We discuss the developmental stages of wh-movement in four Cantonese-dominant bilingual children who are exposed to Cantonese and English from birth, drawing on the Hong Kong Bilingual Child Language Corpus (Yip and Matthews 2000) and the diary records kept by the parents. The longitudinal data suggest that these bilingual children go through an initial wh-in-situ stage, followed by an optional wh-movement stage and a partial wh-movement stage in their English. Our discussion will focus on examples of partial wh-movement where a wh-phrase is moved to some intermediate position, such as the beginning of the embedded clause, with a copy of the wh-word in the clause initial position as in (1):

(1) Who did you t'ink who broken it?
[to father, referring to broken car] (Timmy 3;10;30)

Another option, not discussed in the studies of monolingual English-speaking children but used productively by our bilingual subjects, is to move the wh-phrase only as far as the beginning of the embedded clause, leaving the main clause without any indication that a question is being asked:

Partial wh-movement is attested in several adult languages such as German and Romani (McDaniel 1989), and as a developmental phenomenon in the acquisition of English by monolingual children (Crain & Thornton 1998). These data suggest that developmentally, partial movement is an intermediate step between wh-in-situ and full application of wh-movement. This is consistent with the view of partial wh-movement as a 'transitional behavior' in language acquisition (McDaniel, Chiu & Maxfield 1995). The wh-in-situ stage in bilingual development provides evidence for cross-linguistic influence from Cantonese, a language without wh-movement to English while the partial wh-movement stage suggests availability of an option in Universal Grammar though not instantiated in English.

Are you thinking of submitting your research to the LLL Conference? Or another conference? This workshop will introduce students to the fundamentals of writing abstracts for conferences. We'll cover things like:
- What a conference abstract is
- How a conference abstract is used
- How to write a good abstract
- Who should see your abstract before you submit it

Phonetics Lab (Moore 162) General Lab (TP107*)
*TP107 is located behind the Center for Korean Studies.

What you can learn:
(1) Phonetics Lab: an informal introduction to digitizing
(2) General Lab: an informal introduction to PsyScope and E-prime

- What are the LAE Labs?
- When are the Labs open?
- What can I do in the Phonetics Lab and/or General Lab?
- How do I copy my audiocassettes to a computer?
- How can I look at visual representations of speech sounds?
- How can I investigate learners' perceptions of speech sounds?
- What language databases can I find in the Labs?
- How do I become a lab member? etc.

This presentation is an effort to explain why dependent person markers pattern as pronominal clitics in the Austronesian language Selayarese. To do this, I use the Minimalist Program as my framework. Prior work on the language omits the basics of pronominal cliticization, jumping to more complex phenomena. Although my aim is primarily to answer the question why certain bound pronominal forms pattern as clitics on the verb, three dimensions of the verb phrase are crucial for producing an anaylsis of syntactic structures in Selayarese: clitic doubling, resemblance to “Philippine-type focus system” and ergative case marking. My argument to explain the data is that a definiteness feature of the determiner phrase dictates the distribution of pronominal clitics, and their associated noun phrases, by affecting which entities enter specifier position of vP.

The substrate languages of a creole are the ancestral languages of its speakers – in the case of Hawai‘i Creole (locally known as Pidgin): Hawaiian, Portuguese, Cantonese, Japanese, etc. Bickerton (1981) argues against the influence of substrate languages in creoles, dismissing the notion that a language could be made up of a mixture of features from other languages. He uses the term “cafeteria principle” (attributed to Dillard 1970) to ridicule the idea that a language could select features from various sources like items chosen for lunch at a cafeteria. However, this talk demonstrates that several aspects of the morphosyntax of Hawai‘i Creole have been modelled on features of different substrate languages, and therefore that a certain degree of mixing has occurred. However, the cafeteria principle that is demonstrated is not without constraints, and the talk concludes with a discussion of some of these.

Steyvers et al. (2003) showed that large-scale word association data possess a small-world structure characterized by the combination of highly clustered neighborhoods and a short average path length. In our research, we confirmed this feature and further discovered interesting associative relations between words, by applying graph theory to a semantic network. Our associative semantic network was constructed based on Ishizaki Associative Concept Dictionary of Japanese Words and by applying graph clustering methods called Markov Cluster Algorithm (Van Dongen, 2000) and its customized Recurrent Markov Cluster Algorithm (Jung, 2006). A shortest path of any two words on a semantic network indicates a conceptual relevance and distance between those words, whose information could be more creative according to the searching way of it. A web-based composition support system called ACSS (Associative Composition Support System), which was implemented based on the associative semantic network, provides two types of associative information by different searching ways of a shortest path between words, which expect to be useful for vocabulary learning and composition practicing.