This pilot study investigates optional wh-fronting in Mandarin and Cantonese in connection with topic and focus structures, aiming to uncover the similarities and differences between wh-fronting, topicalization, and focalization in the two languages, and provide a theoretical account for optional wh-fronting phenomena in the two languages. The fact that Mandarin is a topic-prominent language has led most of the previous studies to analyze wh-fronting on a par with topicalization while a recent study has proposed that wh-fronting is best-analyzed as a strategy for licensing identificational focus typically manifested as the cleft constituent in clefted questions crosslinguistically. However, both approaches fall short in providing a detailed comparison of the morphological, syntactic, and semantic properties of the three phenomena in question. Moreover, even though wh-fronting, topicalization, and focalization can be observed in Cantonese, another topic-prominent language akin to Mandarin, a systematic study of the three phenomena is lacking. By investigating the three phenomena in these two contiguous languages, this study can shed light on the debate regarding the syntactic function of wh-fronting in Mandarin.

The project investigates the syntax of clefted questions in Mandarin and French. Built on on-going research on clefted questions, which treats the cleft constituent as contrastive focus, the project aims at addressing the following issues:

- How are clefted questions formed in Mandarin and French? In what ways are the clefted questions in the two languages similar to or different from each other?

- What is/are the licensing condition(s) of contrastive focus in clefted questions?

- Do clefted questions in the two languages share the same syntactic structure?

- How does the study of clefted questions in the two languages shed light on the role of information structure in the grammar?

The study of the syntax of clefted questions in Mandarin and French takes on theoretical significance. In terms of developing a theory of focus structure, while previous studies focus on English clefted questions whose wh-word is required to undergo movement, Mandarin and French have the option of allowing the wh-word to either remain in-situ or be fronted, raising an important question about the licensing conditions available to contrastive foci in Universal Grammar (UG). The results of this study will provide new insights into the licensing conditions of contrastive foci in UG as well as the role of information structure in the grammar. 

Recent theories of generative grammar make the following assumptions about the encoding of temporal relations in natural languages. Both tense and aspect are spatiotemporal ordering predicates holding a topological relation between two time-denoting arguments. Tense is universally associated with a functional head whose structural expression is a linguistic reflex of anchoring referentiality in time and space. Aspect, which in essence realizes the “number” feature of eventuality, is not referential; it does not have an anchoring function. Tense, therefore, is a temporal structural builder whereas aspect is parasitic on tense. On the basis of these assumptions, sentences in natural languages necessarily have a tense projection, with tense being overtly expressed or covertly construed on semantic features. More recent research also shows that how languages structure events may have a bearing on morpho-syntactic realization of tense and licensing of arguments, hence the typological differentiation between “discourse language” and “event language” (Rosen 2007). Mandarin Chinese has been analyzed as a “tenseless” languages which makes use of factors other than overt tense marking to determine the temporal interpretation of sentences (Lin 2005), or various structural means to circumvent an event to anchor tense (Tsai 2007).

This project aims to investigate into the role of various aspects in Mandarin Chinese and some of its relatives in the Tibeto-Burma family, including Jingpo and Yi, which have rich aspect systems but no tense markings. In light of the above theoretical assumptions and analyses, the project intends to address the following questions to arrive at a better understanding of how tense is realized via aspect in these languages :

- what kind of aspects are requisites for interpreting tense in these languages and whether they are derivatives of a temporal predicate which spatiotemporally orders time-denoting arguments in the languages;

- whether there are other elements also spatiotemporal in nature which the languages resort to when relating to time-denoting arguments;

- whether the obligatory presence of spatiotemporal elements are manifestations of a syntactic projection of tense in the language, and

- whether these aspect rich languages are “discourse languages” and what are their relevant structural reflexes.

The study will use data from existing corpora supplemented by confirmation data collected from native speaker consultants.

The project makes an investigation into three of the representative languages spoken by the minority groups in Yunnan, namely Yi, Bai and Jingpo. The study looks at the major grammatical areas of these languages such as word order and argument structure (i.e. properties of verbs and their arguments). The method of investigation involves fieldwork and informant elicitation. Extensive and systematic organization of the language data is called for. On the basis of descriptive data and generalizations, we proceed to analyze aspects of the grammar in line with current linguistic frameworks, addressing issues of universality and variation.

