Post on 03-Jan-2016
transcript
Tom CobbUniversité du Québec à Montréal
Vocab Symposium - Getting the Word Out
AAAL April ‘07
“Vocab learning in a video game?”
3
4
5
6
7
Context (1)
While much is known about making vocab learning + growth efficient
Little of it has an effect on language learners
Who proceed in largely hit-n-miss fashion
8
Context (2)
Word Knowledge has many interlinked facets
1. Meaning(s)
2. Phonological form
3. Orthographic form
4. Syntactic combinability
5. Collocative combinability
6. Procedural (access speed)
Initial acquisition points are 1, 2, or 3
9
Context (3) Pop culture
Spelling-bee mania of recent years Films etc
Education culture Decline of spelling in age of SMS + spell-
checker Obvious impact on writing Subtle on reading
Gaming culture Gamers target learning market
Some seriously Huge potential to get research “out”
10
Spelling games- a non-learning approach
Game presents endless streams of unusual words
Exclusive focus on form
Patterns with little generalizability
Words with minimal communicative value
Only big L1 lexicons can play
11
Spelling games- a learning approach
Game words at player growth edge
Spelling patterns generalize
Game skills = language-use skills–Form cued by meaning–Access speed rewarded
Words with communicative value
Any size lexicons can play• L2 included
12
Desiderata of a learning approach
1. Entire lexicon scaled by level
2. Game-compatible testing
3. Semantic based cueing
4. Recycling of inputs
5. Incorporate access speed
6. Incorporate social nature of language
7. Provide motivating interactions
13
1. Known lexicon categorized by level
From corpus get frequency list No problem since 1960s
But unlemmatized = unuseful Except GSL + AWL
• Make VocabProfile possible
Solved by Nation (06) BNC corpus 14 lemmatized k-lists
Cobb (07) carries job up to 20k The fringe of adult educated lexicon
• Goulden, Nation & Read
14
15
RECENTLY AVAILABLE
100 million word Brit. Nat Corpus (BNC)
=> 17MB word list 694,807 different
word types
GAME NEEDS 20 lemmatized lists
by k-level 3000 4000 5000 …
16
NATION’S WORK
14 lemmatized lists To drive
Vocabprofile program
MY WORK EXPAND LISTS TO
15-20 lemmatized lists by k-level
15,000 16,000 17,000 18,000 19,000 20,000
17
18
19
20
21
22
23
24
25
2. Game compatible level testing Meara et al, Yes/no Checklist
Many words in short time PNW’s as honesty check
PNW’s for other languages Duyck’s WordGen Website
• http://users.ugent.be/~wduyck/Wouter%20Duyck/wordgen.html
Some issue of size…
26
27
3. Semantic cueing Word meanings integrated
• As help / reminder• As game element
From high grade L2-learner dictionary E.g., Cambridge Advanced Learners
Dictionary (CALD)
Whose definitions blend simple language with fairly low frequency words
• (15k plus)
28
29
4. Principled recycling of both hits and misses Classic memory research
Pimsleur et al in psychology Mondria in L2
Forgetting is rapid at first, then slows Geometric progression
This information can be used to prevent forgetting
Retrieval practice at forget-points• Expanding intervals• Goal: minimum number of retrievals to LTM
30
31
32
5. Lex access-speed as game component
Research shows lexical access can be trained
Snellings, van Gelderen, de Glopper (2002)
So, progressively reduce time for game completion
According to power-law algorithm
33
6. Focus on social nature of language learning / use Wireless capacity of DS players
Up to four players Competition, collaboration…
34
35
7. Provide motivation to persist
Language consultant provides content + learning theory
Game professionals provide game elements
Exploit DS’s potential• Sound• Stylus• Split screen
36
Extension projects
Transport game concepts to other languages
Transport game concepts to other language interfaces
French ESL learners Spanish ESL learners Japanese ESL learners…
37
Further reading Frequency lists
– Leech et al, Word Frequencies in Written & Spoken English
Lemmatization procedures– Nation
20k as size of adult educated lexicon– Goulding, Nation & Read (1990)
Yes-No Test– Buxton & Meara (1987)
Spaced recycling– Mondria & Mondria-Wit de Boer (1993)
Reaction-time & practice– Snellings, van Geldeen, & de Glopper (2002)
Easy and hard spelling– Connor (c.1986), N. Ellis (c.1996), Cognitive processes
in spelling
38
Goulden, R., Nation, P., & Read, J.(1990). How large can a receptive vocabulary be? Applied Linguistics 11, 341-358.
Meara, P., & Buxton, B. (1987). An alternative to multiple choice vocabulary tests. Language Testing 4, 142-154.
Nation, P. (2007). How large a vocabulary is needed for reading and listening? Canadian Modern Language Review 63 (1), 1-12.
Mondria, J.-A. & Mondria-De Vries, S.(1993). Efficiently memorizing words with the help of word cards and 'hand computer': Theory and applications. System 22, 47-57.
Snellings, P., van Gelderen, A,, & de Glopper, K. (2002). Lexical retrieval: An aspect of fluent second language production that can be enhanced. Language Learning 52 (4), 723-754.
cobb.tom@uqam.ca
www.lextutor.ca/aaal_07
Stay tuned for upcoming developments…