Markedness and Learnability in Second Language Phonology

描述（由申请人提供）：本研究的目的是表明支配主要语言运作的普遍概括也适用于非主要语言系统，包括非母语（L2）。假设这些共性实际上是对易学性的制约，因此可以作为语言教学的策略。这项研究的重点是第二语言音系，这是第二语言习得中负责非母语语音模式的部分。所提出的研究中使用的普遍性是关于音位对比在语言中可以保持的领域的隐含概括。这一概括表明，如果一种语言在词形复合的词中有音位对比，那么在词形简单的词中也必然有这种对比，但反之则不然。这项研究提出的主张是，这种概括也适用于第二语言，它对人类可以开发的语言系统的类型施加了限制，并且可以用作语言教学的策略。本研究采用的策略是只指导学习者在两个语音领域中的一个进行音位对比，一个是泛化所允许的，另一个是泛化所不允许的。预测是，那些只在被概括排除的语音环境中学习对比的受试者会将学习推广到包括其他领域，从而学习一个被概括认可的系统，但那些只接受被普遍允许的对比训练的学习者不一定会将学习推广到其他环境。计划进行两项研究，涉及来自四种母语背景的120名受试者。本研究针对六个不同的音位对比提出了三个假设，并分析了一个问题中受试者对对比的产生和感知。生产研究将是对普遍概括指定的两个领域的对比的音系分析；知觉研究将探讨当被试感知对比时，内隐概化是否也成立。这项研究的健康相关性在于，我们当前的多元化社会包括许多不同种类的美国英语，包括那些被称为“口音”的英语，这些英语不是母语的人产生的。这些非母语形式的英语是以下因素的产物：1)个人母语的规则，2)英语的模式，以及3)人类学习的原则。这项研究将使我们能够检查第二语言学习者作为显性英语训练的产物的正常语言发展，特别是因为它涉及到所有人在所有语言学习中可能受到的普遍限制。这项研究将有助于更好地理解美国双语和有限英语人群正常第二语言发展的机制。

DESCRIPTION (provided by applicant): The aim of this research is to show that universal generalizations which govern the operation of primary languages hold also for non-primary language systems, including non-native languages (L2). It is hypothesized that these universals are in fact constraints on learnability, and therefore can be used as strategies for language teaching. The research focuses on second language phonology, the part of L2 acquisition responsible for the patterning of non-native speech sounds. The universal used in the proposed research is an implicational generalization about the domain in which a phonemic contrast can hold in a language. This generalization states that if a language has a phonemic contrast in morphologically composite words, it also necessarily has this contrast in morphologically simple words, but not vice versa. The claims underlying the proposed research are that this generalization holds also for L2, that it imposes constraints on the kind of language systems that humans can develop, and that it can be used as a strategy for language teaching. The strategy employed in the proposed research is to instruct learners on certain phonemic contrasts in only one of two phonological domains, one which is permitted by the generalization, and the other which is excluded by it. The prediction is that those subjects who are taught the contrast only in the phonological environment that is excluded by the generalization will generalize the learning to include the other domain, thereby learning a system that is sanctioned by the generalization, but that those learners who are trained only on the contrast that is allowed by the universal will not necessarily generalize this learning to the other environment. Two studies involving 120 subjects from four native-language backgrounds are planned. The studies address three hypotheses focusing on six different phonemic contrasts, and entail analyses of the subjects' production and perception of the contrasts in1 question. The production study will be a phonological analysis of the contrasts in both domains specified by the universal generalization; the perception study will investigate whether the implicational generalization holds also when the subjects perceive the contrasts. The health relatedness of this research is that our current pluralistic society includes many different varieties of American English, including those termed "accented" English produced by people for whom English is not a native language. These non-native forms of English are a product of 1) the rules of an individual's native language, 2) the patterns of English, and 3) principles governing how humans learn. This study will allow us to examine normal language development in second-language learners as a product of explicit training in English, especially as it relates to possible universal constraints that all humans might be subject to in all language learning. This research will lead to a better understanding of the mechanisms of normal L2 development in bilingual and limited-English populations in the United States.