All in all, the thesis tries to fill the research gap in the study of English pronunciation in a Hong Kong context. It exhausts the pronunciation of all types of English consonant clusters by native Cantonese learners of English, provides experimental studies to investigate the most effective approach to the acquisition of English consonant clusters, and applies Optimality Theory to elucidate the pronunciation phenomena found among Cantonese learners of English.
Apart from the inquiry into language pedagogy, this thesis also analyses and expounds the language phenomena of consonant cluster production through the application of Optimality Theory. From the data of the pre-test and post-tests, the pronunciation modification phenomena of subjects were summarised and analysed. It is found that subjects produced pronunciation forms that resembled that of native English speakers, but at the same time they produced forms that deviated from that of native English speakers. The most frequent pronunciation modifications are substitution and deletion.
Optimality Theory argues that phonological differences between languages are the results of the differences in the ranking of universal constraints; therefore, the preference towards certain modification phenomena and the production of certain pronunciation forms by Cantonese speakers can be viewed as the differences in the ranking of constraints between English and Cantonese. In the interlanguage phonology of Cantonese speakers, words like pray /pre I/ and find /faInd/ could become [pe I] and [faIn] phonetically. This can be explained by assuming that the English words are input to a phonology in which *COMPLEX (complex onset or coda is not permissible) is highly ranked, and that in order to observe this constraint, Cantonese learners of English might apply deletion to delete a consonant in syllables with consonant clusters.

There has been a growing concern that students in Hong Kong have low proficiency levels in English. Ths constitutes a challenge to language planning policy, which aims at providing opportunities for extending knowledge and experience through the acquisition of English grammar. Grammar forms a major component in English language teachng, and mostly has priority over other parts of the language. The acquisition of second language syntax is the focus of ths research. The main aim of the present study is to investigate the development of relative clauses in the interlanguage of native Cantonese learners of English within the framework of Keenan and Comrie's Noun Phrase Accessibility Hierarchy (1977, 1979). The study also seeks to distinguish the features of Cantonese-English interlanguage which may be attributed to language transfer from those which may be considered as universal to second language acquisition. The role of L1 is then examined. The study focuses on the theoretical concepts of language transfer, comparative study of Cantonese and English relative clauses as well as language universals from a typological perspective.
The issues are investigated with a database of syntactic judgements and production by a p u p of Cantonese-speaking students of English as a second language in Hong Kong. A number of features, namely, the use of resumptive pronouns, the retention of the head noun, shft of noun functions, nun-adjacency of .head noun and relative pronoun, preposition strandinddeletion, and avoidance of the use of zero relativizer are found in the data. These features are discussed and it is shown that they are more attributable to universal and developmental factors than to direct L1 transfer. One important finding of the study is that there exists an uncommon feature of retention of the head noun. Its presence is suggestive of lower proficiency levels in the target language.

This study investigates how expert-novice scaffolding and peer-peer scaffolding may facilitate the learning of grammatical forms in tertiary Chinese EFL classrooms and attempts to illustrate the features of different modes of interactions. It was a classroom research study with a quasi-experimental design, in which Chinese tertiary level students from two intact classes in Hong Kong and Wuhan were both given a task of revising a student's writing, focusing on the learning of the relative clause in three interactional scenarios: student-student (S-S), teacher-student (T-S), teacher-student + student-student (T-S + S-S).
On the basis of the microgenetic analysis of classroom discourse complemented with quantitative results, the study has found that scaffolding in T-S + S-S interaction generates the most effective impact on learning the target form in the three settings, followed by T-S interaction and S-S interaction. The learner's Zone of Proximal Development (hereafter ZPD) could be promoted most effectively in T-S + S-S interaction due to high quality of scaffolding functions, whereas there is little evidence of extension in the learner's ZPD in T-S and S-S interactions due to a lower quality of scaffolding functions. Moreover, the effective use of the seven scaffolding functions, i.e., Recruitment, Simplifying the task, Direction Maintenance, Marking Critical Features, Frustration Control, Demonstration, and Feedback, in T-S + S-S interaction, seems to stretch the tertiary Chinese students' grammatical competence to a greater extent than in the other two settings. More importantly, the findings have highlighted the importance of helping students notice the gap in their knowledge, promoting self-regulated learning of target forms, and encouraging independent learning when the teacher provides appropriate support.
The study has both theoretical and pedagogical significance. Theoretically, it has shown that grammar learning in EFL classrooms does not happen

