Transcript Document
Researching learner English on a portfolio corpus --A research proposal for diachronic studies Li Wenzhong Learner corpora Things I feel comfortable about the current learner corpora: Recognized learner community Homogeneous texts in terms of types, domains, and styles Strongly diagnostic Contrastive studies REFLECTIONS: LEARNER CORPUS AND CIA • • • • Naturalness and authenticity of the learner data: are they reliable? CIA and CEA: are they valid? Language norms and the learning goals: are they the same? The prospect of developing and using the learner corpus: o small-size corpora specific to the learners in the given context and community; o monitor learner corpus; o multimedia retrieval and analysis (李文中，2009，CIA方法评析，外语电化教学（3），pp. 13-17.) WHY: 1) Students vary in their language competence and have therefore diverse learning problems. The features and problems identified from the learner corpus often mismatch those encountering the individual learners at large in real life; 2) Learner language is changing fast, and what is 'typical' of their language use at "one point of time" is hard to characterize that of another; 3) Learner's language learning is highly individualized and contextualized. So a learner corpus needs to be localized for better description and analysis of the learner language. 4) Any study of learner language must be oriented to the enhancement of learners' communicative competence as a whole. Therefore learning efficiency should be the primary concern of learner corpus research or interlanguage analysis whatsoever. W HAT:portfolio DEFNING CORPUS A learner corpus PORTFOLIO is a collection of the learner texts, both textual and multimedia, in writing and speaking that represents the language progress and development as well as the assessment of the individual learners over time. A learner portfolio corpus has the following features: -- learner perspectives: created and used by both learners and teachers -- dynamic record of the continuous learner language output -- online writing, editing and sharing -- organized in diversified annotations -- well integrated in the learning environment, e.g. ePortfolios -- oriented to the learner accomplishments CONSTRUCT INNOVATIONS A learner portfolio corpus has the following highly innovative constructs: -- learner texts and text revisions -- learner multimedia output: audios, videos -- contrastive studies: contrast of progress; contrast of learners; contrast of varieties; contrast of topics -- developmental feature capture -- learner reflection -- individualized tutoring -- diversified data sources A comparison construct Current learner corpora text Our prospective learner corpora multimedia sampling One point of time Over time Focus Group/community framework goal Individual/group/ community EA/interlanguage/CIA Constructivism Difficulties/ problems accomplishments/ progress RESEARCH OBJECTIVES To investigate and explore: -- how the learners, as an individual and a group, progress and develop in the EFL learning and to what extent their language production represents their accomplishment at large; -- the areas, of an individual learner or learner community, that are underdevelopment and to what extent do they affect their (his or her) overall communicative competence; -- how the learners get benefited from the feedback from both their peers and teachers; -- how each individual learner relates to the community, with which he or she is identified, in language development and underdevelopment; MA ETHOD & INSTRUMENT working plan -- develop the learner portfolio corpus that consists of subcorpora of each individual learners in a given learning community; -- each learner's writing pieces and scripted speaking and the feedback are well documented and pooled for community analysis; -- allow the learners to interact with their peers and accomplishments en route to success; -- online corpus analysis of the learner language as feedback to the learning process; -- get integrated in an e-Learning environment; -- allow the learners to showcase their accomplishments via sharing online; Design Text collection Text types Text Parallel texts Multimedia Text annotation: XML Data retrieval AREAS OF STUDY -- diachronic observations on the learner language -- categorize learners with specific features -- individual learning problems and difficulties -- learning strategy and teaching strategy -- action research in the e-learning environment APPLICATION TECHNOLOGY -- E-learning environment management interface: such as google apps (gmail, docs, translate toolkit, notebook, groups, reader, site, youtube, picasa); -- data mining and process; -- corpus apps: wordlists, KWIC analysis, keyword analysis, collocation study; -- Open Corpus R & D Platform; PREDICTED PROBLEMS AND SOLUTIONS -- methodology of diachronic studies; -- the student privacy; -- the life cycle the corpus; -- noise reduction; -- user training; -- multimedia processing and concordancing; -- put things together; THANKS FOR ATTENTION [email protected]