Sharing the beginning versus sharing the end: Spoken word recognition in the visual world paradigm in Japanese
Hideko Teruya, Vsevolod M. Kapatsinski · The Journal of the Acoustical Society of America · 2011
Several studies have suggested that in English spoken word recognition, words sharing beginnings compete more than words sharing ends (i.e., the cohort competitor has an advantage relative to the rime competitor). We investigated the relative activation differences between the cohort competitor and the rime competitor in Japanese CVCV words using the visual-world paradigm (Allopenna et al., 1998). Results were analyzed using linear mixed effects models (Magnuson et al., 2008). Participants fixated referents of rime competitors more than those of cohort competitors when the cohort competitor shared only the first consonant with the target while the rime shared the rest of the word (VCV). When both competitors shared a syllable with the target, the rime competitor was fixated as much as the cohort competitor and the recognition of the target was delayed. The results suggest a syllabic or moraic basis to Japanese spoken word recognition, where lexical activation requires at least one mora of the word to match the signal. The weakness of cohort effects in Japanese challenge fully incremental models of word recognition. Whether the results would also hold for English is unclear as previous visual world spoken word recognition studies in English did not systematically manipulate degree of overlap.