This study explores the underlying cognitive structure of a small number of bilinguals, seeking to determine whether or not it is cognitively possible to develop symmetrically conceptual mediation between the two languages of a bilingual. Previous research has consistently found asymmetries in response and priming times in experiments with bilinguals. The pattern of these asymmetries and the conditions under which they were obtained, have motivated the development of the current model of bilingual representation and processing, the Revised Hierarchical Model (RHM) of Kroll and Stewart (1994). Level of proficiency has been an important factor in explaining the asymmetric connections posited by this model, arguing that the more proficient bilinguals become, the more heavily they rely on conceptual mediation between the two languages. This account implies that a bilingual who began learning both languages from early childhood would develop a fully, conceptually mediated system of language interconnection. This symmetry in lexical architecture would be reflected in symmetric, as opposed to asymmetric priming effects in cross-language tasks.; In order to test this prediction, a series of single-subject, cross-language experiments were conducted with three native, and two non-native Chinese-English bilinguals. Masked priming was used in both lexical decision and episodic recognition tasks. It was reasoned that if the hypothesized asymmetric structure of the RHM is truly a consequence of proficiency, that the native bilinguals would show symmetric priming effects, and the non-natives asymmetric effects. On the other hand, if the asymmetric structure of the RHM is not a consequence of proficiency, both native and non-native bilinguals would show asymmetric priming effects.; Among both the native and later-learning bilinguals, a consistent pattern of asymmetric priming was found in lexical decision utilizing the same presentation procedures which produced within-language priming. Cross-language episodic recognition tasks followed the same asymmetric pattern of priming. These results suggest that the levels and types of interconnection between a bilingual's two lexicons, while affected by proficiency, are not absolutely determined by it. It is concluded that the development of symmetrical conceptual mediation between the two languages of a bilingual may not be possible.
展开▼
机译：这项研究探索了少数双语者的潜在认知结构，试图确定在双语的两种语言之间发展对称的概念调解是否在认知上是可能的。先前的研究一直在双语者的实验中发现响应和启动时间的不对称性。这些不对称的模式及其获得的条件促使了当前的双语表示和处理模型的发展，即Kroll and Stewart（1994）的修订的层次模型（RHM）。熟练程度一直是解释此模型所建立的不对称联系的重要因素，认为双语者变得越熟练，他们越依赖两种语言之间的概念性中介。这说明一个双语者从幼儿期就开始学习两种语言，将会发展出一个完整的，概念上介导的语言互连系统。与跨语言任务中的非对称启动效应相反，词汇体系中的这种对称性将反映为对称。为了检验这一预测，对三位母语和两名非母语汉英双语者进行了一系列的单主题跨语言实验。在词汇决策和情节识别任务中都使用了屏蔽启动。有理由认为，如果假设的RHM不对称结构确实是熟练程度的结果，那么本地双语者将显示对称启动效应，而非本地人则显示非对称效应。另一方面，如果RHM的不对称结构不是熟练程度的结果，那么本地和非本地双语者都会显示不对称的启动效应。在本地和以后学习的双语者中，在词汇决策中使用产生语言内启动的相同表示程序，发现了一致的不对称启动模式。跨语言情景识别任务遵循相同的不对称启动模式。这些结果表明，双语者的两个词典之间的相互联系的水平和类型虽然受熟练程度的影响，但并不是绝对决定的。结论是，双语的两种语言之间不可能进行对称的概念调解。
展开▼
