Howson Phil J, Redford Melissa A
Department of Linguistics, University of Oregon, Eugene.
J Speech Lang Hear Res. 2021 Mar 17;64(3):734-753. doi: 10.1044/2020_JSLHR-20-00391. Epub 2021 Mar 1.
Purpose Liquids are among the last sounds to be acquired by English-speaking children. The current study considers their acquisition from an articulatory timing perspective by investigating anticipatory posturing for /l/ versus /ɹ/ in child and adult speech. Method In Experiment 1, twelve 5-year-old, twelve 8-year-old, and 11 college-aged speakers produced carrier phrases with penultimate stress on monosyllabic words that had /l/, /ɹ/, or /d/ (control) as singleton onsets and /æ/ or /u/ as the vowel. Short-domain anticipatory effects were acoustically investigated based on schwa formant values extracted from the preceding determiner (= ) and dynamic formant values across the /ə#LV/ sequence. In Experiment 2, long-domain effects were perceptually indexed using a previously validated forward-gated audiovisual speech prediction task. Results Experiment 1 results indicated that all speakers distinguished /l/ from /ɹ/ along F3. Adults distinguished /l/ from /ɹ/ with a lower F2. Older children produced subtler versions of the adult pattern; their anticipatory posturing was also more influenced by the following vowel. Younger children did not distinguish /l/ from /ɹ/ along F2, but both liquids were distinguished from /d/ in the domains investigated. Experiment 2 results indicated that /ɹ/ was identified earlier than /l/ in gated adult speech; both liquids were identified equally early in 5-year-olds' speech. Conclusions The results are interpreted to suggest a pattern of early tongue-body retraction for liquids in /ə#LV/ sequences in children's speech. More generally, it is suggested that children must learn to inhibit the influence of vowels on liquid articulation to achieve an adultlike contrast between /l/ and /ɹ/ in running speech.
目的 对于说英语的儿童而言,流音是他们最后习得的语音之一。本研究从发音时长的角度考察流音的习得情况,通过调查儿童和成人言语中/l/与/ɹ/的预期姿势。方法 在实验1中,12名5岁儿童、12名8岁儿童和11名大学生说出包含单音节词的载体短语,这些单音节词以/l/、/ɹ/或/d/(对照)作为词首单音,以/æ/或/u/作为元音,且倒数第二个音节重读。基于从前置限定词中提取的央元音共振峰数值以及/ə#LV/序列中的动态共振峰数值,从声学角度研究短域预期效应。在实验2中,使用先前验证过的前向门控视听语音预测任务从感知角度对标长域效应进行索引。结果 实验1的结果表明,所有发音者都能根据第三共振峰区分/l/和/ɹ/。成人能以较低的第二共振峰区分/l/和/ɹ/。年龄较大的儿童呈现出更接近成人模式的细微差异;他们的预期姿势也更容易受到后续元音的影响。年龄较小的儿童无法根据第二共振峰区分/l/和/ɹ/,但在所研究的音域中,这两个流音都能与/d/区分开来。实验2的结果表明,在门控成人言语中,/ɹ/比/l/更早被识别;在5岁儿童的言语中,这两个流音被识别的时间相同。结论 研究结果表明,儿童言语中/ə#LV/序列的流音存在早期舌体后缩的模式。更普遍地说,这表明儿童必须学会抑制元音对流音发音的影响,以便在连贯言语中实现类似于成人的/l/和/ɹ/对比。