This chapter is based on long-term experiences of language policy development, primarily within heritage and minority language education, in Sweden and Canada. In the case of Sweden, two recent inquiries on national minorities and national minority education are used to summarise the (official) situation for the five national minority languages (NMLs) within the language policy and educational fields. In the case of Canada, the many overviews and critical publications on multilingual education are used in a similar way. The international standards forming language rights, and set up by the Council of Europe (CoE), also work as a frame of reference for the comparison of the possibilities for the heritage and minority languages within the educational systems of Sweden and Canada.

The chapter deals with the lack of fit between Sweden's ratification of international conventions on linguistic human rights for national minorities and their implementation. The author discusses two main aspects of this: recurrent attempts to prohibit the use of national minority languages in public (work, school settings), and the right to education in and of national minority languages. He uses the European Charter for Regional or Minority Languages as one main starting-point, but also discusses Swedish ratifications of UN, EU, and CoE human rights conventions. In addition he contrasts these with the Swedish constitution, the Discrimination Act, in which language as a ground for discrimination is absent, the Law on National minorities and national minority languages, and the School Act. Mismatches exist both between the international conventions and Swedish adaptations of them, and between both types of legal instruments and practice.

This article aims to establish the ways in which different aspects of the Swedish national minority language policy have facilitated or hindered the fulfilling of the aims to revitalize these languages. Finnish is used as a test case, and we apply a two-step approach to analyzing the development of its use during the 2000s: we adapt the COD model (Capacity, Opportunity, Desire), proposed by Grin in the early 1990s, and apply it at three societal levels: macro, meso and micro. The macro level covers national policies, laws and other regulations. The meso level is represented by the adoption and implementation of these locally. The micro level consists of linguistic biographies by informants of a bilingual project started in 1991, and revisited in 2013. Education in or of Finnish as an example of language policy is central to the study, although other fields are considered.

Our analysis shows that policies established at the national level recur in different forms at the meso and micro levels. Although there is an overlap between these societal levels, there is a discrepancy between legislation and practice. The dimension of capacity in Finnish seems to account for much of the development in language use, compared to opportunity and desire. The COD model, in combination with its multilevel adaptation, is a functional, analytic option for the study of language policy decisions and their effect on language use patterns. It therefore could also be adapted to the other national minority languages of Sweden.