Picture-word interference is a Stroop effect: A theoretical analysis and new empirical findings.

The picture-word interference (PWI) paradigm and the Stroop color-word interference task are often assumed to reflect the same underlying processes. On the basis of a PRP study, Dell'Acqua et al. (Psychonomic Bulletin & Review, 14: 717-722, 2007) argued that this assumption is incorrect. In this article, we first discuss the definitions of Stroop- and picture-word interference. Next, we argue that both effects consist of at least four components that correspond to four characteristics of the distractor word: (1) response-set membership, (2) task relevance, (3) semantic relatedness, and (4) lexicality. On the basis of this theoretical analysis, we conclude that the typical Stroop effect and the typical PWI effect mainly differ in the relative contributions of these four components. Finally, the results of an interference task are reported in which only the nature of the target - color or picture - was manipulated and all other distractor task characteristics were kept constant. The results showed no difference between color and picture targets with respect to all behavioral measures examined. We conclude that the assumption that the same processes underlie verbal interference in color and picture naming is warranted.

In defense of the lexical-competition account of picture-word interference: a comment on Finkbeiner and Caramazza (2006).

On the basis of two empirical observations, Finkbeiner and Caramazza (2006) take issue with the generally accepted interpretation of semantic interference in the picture-word interference task in terms of lexical competition. As an alternative, they propose a response-selection account, in which semantic interference is attributed to the time needed to remove the inappropriate (word-reading) response from an output buffer. In this comment we argue that the empirical work discussed provides an interesting challenge for current models of language production, but that the authors' alternative account is at variance with at least three robust empirical findings in the language production literature.

Response selection in visual search: the influence of response compatibility of nontargets.

The authors used visual search tasks in which components of the classic flanker task (B. A. Eriksen & C. W. Eriksen, 1974) were introduced. In several experiments the authors obtained evidence of parallel search for a target among distractor elements. Therefore, 2-stage models of visual search predict no effect of the identity of those distractors. However, clear compatibility effects of the distractors were obtained: Responses were faster when the distractors were compatible with the response than when they were incompatible. These results show that even in parallel search tasks identity information is extracted from the distractors. In addition, alternative interpretations of the results in terms of the occasional identification of a distractor before or after the target was identified could be ruled out. The results showed that flat search slopes obtained in visual search experiments provide no benchmark for preattentive processing.