How do we code the letters of a word when we have to write it? Investigating double letter representation in French

Abstract : How do we code the letters of a word when we have to write it? We examined whether the orthographic representations that the writing system activates have a specific coding for letters when these are doubled in a word. French participants wrote words on a digitizer. The word pairs shared the initial letters and differed on the presence of a double letter (e.g., LISSER/LISTER). The results on latencies, letter and inter-letter interval durations revealed that L and I are slower to write when followed by a doublet (SS) than when not (ST). Doublet processing constitutes a supplementary cognitive load that delays word production. This suggests that word representations code letter identity and quantity separately. The data also revealed that the central processes that are involved in spelling representation cascade into the peripheral processes that regulate movement execution.