The findings of the project enriched our knowledge of the linguistic structure of the minority languages and inform linguists working in both diachronic and synchronic traditions.

This project aims at investigating the relation of word order and constituency structure of a Tibeto-Burman language, Kachin (also known as Jingpo), which has not been systematically studied in the generative framework. The objectives of the research are (i) to provide a descriptively adequate account of word order and phrase structure in Kachin; (ii) to examine how word order and the hierarchical representation of phrases are related in the language, and (iii) to explore how the relation of word order and constituency structure of the language complies with principles of Universal Grammar as formulated in the generative framework. The study will center on research issues key to the current understanding of word order and hierarchical representation of constituents in natural languages.

The projected research results provide evidence substantiating whether word order in human languages is invariably determined by constituent hierarchy or whether word order and structural hierarchy are symmetrical cross-linguistically. This has significant bearing on a better understanding of the architectural design of human language structure. The set of linguistic analyses developed from the project can also inform researchers working on similar and related languages. The projected output of the research helps enrich the archive of world languages, ultimately benefiting researchers in theoretical linguistics, historical linguistics, field linguistics, typological linguistics, language evolution, and other related fields. Publications include some book chapters and journal articles (Journal of East Asian Linguistics, Linguistic Sciences) as well as conference presentations.

The capacity for symbolic reference and the ability to organize words into hierarchical groupings and recursive structures are cognitive traits that distinguish humans from other animals. How do reference and syntax develop in infants and toddlers? This study addresses this question by analyzing the longitudinal development of Chinese children from 8 months to 30 months of age, focusing on the transition from babbling into first words, and the emergence of word combinations from single word utterances. Several issues of continuity are investigated: (a) Is early lexical development marked by a naming insight and if so, what are the indicators of this sudden awareness? (b) To what extent are syntactic relations already present in the one-word stage, in the form of combinations of gestures and speech? (c) When do children show evidence for syntactic categories (such as nouns and verbs)? Should these be taken as innately given units, or are they induced from distributional regularities in the input? (d) Are children's early word combinations pivoted around particular words, or are they instantiations of abstract syntactic structures? Our findings will shed light on the nature of early linguistic competence, and the contention between nativist and usage-based approaches to language acquisition.

On the question of lexical spurt, we have found that the rate of lexical growth of the two children cannot be fitted by a logistic curve; instead, the subjects' vocabulary development reflects a transition from a period of zero growth to a phase showing steady increase. The point of inflection was found when the children's cumulative vocabulary numbered between 12 and 17 words, at the ages of one year four months, differing from standard accounts. A close connection between facility in spontaneous naming and the lexical spurt was observed. In the transition from the one-word to two-word stage, gestures are used primarily to accompany speech. Our data indicate the important role of deictic gestures in gesture-word combinations, and the special status of supplementary gesture-word combinations that express argument(s) + predicate relations; these semantic relations were first expressed by gesture-word combinations before they were realized as two-word combinations. A frequent frame analysis of adult input and child utterances indicates that distributional bootstrapping is a viable route for children to tap into the language-specific properties of syntactic categories. This, however, does not solve basic learnability problems posed by abstract syntactic knowledge that refers to categories. Our initial analysis of longitudinal data from two Mandarin-speaking children shows that early syntactic development is not piece-meal or conservative.

The proposed research investigates how young Chinese-speaking children understand the referential properties of numeral phrases such as 'yi ge laoshi' ("a teacher") or 'san ben xiaoshuo' ("three novels"), i.e. expressions of the form 'numeral+classifier+noun', in which the numeral may be omitted. The referentiality of numeral phrases in Chinese depends on complex factors such as the internal structure of the Chinese noun phrase, the syntactic position it occurs in, and the subject constraint on numeral phrases in the language. Depending on form and context, numeral phrases may be understood as referential or non-referential, specific or non-specific, and definite or indefinite. A command of numeral phrase referentiality is an essential component of one's basic linguistic competence.

