Professor Madhav Gadgil gave a talk at IIIT Hyderabad today. Among other things, he stressed the need to empower local communities with information on and knowledge of how various decisions by governments and corporations will impact their environment. Government-generated environmental information is poor, not-transparent, and frequently manipulated, he said. The Biotechnology Regulatory Authority of India Bill, 2013 came in for special castigation (see also this analysis of the bill by PRS Legislative Research).
He was tremendously optimistic about citizen-science initiatives (like the Australian WaterWatch) and using the Right to Information Act to generate knowledge for informed debate. He was equally optimistic that machine translation "within about 30-40 years" will make this knowledge available in any language.
In the Q&A that followed, I took the following line (spelt out here in greater detail than I had time for in the session):
(a) biodiversity hotspots are also hotspots of linguistic diversity -- researchers in fact speak of "biocultural diversity";
(b) knowledge of biodiversity maintenance is typically encoded in indigenous languages, and many of these are very small languages indeed -- the median number of speakers of the world's languages is a mere 7000 (compare it to the mean -- 878 thousand);
(c) thus, even from a purely instrumental point of view (setting aside any ethical arguments), there is a good case for the flourishing of these languages;
(d) but this linguistic diversity is disappearing very fast -- "between 1970 and 2005 the number of languages spoken globally has decreased by 20%";
(e) among the most effective mechanisms to ensure that a language disappears is an assimilationist language policy in education -- specifically, the use of a non-mother-tongue as medium of instruction -- what we have elsewhere called "silent ethnocide -- a low-intensity warfare through formal education";
(f) thus, if information and knowledge are to benefit the environmentally most vulnerable in society, a mother-tongue-based multilingual education is necessary -- machine-translation technologies, while welcome, will not by themselves be enough.
He was tremendously optimistic about citizen-science initiatives (like the Australian WaterWatch) and using the Right to Information Act to generate knowledge for informed debate. He was equally optimistic that machine translation "within about 30-40 years" will make this knowledge available in any language.
In the Q&A that followed, I took the following line (spelt out here in greater detail than I had time for in the session):
(a) biodiversity hotspots are also hotspots of linguistic diversity -- researchers in fact speak of "biocultural diversity";
(b) knowledge of biodiversity maintenance is typically encoded in indigenous languages, and many of these are very small languages indeed -- the median number of speakers of the world's languages is a mere 7000 (compare it to the mean -- 878 thousand);
(c) thus, even from a purely instrumental point of view (setting aside any ethical arguments), there is a good case for the flourishing of these languages;
(d) but this linguistic diversity is disappearing very fast -- "between 1970 and 2005 the number of languages spoken globally has decreased by 20%";
(e) among the most effective mechanisms to ensure that a language disappears is an assimilationist language policy in education -- specifically, the use of a non-mother-tongue as medium of instruction -- what we have elsewhere called "silent ethnocide -- a low-intensity warfare through formal education";
(f) thus, if information and knowledge are to benefit the environmentally most vulnerable in society, a mother-tongue-based multilingual education is necessary -- machine-translation technologies, while welcome, will not by themselves be enough.
No comments:
Post a Comment