This research is motivated by the scarcity of investigation into L2 learners' online processing problems in listening comprehension, in particular those relating to L2 proficiency. It aims to identify the online processing problems encountered by Hong Kong students of differing L2 proficiency and whether these problems can be overcome by compensatory schema use. Two questions are addressed: 1) Given the same listening situation, do high and low proficiency students experience the same online processing problems? 2) Does success in compensatory schema use vary according to L2 proficiency? Four data collection methods (semi-structured interviews, questionnaire survey, partial transcription, and introspection) and both quantitative and qualitative data analyses are employed. Exploratory semi-structured interviews indicate that apart from the problems of unfamiliar topics, vocabulary and accents, students report experiencing what might be called an "acoustic blur" during L2 listening. They hear streams of English speech sounds but cannot identify the key words. Even when they think they have identified the words, these words often do not fit the context. The questionnaire survey was designed using the data gathered from the semi-structured interviews. The survey results show that practice frequency has a significant effect on students' perception of overall listening difficulty; the more often the practice, the less often listening problems are perceived. Slightly more than one third of respondents have almost no listening practice in English outside class. L2 proficiency impacts significantly on students' perceptions of processing problems such as those caused by unfamiliar topics and vocabulary, and failure in sound segmentation and word recognition. In their self-assessment of word recognition, low proficiency students' word recognition in an average lecture and TV program is less than 50% and 40% respectively, whereas that of proficient students is 80% and 65%. These

The English verb BE, with its complicated forms and multiple functions, has posed great learning difficulties for learners of English both as the first and a second language. Most previous studies on the acquisition of BE have only examined a small number of participants and used data from experimental tasks, which have failed to provide a full picture of how BE is acquired and used by learners. By adopting a corpus-based approach, this study presents a fuller picture of the forms and functions of BE in the interlanguage grammars of Chinese learners of English.
Two learner corpora were used in this research, namely, Hong Kong Baptist University Interlanguage Corpus (HKBUILC) and Chinese Learner English Corpus (CLEC), consisting of English essays written by students from Hong Kong and Mainland China, respectively. Three subcorpora of CLEC (i.e. ST2, ST3, ST4) were adopted, representing Chinese learners at three different English proficiency levels. To obtain a comprehensive distribution of interlanguage BE, all the tokens of BE, including non-target-like, overgenerated BE, as in " Mary was lost the necklace she borrowed from her friend ", and omission of BE, as in "My mother would O angry with me", were annotated with respect to their syntactic and semantic properties. The data were then retrieved type by type for examination by using WordSmith, a concordancing software.
To account for the variable use of BE in learners' interlanguage grammars, this study further investigated a few factors that may influence the production and omission of BE, such as the following linguistic distinctions: (1) copular vs. auxiliary BE; (2) finite (am, is, are, was, were) vs. nonfinite (be, being, been) BE; (3) stage-level vs. individual-level predicate types (Carlson, 1977); (4) different post-BE constituents · · e.g. ergative vs. unergative verbs (Perlmutter, 1978); and (5) the differences between L1 and L2 (i.e. English BE vs. Mandarin Shi/Cantonese Haih).
The concordancing