In this project, we explore how children interpret the specificity of numeral phrases in various linguistic contexts; when they develop knowledge of the constraint on numeral phrase subjects as well as the modal contexts that license them; and whether children's understanding of quantifier scope is connected with their interpretation of numeral phrases. Our study will shed light on Chinese-speaking children's knowledge of referentiality as a central aspect of their semantic competence. If children's understanding of noun phrase structure and syntactic context can be shown to guide their interpretations of the numeral phrase, our study will provide support for universal syntax-semantic mapping principles. If the principles for determining quantifier scope in Chinese can be derived from specificity properties of numeral phrases in the language, then one would expect such scope principles to be acquired late. The question whether differences in scope interpretation between languages such as English and Chinese should be construed as parametric variation will then need to be re-examined.

This project investigates how bilingual children acquire two rhythmically different languages: Cantonese and English. Cantonese is a typical syllable-timed language while English is a typical stress-timed language. Recently, a number of acoustic measures were developed which can categorise languages spoken by adult speakers into distinct rhythmic clusters using only durational measurements. However, speech rhythm of bilingual children is largely unexplored. This study applies these rhythmic measures to compare Cantonese-English bilingual children with monolingual Cantonese and monolingual English children, as well as with monolingual adult speakers of the two languages for a thorough investigation of bilingual acquisition of speech rhythm.

- Monolingual Cantonese and English children display distinct rhythmic patterns already at the age of 3.

- Rhythmic patterns of the bilingual children follow those of their monolingual counterparts, but are also different from them.

- Language dominance of the bilingual children influences their rhythmic development

Early words provide the building blocks of early sentences. This project investigates the development of vocabulary and its relationship with the emergence of grammar in childhood bilingualism. It will chart the lexical development of our bilingual children, combining existing and new longitudinal corpus data to investigate the categories of their first words and how they are combined to form sentences. A number of research questions in lexical development will be addressed including:

-  How do bilingual children acquire the lexicons of two languages simultaneously?

- What are the principles governing their lexical development? At what point is there evidence for differentiation of two distinct lexicons?

- In what ways is bilingual children’s lexical development similar to and different from monolingual children in terms of rate and trajectory of development?

Moreover, the relationship between lexicon and syntax will be investigated:

-  Are early multiword combinations item-based or rule-based?

-  How do children acquire knowledge of syntactic categories such as nouns and verbs?

These questions will be addressed through corpus-based longitudinal studies and cross-sectional experiments. The findings will have theoretical implications for the application of influential principles proposed in child language acquisition research, contributing to the debate on language differentiation in bilingual children. Early lexical differentiation serves as the basis for syntactic differentiation, the precursor for the construction of two grammars. The relationship between the lexicon and syntax in bilingual development is important to our understanding of how the acquisition of individual words is correlated with the ability to put words together to form meaningful sentences in the bilingual child’s two linguistic systems. We aim to contribute to the on-going debate between usage-based and nativists approaches by addressing key theoretical issues in the emergence of syntax in bilingual children. The role of the lexicon represents a new frontier in our team’s research in childhood bilingualism. The research will significantly extend our current line of work in syntactic development to cover the first words in the single word stage before the emergence of syntax, yielding a more comprehensive picture of early bilingual development. The practical significance of the project lies in the establishment of baseline data for typically developing bilingual children which is required for comparison with atypically developing bilingual children. The findings will be useful for assessment of bilingual children and the learning of early words in Cantonese and English.

Children who are exposed to both Cantonese and English in their preschool years achieve various degrees of bilingualism. This project focuses on the interface between childhood bilingualism and child second language acquisition, investigating the developmental processes underlying Hong Kong children’s acquisition of bilingual and second language competence. In the two acquisition contexts we seek to address (a) empirical issues regarding the ways in which different groups of child learners exhibit similarities and differences in terms of their developmental profiles and ultimate attainment; and (b) theoretical issues regarding the initial state of acquisition, epistemology, underlying processes such as cross-linguistic influence and directionality of transfer.

