Crowd-sourced Technical Texts can help Revitalise Indian Languages
Paper i proceeding, 2018
concepts, however elementary, ignoring the STEM vocabulary in IL textbooks up to high-school. People assume English is
necessary, and ILs are unfit, for STEM and higher education generally. English and STEM competence also mark wealth,
so parents now abandon first language (L1) schools for often woeful “English” ones even at primary level. So children learn
everything poorly: L1, English and content. To reverse this collapse, people need to use L1 more broadly. This paper calls
for IL STEM texts, crowd-sourced from STEM-trained IL-speakers, to seed such usage. We note how the texts would fit
in the linguistic landscape. They would also be important new data for computational linguistics. STEM-trained people
with rusty L1 writing, like us, will find that with the dictionaries and text online, they can write in L1—we comment on
vocabulary and help from related languages. Crowd-sourced texts vary in quality, but they can help people to use L1 for
STEM topics, and to realise that children learn content better in L1 than in bad English.
Indian languages
crowd-sourcing
translations
STEM
Författare
K V S Prasad
Chalmers, Data- och informationsteknik, Funktionell programmering
Shafqat Mumtaz Virk
Chalmers, Data- och informationsteknik, Datavetenskap
Miki Nishioka
Osaka University
C. A. G. Kaushik
Volvo Group
Language Resources and Evaluation
1574-020X (ISSN)
Vol. 2018 11-16Miyazaki, Japan,
Styrkeområden
Informations- och kommunikationsteknik
Drivkrafter
Hållbar utveckling
Ämneskategorier
Språkteknologi (språkvetenskaplig databehandling)
Jämförande språkvetenskap och allmän lingvistik
Studier av enskilda språk
Lärande och undervisning
Pedagogiskt arbete