The notion of 'balanced' bilingual development has often been used to refer to two languages which are spoken with equal proficiency during bilingual language acquisition. However, it is still not known how 'equal' the proficiencies of the two languages should be in order to achieve 'balanced' bilingualism. This case study aims to capture this notion of 'balanced' by using the following quantitative measures with respect to grammatical development: mean length of utterance (MLU), upper bound (UB) and lexical diversity (D). Direct comparison between two sets of absolute values does not reliably predict a bilingual child's language dominance due to potential typological differences two languages. The obtained values of these measures are shown to be comparable with those of monolingual children acquiring each language, and thus it is argued that bilingual children who attain such competence in each language are considered as having 'balanced' bilingual development.
This study concerns particularly two aspects of bilingual children: a) grammatical development and b) language use in terms of mixing. With regard to a), given that cross-linguistic influence is attested across studies of bilingual language development, this case study investigates the domains of null objects and wh-interrogatives to see if any cross-linguistic influence takes place. The qualitative similarities found in all Cantonese-English bilingual subjects suggest that the nature of the cross-linguistic influence is similar, whereas the quantitative differences imply that language dominance has determining effect on the phenomenon.
Concerning b), since language proficiency is often taken as one of the crucial factors predicting language mixing pattern, this present study examines the relationship of the mixing pattern in the Cantonese-English balanced bilingual child and her language proficiency. The results show that mixing occurs more frequently in Cantonese context than in English context.

This study examined the roles of sub-character processing, phonological awareness, morphological awareness, and orthographic knowledge, measured using twelve different tasks hypothesized to indicate these four broad constructs, in Chinese character recognition and English word reading among 536 Hong Kong Chinese kindergartners, second- and fifth-graders. The twelve tasks generally showed an increase in performance with grade level. Confirmatory factor analyses comparing alternative models of these four constituents of Chinese word reading revealed a dynamic pattern of children's latent linguistic or reading processing skills development: The best-fitting model of kindergartners' processing was one that included two broad constructs, broadly termed metalinguistic processing and orthographic processing. In contrast, second-graders showed a fine-grained sensitivity to four distinct skills of sub-character processing, phonological awareness, morphological awareness, and orthographic knowledge. Finally, the latent processing skills of the fifth-graders converged into phonological and orthographic processing. The contributions of each of these initially specified constructs, i.e., sub-character processing, phonological awareness, morphological awareness, and orthographic knowledge, to Chinese word reading varied across each separate grade in regression analyses. The sub-character processing construct was uniquely associated with kindergarten Chinese word reading only. In contrast, the morphological awareness construct was uniquely associated with Chinese word reading in both second- and fifth-graders. The orthographic knowledge construct was uniquely associated with word reading across ail three grades. However, the phonological awareness construct was not uniquely associated with Chinese word reading in any of the groups of children, though it was uniquely associated with English word reading, even with Chinese character recognition skill statistically controlled. These findings

Two scales of mothers' mediation of their children's writing based on Aram and Levin (2001) were developed and tested in 67 mother-child Hong Kong Chinese dyads in three grade levels - second-year kindergarten, third-year kindergarten, and first grade. With children's ages, grades, and non-verbal IQs, as well as mothers' education levels statistically controlled in a regression equation, the four strategies of the 'Print Mediation' scale, a measure of mother-child joint writing productions, explained a unique 10% of the variance in children's word reading. These four strategies tended to change with age such that higher autonomy was more common among older children and lower autonomy was more prevalent among younger children. In a separate hierarchical regression equation, the 'Literate Mediation' scale, tapping specific writing strategies, explained a unique 8-11% of the variance in children's word reading. A prevalence of copying strategies tended to be negatively associated with reading skill and a prevalence of morphologically based strategies tended to be positively associated with reading skill, even apart from age and grade level. Findings demonstrate some strong developmental trends in the writing process but also suggest that a more analytic approach to Chinese character writing may be helpful for literacy development, even among the youngest children.
[Copyright of Language and Cognitive Processes is the property of Psychology Press. Full article may be available at the publisher's website: http://dx.doi.org/10.1080/01690960801970615]