A number of complex grammatical structures which are potential ‘vulnerable domains’ in language development, including relative clauses and dative constructions, will be investigated in depth to ascertain the relationship between word order and processing effects in comprehension and production. A two-pronged strategy is used in the project:

-  longitudinal study - 2 bilingual and 2 child second language learners will be video- and audio-taped over a period of one to two years. The resulting corpus data will be transcribed in a standard format, digitized and linked with sound and video files.

- experimental study - cross-sectional experiments will be carried out to investigate the acquisition of relative clauses and dative constructions in Cantonese and English.

The data will constitute two new multimedia corpora, one bilingual and one child second language corpus. The data will be contributed to the CHILDES online database of child language.

The study investigates the lexical development of bilingual children acquiring Cantonese and English simultaneously from the one-word stage to the multiword stage. The longitudinal data of six bilingual children from the Hong Kong Bilingual Child Language Corpus will be analyzed to address the following questions:

-  What is the type and token frequency of nouns and verbs, two major lexical categories in each language throughout their development from one and a half to four and a half? How does the noun-verb ratio in English compare with that in Cantonese? How does the lexical development of the bilingual child’s dominant language compare with that of the less dominant language? In what ways is bilingual children’s lexical development similar to and different from that of their monolingual counterparts?

-  At what stage do the bilingual children show evidence of lexical differentiation? How early do they acquire translation equivalents, i.e. using two words to refer to the same object? What is the percentage of these translation equivalents in different stages of development?

- What coinages are created by the bilingual children which are not exemplified in the adult input? Do they code-mix within newly coined words?

The findings will contribute to a number of controversial issues in lexical development and form the basis for a large-scale CERG proposal investigating both lexical and functional categories in bilingual development and the interface of acquisition of lexicon and syntax in bilingual children.

The project aims to revisit selected issues in the grammar of Cantonese from three complementary perspectives which have been investigated in the researchers' recent projects. The typological perspective offers a comparative dimension; the processing perspective focuses on how speakers of the language process incoming speech and formulate sentence structures in real time; and the acquisition perspective considers how the same structures develop in children acquiring Cantonese in monolingual and bilingual contexts. The intention is to use these complementary perspectives as converging evidence to illuminate Cantonese grammar.

This is an innovative study combining data from corpora and psycholinguistic experiments conducted by the research team to shed light on the analysis of grammatical phenomena in Cantonese. Besides leading to specialized studies of grammatical topics in the form of journal articles, the work will contribute to a substantially revised edition of the investigators’ influential work, Cantonese: A Comprehensive Grammar, first published by Routledge in 1994. The results are expected to contribute to theoretical and empirical issues in the study of Cantonese grammar. The findings of the project will also be useful to a wide range of professionals and researchers including linguists, psychologists, educators and clinicians whose work involves Cantonese.

This project aims to develop an Internet-based, computer-aided pronunciation training (CAPT) platform to help our students (both undergraduate and graduate) improve their English pronunciation. Pronunciation training faces the challenge of negative language transfer, which refers to the influence of well-established perception and production of sounds in the first language (L1) on the second language (L2). Such influences are ingrained with age and hamper acquisition of a second language. Pronunciation improvement requires persistent practice. In this regard, CAPT has much to offer, in terms of an around-the-clock, self-paced and personalized platform for the learner to practice pronunciation drills. In this project, we focus on language transfer effects due to L1 Chinese (Cantonese and Mandarin) on L2 English. We propose to phonetically transcribe a large corpus of L2 English speech recordings from native Chinese speakers. Analysis of the phonetic transcriptions will enable us to identify pronunciation deviations from the target productions, which constitute accentual variations and mispronunciations. We will organize the mispronunciations according to a gradation (i.e. from subtle to salient) for pronunciation training. We will also develop an automatic speech recognizer (ASR) that focuses on the detection and diagnosis of the identified mispronunciations. The ASR will be integrated into an Internet-based, interactive CAPT system that is accessible by a large number of learners for pronunciation drills. This can significantly augment the language educators’ efforts in training the learners and create impact in enhancing learners’ perception and production of English sounds.

