Abstract：
Children achieve better long-term language outcomes than adults. However, it remains unclear whether children actually learn language more quickly than adults during real-time exposure to input-indicative of true superior language learning abilities-or whether this advantage stems from other factors. To examine this issue, we compared the rate at which children (8-10 years) and adults extracted a novel, hidden linguistic rule, in which novel articles probabilistically predicted the animacy of associated nouns (e.g., \"gi lion\"). Participants categorized these two-word phrases according to a second, explicitly instructed rule over two sessions, separated by an overnight delay. Both children and adults successfully learned the hidden animacy rule through mere exposure to the phrases, showing slower response times and decreased accuracy to occasional phrases that violated the rule. Critically, sensitivity to the hidden rule emerged much more quickly in children than adults; children showed a processing cost for violation trials from very early on in learning, whereas adults did not show reliable sensitivity to the rule until the second session. Children also showed superior generalization of the hidden animacy rule when asked to classify nonword trials (e.g., \"gi badupi\") according to the hidden animacy rule. Children and adults showed similar retention of the hidden rule over the delay period. These results provide insight into the nature of the critical period for language, suggesting that children have a true advantage over adults in the rate of implicit language learning. Relative to adults, children more rapidly extract hidden linguistic structures during real-time language exposure. RESEARCH HIGHLIGHTS: Children and adults both succeeded in implicitly learning a novel, uninstructed linguistic rule, based solely on exposure to input. Children learned the novel linguistic rules much more quickly than adults. Children showed better generalization performance than adults when asked to apply the novel rule to nonsense words without semantic content. Results provide insight into the nature of critical period effects in language, indicating that children have an advantage over adults in real-time language learning.
摘要：
儿童比成人获得更好的长期语言结果。然而,目前尚不清楚儿童在实时接触输入时是否比成人更快地学习语言-这表明了真正的高级语言学习能力-或者这种优势是否源于其他因素。为了研究这个问题，我们比较了儿童（8-10岁）和成人提取小说的比率，隐藏的语言规则，其中新颖的文章概率地预测了相关名词的敌意(例如，\“gilion\”）。参与者根据一秒钟对这些两个单词的短语进行了分类，明确指示了两届会议的规则，被一整夜的延迟分开。儿童和成人都通过仅仅接触这些短语就成功地学会了隐藏的敌意规则，显示响应时间较慢，对违反规则的偶尔短语的准确性下降。严重的,儿童对潜规则的敏感性比成人更快地出现；儿童从学习的早期就表现出违规试验的处理成本，而成年人直到第二次会议才对该规则表现出可靠的敏感性。当被要求对非词试验进行分类时，儿童还表现出对隐藏动物规则的优越概括(例如，\"gibadupi\")根据隐藏的动画规则。在延迟期内，儿童和成人对潜规则的保留相似。这些结果提供了对语言关键时期性质的洞察，这表明儿童在内隐语言学习的速度上比成人有真正的优势。相对于成年人,儿童在实时语言暴露过程中更快地提取隐藏的语言结构。研究重点：儿童和成人都成功地隐含地学习了一部小说，未经指导的语言规则，完全基于对输入的暴露。孩子们学习新的语言规则比成年人快得多。当被要求将小说规则应用于没有语义内容的无意义词时，儿童表现出比成年人更好的泛化表现。结果提供了对语言中关键时期效应性质的洞察，这表明儿童在实时语言学习方面比成人更有优势。
