An investigation in to the effect of pre-lexical conceptual processing on naming latencies in the PWI paradigm

The ability to accurately and efficiently retrieve a word from the mental lexicon is a fundamental communication skill. Theoretical models of speech production which seek to explain this skill are informed by experimental naming paradigms such as the Picture Word Interference (PWI) paradigm, in which participants name pictures of common objects as quickly and as accurately as possible whilst ignoring a superimposed distractor word. In this paradigm, different semantic relations between picture-word pairs lead to differing effects — interference for categorically related PICTURE-word pairs (e.g., MOUSE-fox), and facilitation for associatively related picture-words pairs (e.g., MOUSE-cheese) and part-whole related picture-word pairs (e.g., MOUSE-tail), compared to unrelated picture-word pairs (e.g., MOUSE-violin). Rival theoretical models of speech production (e.g., Abdel Rahman & Melinger, 2009; Levelt, Roelofs, & Meyer, 1999; Mahon, Costa, Peterson, Vargas, & Caramazza, 2007) commonly assume pre-lexical conceptual mechanisms are responsible for semantic facilitation effects. Despite this common assumption, to date no PWI study has employed a manipulation of conceptual features. Thus, the aim of the current study is to investigate the effect of pre-lexical conceptual processing on naming latencies in the PWI paradigm by examining the effect of semantic feature richness of distractor words on naming unrelated pictures. Semantic features(characteristics that describe the meaning of a word) are assumed to be involved in initial computation of word meaning and it has been demonstrated that words with many semantic features are responded to more quickly than words with few semantic features in other speeded semantic tasks. A significant effect of semantic feature richness on naming latencies failed to emerge, however, un-expectedly a significant effect of number of features was observed on semantic naming errors. In addition, qualitative analysis of errors revealed that all errors were categorical or visual/structural in origin and most involved response set members