The project aims to examine how deaf children acquire the classifier predicates in Hong Kong Sign Language. An experiment was set up in which narratives in HKSL based on 6 comic strips with clear and straightforward story lines were collected from 14 deaf children, The narratives transcribed using ELAN, a software package developed by the Max Planck Institute of Psycholinguistics, Nijmegen, Netherlands. Based on their narrative productions, the 14 children were grouped into three levels of proficiency (Level 1-3) and systematic analysis was conducted.

- HKSL systematically displays a set of verbal classifiers which can be functionally categorized into semantic, size and shape specifiers, and handling classifiers

- Deaf children had experienced initial difficulty in acquiring the system, leading to various errors in terms of movement, handshape, spatial configuration between the dominant and the nondominant hands.

- The results lead to a hypothesis that deaf children initially treat classifier verbs as plain verbs until they have developed knowledge about the compositional nature of the classifier predicats, as against the lexical nature of plain verbs.

The Asia Pacific Sign Linguistics (APSL) Research and Training Program supported by the Nippon Foundation is a multi-country and multi-phase project that aims at developing sign linguistics as a research discipline in the Asia Pacific region. It provides expert training in sign language research and sign language teaching to Deaf and hearing researchers from different parts of Asia. It also aims to develop research to document the sign languages in the region, through the production of sign language dictionaries and teaching materials. The first phase of the project ‘Practical Dictionaries of Asian-Pacific Sign Languages’ involved four sites in Asia: Vietnam, Cambodia, Philippines, and Hong Kong. The second phase of the research project will last for 6 years and the training is organized through a series of sub-degree programs for the Deaf and postgraduate programs for the hearing researchers. Universities in Asia that have an interest in developing Sign Linguistics research are invited to participate in the Program. Currently, the Centre is hosting Deaf and Hearing Researchers from Indonesia, Sri Lanka, and HK, as well as collaborating with the University of Indonesia and University of Kelanya.

- The Program has led to the publication of a set of user-friendly dictionaries based on Hong Kong Sign Language, Ho Chi Ming Sign Language, and Cambodian Sign Language. It has also supported the documentation of indigenous signs used in the various parts of the Philippines.

- The Program also supports the development of a sign language database “Asian SignBank” to support future Sign Linguistics research developments in the region.

Dr. Alex K Yasumoto has generously donated to the Centre to support the project “Child Sign Language Documentation”. The donation has been used to support a team of Deaf and hearing researchers in continuing to construct the Child HKSL Corpus as well as developing a HKSL-Cantonese bilingual acquisition Corps. The Child HKSL Corpus was initiated based on grants from the University Grants Council. Thanks to Dr. Yasumoto’s donation, the team can continue to build the corpus and to work with Prof. Brian MacWhinney, director of CHILDES, Carnegie Mellon University, on aligning the sign language acquisition data with the format of CHILDES, in hope of anchoring the data to support future child data exchange activities internationally.

- Deaf children are capable of learning language, not only sign language, but also spoken language. In other words, deaf children can become sign bilingual.

- The team has been using the data generated from the corpus to analyze the development of verb agreement, classifier constructions, modals, negation, verb types, and perfective aspect, and handshapes phonology in child Hong Kong Sign Language.

This 4-year project symbolizes ground-breaking research on sign language acquisition in HK and Asia, focusing on the development of Hong Kong Sign Language (HKSL) by a deaf child The research team collected data from the deaf child weekly starting from the child’s first exposure to HKSL at one year nine months of age. The project aimed at establishing a corpus of child HKSL in order to support further research on how deaf children acquire language. Activities include weekly recordings of the deaf child’s interactions with his mother, siblings and a native deaf researcher. The period of data collection is 3 years and the transcription is done on a monthly basis using a computer program ELAN, a multimedia annotator developed by the Max Planck Institute of Psycholinguistics, Nijmegen, The Netherlands.

- Establishment of Child HKSL Corpus

- The deaf child’s HKSL development show a similar sequence starting from gestures and pointings, gradually evolving into one sign stage, two sign stage and finally the formation of simplex and complex sentences.

- In terms of the development of specific linguistic structures, the deaf child show delays in acquiring verb agreement in HKSL. It is hypothesized that the nature of verb system in HKSL as well as the constraints involved in person agreement lead to this difficulty.

