University of Illinois at Urbana-Champaign, USA.
J Speech Lang Hear Res. 2011 Apr;54(2):549-66. doi: 10.1044/1092-4388(2010/09-0216). Epub 2010 Aug 18.
Theories of morphosyntactic development must account for between-child differences in morphosyntactic growth rates. This study extends Legate and Yang's (2007) theoretically motivated cross-linguistic approach to determine if variation in properties of parent input accounts for differences in the growth of tense productivity.
Fifteen toddlers (and parents) participated. None were producing tense morphemes productively at 21 months. Two dependent measures of morphosyntactic growth between 21 and 30 months were used: empirical Bayes linear coefficients at 21 months and predicted productivity scores at 30 months. Predictor variables included child sex, vocabulary, and mean length of utterance as well as 4 measures of parent language input at 21 months.
Input informativeness for tense was the most consistent predictor of morphosyntactic growth, explaining 28.3% of the unique variance in children's linear growth coefficients at 21 months and 23.0% of the unique variance in predicted tense productivity scores at 30 months. General input measures were unrelated. Child sex explained an additional 24.7% of the variance in early linear growth. Child vocabulary at 21 months did not explain a significant proportion of unique variance.
The findings provide evidence that input informativeness, an abstract and distributed property of input, contributes to morphosyntactic growth.
形态句法发展理论必须解释儿童之间形态句法增长率的差异。本研究扩展了 Legate 和 Yang（2007）的理论动机跨语言方法，以确定父母输入的属性变化是否解释了时态生产增长的差异。
15 名幼儿（及其父母）参与了研究。在 21 个月时，没有幼儿能生成时态词。在 21 至 30 个月之间使用了两个形态句法增长的依赖度量：21 个月时的经验贝叶斯线性系数和 30 个月时的预测生产力得分。预测变量包括儿童的性别、词汇量和平均话语长度，以及 21 个月时的 4 种父母语言输入测量。
时态的输入信息量是形态句法增长最一致的预测因素，解释了儿童 21 个月时线性增长率的 28.3%的独特方差，以及 30 个月时预测时态生产力得分的 23.0%的独特方差。一般输入测量没有关联。儿童性别解释了早期线性增长的另外 24.7%的方差。21 个月时的儿童词汇量没有解释独特方差的显著比例。
这些发现提供了证据表明，输入信息量作为输入的抽象和分布式特征，有助于形态句法的增长。
