Beyond alphabetic orthographies: The role of form and phonology in transposition effects in Katakana

Abstract

In the past years, there has been growing interest in how the order of letters is attained in visual word recognition. Two critical issues are: (1) whether the front-end of the recently proposed models of letter position encoding can be generalised to non-alphabetic scripts, and (2) whether phonology plays an important role in the process of letter position encoding. In the present masked priming lexical decision experiments, we employed a syllabic/moraic script (Katakana), which allows disentangling form and phonology. In Experiment 1, we found a robust masked transposed-mora priming effect: the prime a.ri.me.ka facilitates the processing of the word a.me.ri.ka relative to a double-substitution prime (a.ka.ho.ka, ). In Experiment 2, we failed to obtain any signs of a masked phoneme transposition effect (a.re.mi.ka-a.me.ri.ka vs. a.ke.hi.ka-a.me.ri.ka). In Experiment 3, we failed to find any signs of a masked phonological priming effect when the order of the consonant/vowel phonemes of the internal morae was the right one (e.g., a.ma.ro.ka-a.me.ri.ka vs. a.ka.ho.ka-a.me.ri.ka). Thus, masked transposed-mora priming effects are orthographic (rather than phonological) in nature. We discuss how the recently proposed input coding schemes can be generalised to a syllable-based script.