Learning in school using Indonesian as L2

A message that is not understood is useless. But education theory tends to be developed in monolingual countries with monolingual blinders. However most of the world's societies are multilingual, trying to “do education”, while ignoring the language factor, resulting in education systems that are not nearly as effective as they could be and interfering with achieving “education for all”. With around 700 languages in Indonesia, as in many multilingual countries, most children start school with the language of instruction being different than the language spoken in the home and community. However, a basic principle of education is to begin from what is already known, to teach what is unknown. Yet when children start school, the language they have been using for years to learn about the world around them is declared invalid, and they are forced to use a second language they only poorly understand. So it is not surprising that large numbers of children are demoralised, drop out and do not reach higher levels of education (in one province, 76% have not gone beyond elementary levels and less than 2% reach tertiary levels). More than 60 years of research in multilingual education show us how we can get children learning better, reaching higher levels of education, learning the national language better, have fewer dropouts and repeats, and have greater participation by the parents and community. One of the keys to education is reading skill. But by forcing children to learn to read in a language they don't speak in the home, that step is massively greater and more difficult than learning to read in one's own language. Three different approaches to multilingual education are described, including discussion of why some poorly implemented bilingual programs are not delivering the expected and desired results. Within the limits of current education policy and curriculum in Indonesia, how can the language factor be leveraged to achieve better educational outcomes for more children?