-exclusion account of performance in Stroop-like tasks against data from the word-word interference (WW1) task, and conclude that whereas the WEAVER++ successfully accounts for those data, the response-exclusion hypothesis fails. Here we show that once recent data from the WW1 task are considered, both models fail.Keywords: word production; word-word interference task; lexical access; reading aloud In the word-word interference (WWI) task, participants are instructed to read aloud one of two simultaneously presented printed words while ignoring the other. Roelofs, Piai, and Schriefers (2013) use data from this task to test two models: the WEAVER++ (Levelt, Roelofs, & Meyer, 1999), a model of word production, and the responseexclusion hypothesis (REH, Finkbeiner & Caramazza, 2006a), a model of performance in Stroop-like tasks (both described below). Roelofs et al. (2013) conclude that whereas the WEAVER++ accounts for the performance in the WWI task, the REH fails. Here we will demonstrate that, instead, both the WEAVER++ and the REH fail to account for the results typically obtained in WWI tasks. Then, we will list a few basic theoretical issues any model of performance in WWI must consider. Roelofs et al. (2013) list three findings obtained with the WWI task: (1) it takes longer to read aloud a word when it is accompanied by another (unrelated) word than when it is accompanied by a neutral orthographic string (e.g., a row of Xs), thus a distractor word has a "general interference effect" on the reading of the target word; (2) it takes less time to read aloud a word when it is accompanied by an identical word than when it is accompanied by a neutral orthographic string,
Semantic effect in WWI