Previous studies on the sensitivity of Subjacency by adult L2 learners whose native language does not observe the rule have drawn different conclusions concerning adult sensitivity to Universal Grammar (UG) principles. This study further explores this issue by investigating not only Subjacency but also the Empty Category Principle (ECP). Using Chinese L2 learners of English, the present study tests their limitations on extraction out of several island conditions and their sensitivity to Wh-arguments (what, who, which) and Wh-adjuncts (when, where, how and why). Participants in the study included 180 Chinese freshmen and sophomores in a Chinese university, who were non-English majors and had never been exposed to an English speaking country and 16 Chinese L2 learners who were studying at University of Massachusetts at the time of study and who had at least 3 years of intensive English training before and had continually employed English afterwards. 25 English-speakers also participated in the study as a control group. They were asked to perform a grammaticality judgment task and a reading comprehension task on Subjacency and the ECP. The proficiency of the first group was measured with CELT and Assessment of Syntactic Capabilities tests. The study has found Chinese L2 learners demonstrated limitations on extraction from island conditions. Once they had sophistication in English, their performance score on Subjacency tasks showed no difference from that of the native English-speaking group. The informants also treated the different island conditions differently. They also distinguished Subjacency violations in relative clauses from that in noun complement clauses. In the reading comprehension task, the 180 Chinese informants had similar patterns to the control group and the children in DeVilliers' study. They allowed Wh-LD movement when the COMP in the embedded clause was not filled in English; when the COMP in medial was filled, they (like children and native speakers), gave answers to the lower clause when the trace was properly governed; they distinguished argument questions from adjunct questions by giving more answers to the former than the later questions. The study considers the implications of the above results for L2 teaching.

This dissertation reports findings from a sociolinguistic ethnography that examined relationships between literacy practices and personhood. The study involved the formulation of a writing club at an urban middle school, involving a multiracial group of women from the lowest academic track; two were described as special education students. They researched and wrote about their communities, investigating questions of personal concern about issues of racism and sexism. Students interviewed community members, including artists, organizers, neighbors, and peers. Students wrote up and published their findings. I collected data on the writing project, including forty-five hours of taped data. Analysis involved thematic and textual analyses of the students' written artifacts, and microanalysis of videotaped events. A microethnographic analysis examined sociolinguistic processes that research suggested is important. Attention was paid to the social construction of intertextuality during writing activities. The findings show that the nature of literacy practices and personhood is such that they are continuously and inherently constructed within particular fields of intertextual semantic potentials. These intertextual potentials are described along five dimensions: (1) ways students' definitions of personhood changed over the course of the project, (2) strategies students, community members, and myself used to position students, (3) how the project's structure positioned students, (4) community literacy practices and how they positioned people, (5) how students used community literacy practices to position themselves and others. The student's definitions of personhood changed. The established field of intertextual semantic potentials was influenced by changes in literacy practices that led to changes in literacy practices that led to changes in the students' definitions of writing, their views about themselves and life in the community. Literacy practices established in the writing project built on ones students encountered as they researched their communities. Community members shared ways of acting for social justice, including the importance of reclaiming cultural heritage, learning history from the community's perspective, analyzing multiple forms of oppression. Students' ethnographic research helped them reflect on their communities by enhancing their understanding of the cultural dynamics in which they live. Students recreated methods and theoretical frameworks to address the issue of personhood as students, as community members, and as ethnographers of their own communities.

Research has shown that bilinguals have slower performance than monolinguals on tasks measuring word recognition, number naming, and picture naming speed (Magiste, 1979; Ransdell & Fischler, 1987). Specifically, English speaking bilinguals have been shown to be slower than native English speaking monolinguals on lexical decision tasks performed in English (Ransdell & Fischler, 1987). The present study examined whether differences in visual word processing speed between bilinguals and monolinguals could be accounted for by the presence of orthographic structure similarities across the two languages of bilinguals. This was tested by having native English speaking monolinguals and English-Spanish bilinguals name or make lexical decisions on English words that had language-specific orthographic structures or nonspecific orthographic structures. The study predicted that orthographic structure similarity would slow down the visual word processing of English-Spanish bilinguals relative to monolinguals. The findings from Experiments 1 and 2 did not provide evidence for differences between monolinguals and English-Spanish bilinguals in the time required to process English words. The results of Experiment 2 provided evidence consistent with the hypothesis that similarities in the spelling patterns of Spanish and English words affected the word naming times of English-Spanish bilinguals relative to monolinguals. The results were interpreted as consistent with models of interactive language functioning in bilinguals.