Guest Lecture ComputerAssisted Language Learning Matthew Kam Department

Guest Lecture: Computer-Assisted Language Learning Matthew Kam Department of Electrical Engineering and Computer Sciences, and Berkeley Institute of Design University of California at Berkeley, USA This material is based in part upon work supported by the National Science Foundation under Grant No. 0326582.

Relevance of ESL to Third World English is a global language: 1. 2 to 1. 5 billion people in >170 countries (Crystal 1997) ESL is sought after by fair proportion of low-income populations in Third World regions (e. g. Clegg, Ogange & Rodseth 2003, Faust & Nagar 2001, Kapadia 2005) Education: medium of instruction in further education Economic opportunities: rural BPO, government, MNCs Computer literacy: ~80% of Internet content Social status: membership in upper classes 2 This material is based in part upon work supported by the National Science Foundation under Grant No. 0326582.

Case for Out-of-School Learning Schools in developing countries have limited impact Shortage of qualified ESL teachers, communicated with us through interpreters In India, non-attendees comprise 43% to 61% of school-going age children (Azim Premji Foundation 2004, NFHS II and Tilak 2000). 15% to 43% cite lack of interest in studies 13% to 31% cite need to work in fields or home 3 This material is based in part upon work supported by the National Science Foundation under Grant No. 0326582.

Case for IT and Educational Games Student motivation and learning (Jenkins 2005) Videogames can incorporate good learning principles (Gee 2003) Longitudinal randomized experiment: 2 years, >10, 000 urban slums students in India (Banerjee et al. 2005) Collaboration b/w MIT and the NGO Pratham Played math computer games twice per week Significant gains in math test scores 4 This material is based in part upon work supported by the National Science Foundation under Grant No. 0326582.

Krashen’s Influential Theory of L 2 Acquisition-learning hypothesis Natural order hypothesis Monitor hypothesis Comprehensible input hypothesis Affective filter hypothesis 5 This material is based in part upon work supported by the National Science Foundation under Grant No. 0326582.

Panchatantra Indian equivalent of Aesop’s Fables Can be digitized into video clips Demo (vocabulary teaching phase) Demo (digital story phase) Q: is this acquisition or learning? 6 This material is based in part upon work supported by the National Science Foundation under Grant No. 0326582.

Ladybird’s Key Word Reading Scheme Peter and Jane series of books Words are introduced and then repeated 12 words make up ¼ of all English words that we read and write 100 words make up ½ of all English words that we use in a normal day 300 words make up ¾ of our verbal output 7 This material is based in part upon work supported by the National Science Foundation under Grant No. 0326582.

8 This material is based in part upon work supported by the National Science Foundation under Grant No. 0326582.

Krashen’s Influential Theory of L 2 Acquisition-learning hypothesis Natural order hypothesis Monitor hypothesis Comprehensible input hypothesis Affective filter hypothesis 9 This material is based in part upon work supported by the National Science Foundation under Grant No. 0326582.

Discussion Q: What are some feasible sources of comprehensible input? More experienced learners, i. e. Krashen’s idea of the handcrafted book 10 This material is based in part upon work supported by the National Science Foundation under Grant No. 0326582.

Special English Demo Reactions? URL: http: //www. voanews. com/specialenglish/ 11 This material is based in part upon work supported by the National Science Foundation under Grant No. 0326582.

Special English Used in Voice of America radio broadcasts Radio transmissions over low-frequency channels, or downloadable MP 3 file accompanied by text transcript 2/3 the speed of normal speech Core vocabulary of 1, 500 words 12 This material is based in part upon work supported by the National Science Foundation under Grant No. 0326582.

Discussion Q: Limitations with Voice of America broadcasts? 13 This material is based in part upon work supported by the National Science Foundation under Grant No. 0326582.

Paraphrasing Through Repetitions Recall the Panchatantra digital stories? SMIL - Synchronized Multimedia Integration Language Benefits of vector-based graphics over static video clip for mobile devices Storage efficiency Randomization promotes replay value 14 This material is based in part upon work supported by the National Science Foundation under Grant No. 0326582.

Electronic Dictionary Explanation using pictures, native language and/or target language Audio pronunciation Paper printout feature Affordances of paper 15 This material is based in part upon work supported by the National Science Foundation under Grant No. 0326582.

Cameraphone Dictionary Sp’ 06 CS 160 class project by Anand Raghavan et al. Words are explained using photos from local contexts that student can relate to Consistent with personalized dictionary approach by Project Pygmalion and others 16 This material is based in part upon work supported by the National Science Foundation under Grant No. 0326582.

Cameraphone Dictionary Seed with high-frequency words Voice of America’s Special English TV and movie scripts Project Gutenberg Populate with definitions, etc. from Wiktionary 17 This material is based in part upon work supported by the National Science Foundation under Grant No. 0326582.

Vocabulary Teaching and Testing Talk Now! Spanish from Topics Entertainment Demo Reactions? 18 This material is based in part upon work supported by the National Science Foundation under Grant No. 0326582.

Teaching for Transfer RABBIT 19 This material is based in part upon work supported by the National Science Foundation under Grant No. 0326582.

Initiation-Reply Sequences Tell Me More from Auralog Demo Any reactions? 20 This material is based in part upon work supported by the National Science Foundation under Grant No. 0326582.

Discussion Q: What are the limitations with this approach of language teaching? 21 This material is based in part upon work supported by the National Science Foundation under Grant No. 0326582.

Pimsleur Audio CDs Demo (00: 00 to 07: 20) Any reactions? 22 This material is based in part upon work supported by the National Science Foundation under Grant No. 0326582.

Paul Pimsleur Four principles: Organic learning Core vocabulary Anticipation Graduated interval recall Implemented in the old days (1960’s) using cassette tape technology 23 This material is based in part upon work supported by the National Science Foundation under Grant No. 0326582.

Pimsleur Generator Text files Female: Male: Female: Hello. Audio Hello Ma’am files Are you from India. Yes I’m from India. Do you understand Hindi? No, I don’t understand. Oh you understand English. Male: Yes I understand English. Female: You understand very well. Pimsleur Generator MP 3 file Metadata 24 This material is based in part upon work supported by the National Science Foundation under Grant No. 0326582.

Reading Acquisition Oral Language Written Language 25 This material is based in part upon work supported by the National Science Foundation under Grant No. 0326582.

Phonics Instruction Clifford: The Big Red Dog from Scholastic Demo 26 This material is based in part upon work supported by the National Science Foundation under Grant No. 0326582.

Phonics Instruction Reader Rabbit 27 This material is based in part upon work supported by the National Science Foundation under Grant No. 0326582.

More Phonics Instruction Reader Rabbit 28 This material is based in part upon work supported by the National Science Foundation under Grant No. 0326582.

Book. Box Commercial spin-off from Same Language Subtitling Demo (~ 7 minutes) Q: What are its strengths and limitations? 29 This material is based in part upon work supported by the National Science Foundation under Grant No. 0326582.

Simulated World Who is Oscar Lake? from Language Publications Interactive Demo Others in this category include DARPA’s Tactical Language Training System Prohibitively expensive to develop 30 This material is based in part upon work supported by the National Science Foundation under Grant No. 0326582.

Summary People learn a language through acquiring comprehensible input Contextual inferencing Extralinguistic context Paraphrasing and repetitions Challenge: how can we create comprehensible input without incurring prohibitive content development costs? 31 This material is based in part upon work supported by the National Science Foundation under Grant No. 0326582.