本研究的主要目的是在強調社會互動中大人所提供的各種「鷹架」(scaffolding) 在語言發展中的角色與價值，並試圖用鷹架的概念來解釋小孩為何能在短短幾年中學會母語 ，又為什麼在教室裡學習第二語言的效果則差得很多。首先我透過文獻探討，簡略討論了三 種語言發展的觀點，包括行為主義學派的學習理論、語言天賦論及社會互動論的觀點。本文 指出，語言發展大體是在有「鷹架」支撐的社會互動過程中完成的，此一發展與學習的過程 可用 Vygotsky 「最佳發展區」 (zone of proximal development) 的概念來理解 (Vygotsky, 1978)。本文進一步描述、分析了一個母語學習情境 (為一個父親與一歲多的女 兒的互動組 )，以及另一個以英語為第二語的教學情境。本文主要的研究發現有二：第一， 在母語學習情境裡，大人提供了語言鷹架、概念鷹架、與結構鷹架，這些鷹架能幫助幼兒瞭 解正在進行的活動，以及活動中所包含的語言與概念。這些鷹架通常是結合在一起的，其中 結構鷹架可說最為普遍。有了這些鷹架，幼兒就比較能瞭解語言所傳達的意義，進而「破解 」聲音的符號系統。第二，在第二語的教學中，老師也在語言形式及概念難度方面，做了一 些調整。但是由於師生的比例差距很大，語言鷹架及概念鷹架比較難落實到個別學習者的身 上。整體來說，不管是互動的量或質，教室裡的第二語教學所提供的學習鷹架比母語學習的 情境來得不理想。

　　　　　The main purpose of this study was to examine what role scaffolding plays in children's language development. This paper starts with a discussion of three major theoretical perspectives on language acquisition. In the discussion, I point out that the behaviorist's attempt to explain language acquisition through a general theory of learning, i.e., through a mechanism of stimulus, response, and reinforcement, seems to overestimate the shaping power of the environment while underestimating the child's potential to actively construct hypotheses about linguistic rules. Then I point out that the innatist's LAD hypothesis, though supported by evidence from recent neurolinguistic research, falls short of giving a detailed account of how a language is acquired and why sometimes acquired in different ways across learners and cultures. I take the interactionist view, which sees language acquisition as a process in which a child fulfills his/her language potential through interaction with the society, to be the most competent theory. Undergirded by the interactionist belief that scaffolded social interaction is crucial for language development, I present findings from my observation of two language learning situations, the first on first language socialization and the second on learning English as a second language in a classroom setting. First, linguistic scaffolding, conceptual scaffolding, and structural scaffolding were found to be important processes that provide first language learners with strong support in making sense of the on-going interaction and in learning new linguistic forms and concepts. Second, in the English as Second Language classroom, the teacher also made necessary modification in language input and conceptual level. However, due to the disparate teacher-student ratio, linguistic and conceptual scaffoldings became very difficult processes when considering the needs and levels of each individual student. This is especially true when there was no or little structural scaffolding to help students understand the on-going activity. At the individual level, the quality of language socialization process in the ESL classroom was thus found to be inferior to that found in the first language learning situation.