Do words compete as we speak? A systematic review of picture-word interference (PWI) studies investigating the nature of lexical selection

This review synthesizes findings from 117 studies that have manipulated various picture-word interference (PWI) task properties to establish whether semantic context effects reflect competitive word retrieval, or are driven by noncompetitive processes. Manipulations of several PWI task parameters (e.g., distractor visibility) have produced contradictory findings. Evidence derived from other manipulations (e.g., visual similarity between targets and distractors) has been scarce. Some of the manipulations that have furnished reliable effects (e.g., distractor taboo interference) do not discriminate between the rival theories. Interference from nonverbal distractors has been shown to be a genuine effect dependent on adequate lexicalization of interfering stimuli. This supports the swinging lexical network hypothesis and the selection-by-competition-with-competition-threshold hypothesis while undermining one of the assumptions of the response exclusion hypothesis. The contribution of pre-lexical processes, such as an interaction between distractor processing and conceptual encoding of the target to the overall semantic context effect is far from settled.