Reduction, the shortening, warping, or deletion of speech sounds, is a natural process in speech communication, especially in casual or fast speech. While reduced words are easier to produce, they require more perceptual effort. Also, words with consistently spelled sounds are understood faster and more accurately. Taking reduction and consistency together, we hypothesize that reduction is more likely in words with consistent written forms. Unlike English, Japanese implements logographic orthography, in which written symbols represent an entire word or concept, rather than individual sounds. The present study analyzes production data from a delayed repetition task previously collected by Mukai et al. (2023) and examines the relationship between orthographic consistency and reduction among Japanese speakers. We also compare participants’ productions to the speech stimuli. Our results indicate that speakers align reduction characteristics in the target speech: the duration of reduced targets are shorter than the unreduced counterparts. We also find a consistency effect in which speakers decrease the duration of unreduced targets as the consistency increases. The results are discussed regarding their implications for how listeners align their production when listening to reduced speech and the role of the orthographic form in speech processing.
Mukai et al. (Wed,) studied this question.