Date created: | Last Updated:

Category: Project

Description: Abstract: A distinctive feature of tonal languages such as Mandarin Chinese is that the same consonant-vowel (CV) sequence is a different word depending on the tone it is spoken with. Ma with Tone 1 (level) means mother, but with Tone 2 (rising) means hemp. According to some linguistic theories, these tones “float above” the CV-structure of a word, raising the question of how tones are represented and produced. We measured how quickly speakers produced sequences of CV-tone syllables, independently manipulating the repetition pattern of the C, V, and tone elements. Surprisingly, the number of unique CV or tone units in the sequence did not predict speech rate, nor did their repetition pattern. Instead, post-hoc analyses revealed that speech rate was robustly faster when each CV was paired with only one tone (speakers produced both ba1 li2 bi2 la1 and ba1 ba1 ba1 ba1 about equally quickly), compared to when a particular CV needed to be produced with more than one tone (speakers produced ba1 ba2 ba1 ba2 slowly). We suggest that Mandarin speakers represent CVs as syllable “chunks,” programming each with tone upon phonetic encoding, so that producing the same chunk with more than one tone in a sequence is difficult.