Experiment paradigm

Two different versions of the tone production paradigm were performed. Two subjects (S1 and S2) participated in the first paradigm, while the other subjects participated in the second paradigm. For each trial in the first version, a single word was presented on a screen 20 cm from the participant. The subject was instructed to name the word three times in a row after a go cue. In each block, a total of 60 different words were used (four lexical tones of /e/, /ye/, /yue/, /wo/, /ei/, /wei/, /ou/, /you/, /a/, /ya/, /wa/, /ai/, /wai/, /ao/, /yao/), and each word was repeated 3 times, which yielded 180 tokens in total (45 repetitions per lexical tone). Subject S1 completed one block (45 repetitions per lexical tone), and Subject S2 completed two blocks (90 repetitions per lexical tone). For each trial in the second version, a single word was played through the speaker of a laptop, and the participant was instructed to repeat the exact word three times in a row. In each block, a total of 8 different words were used (four lexical tones of /ma/ and /mi/, i.e., /mā/, /má/, /mǎ/, /mà/, /mī/, /mí/, /mǐ/, /mì/), and each word was repeated 15 times, which yielded 120 tokens in total (30 repetitions per lexical tone). Each patient completed four blocks (120 repetitions per lexical tone).

In the sentence production task, we selected 20 sentences from the Fu-sentence corpus¹⁶, which consists of phonetically balanced 7-character Mandarin sentences (e.g., “北京近来很寒冷“, /běi jīng jìn lái hěn hán lěng/). For each trial in the task, a single sentence was presented on a screen positioned 20 cm away from the participant. The participant was instructed to read the entire sentence aloud once. Within each block, each sentence was repeated twice. Each participant completed 2-5 blocks, resulting in a total of 80-200 sentences (560-1400 tones) in total.