In August 2006, the Centre received a major donation from The Hong Kong Jockey Club Charities Trust to run a research program experimentating on a form of education that potentially benefits both deaf and hearing students linguistically and academically. This 7-year project explores an alternative model of deaf education that draws insights from recent advances in sign linguistics, sign bilingualism as well as co-enrolment in deaf education. The project also evaluates the effectiveness of this model of deaf education on the language development of deaf and hearing students, from the perspective of sign language development, literacy and oral language development. The program also looks into the nature of a sign-bilingual and co-enrolment classroom through classroom process research. The programme has the capacity of nurturing local expertise in conducting research on deaf education and sign linguistics, as well as teachers who have knowledge of teaching deaf children effectively. The research programme will also lead to a body of research outputs that sets the history of language in deaf education in Hong Kong.

- Both deaf and hearing children are benefiting from the pedagogical processes involving both deaf and hearing teachers, with evidence coming from their classroom participation, academic results, and language assessment results.

- Deaf children benefit from this programme in various degrees and aspects. There is an overall improvement in HKSL, which is crucial for them to extract information in classroom teaching. There is also improvement in their oral language development, suggesting that the learning of HKSL does not impede their oral language development.

- There is an interesting code switching phenomenon in the classroom processes, suggesting that both the deaf and hearing children are becoming more and more sign bilingual when learning in this environment.

“Promotion of Deaf Communication through Hong Kong Sign Language" is a Deaf awareness project conducted jointly with a local deaf school and the Audiological Services Section of the Education and Manpower Bureau (EMB). The aim of the project was to explore the effect of adopting HKSL as a language of communication in a deaf school setting and to help hearing teachers of deaf students and the hearing students in the neighborhood to appreciate the nature of sign language in the Deaf community. The project consisted of sign-supported reading programs, seminars, workshops and sign language classes for parents, school teachers, hearing students in the neighborhood, and an experimental program adopting Hong Kong Sign Language as the medium of instruction in Chinese lessons. Through classroom observations, researchers investigated the processes involved in the students’ learning activities and the teachers’ readiness to accept natural sign language as a language of communication in various classroom learning situations.

- Introducing HKSL in the experimental classroom saw an increasing interest in reading in Chinese by deaf students and motivation in classroom participation.

- Deaf students benefit from reading in Chinese supported by HKSL with varying degrees, depending on their language foundation. Those with better signing skills show more improvement in literacy than those with less.

- Total Communication that biases towards spoken language and limited signing impedes classroom learning.

This research project examined the grammatical properties of the classifier predicate system in Hong Kong Sign Language. The aim was to verify whether HKSL has verbal classifiers like other sign languages reported in the literature and to categorize them according to their morphosyntactic characteristics. The activities involve working with native Deaf people to identify and collect pictures of signs for analysis.

- HKSL displays a set of classifier handshapes which contribute to the construction of verbal and adjective predicates. The movement of the sign represents the verbal or descriptive root and the classifier handshapes represent the arguments in the predication.

- The classifiers can be categorized into different subtypes: semantic, size and shape specifiers and handling.

Research in the domain of second language acquisition has found that adult second language learners have not lost the ability to distinguish novel foreign speech sounds in the target language. However, they must rely on repeated focused trials to link changes in the auditory syllable. This project aims to use optimized online Pinyin dictation trainings to assist Cantonese speakers to strengthen the perception of Putonghua segmental and suprasegmental distinctions, and to refine the skill of transcribing Chinese syllables into Pinyin Romanized characters. The project is composed of two lines of research. The first line systematically investigates the Pinyin spelling errors produced by Cantonese learners to determine the confusions Cantonese learners encounter. This detailed characterization will serve as a solid empirical guide to the instruction and learning of Putonghua phonology and Pinyin orthography in Hong Kong. The second line evaluates the effectiveness of the newly developed component-based training module by examining both the accuracy and fluency of Pinyin spellings, and comparing the corresponding statistics with the traditional whole-item-based training module. It is expected that learners in the new module treatment group will master Pinyin transcription skills faster, because redundant practices of items that contain mastered components are avoided. Findings from the study will bring new understandings of how training can assist the restructure of second language phonological system.