This paper aims to contribute to the concept of educational governance (Altrichter et als., 2007) from a knowledge-analytical point of view. Recent developments illustrate that the educational system is not solely steered by educational actors exclusively (Kade et als., 1991; Fuchs et al., 2000). The latter are rather backed up by diverse ‘external’ actors, e.g. from economy, civil-societal organisations etc. This development becomes evident e.g. in conjunction with the implementation of the worldwide UN-Decade “Education for Sustainable Development” (2005-2014): Heterogeneous actors get involved in order to implement the overarching objectives of the Decade into all educational fields of the respective national educational system, e.g. adult education, higher education, schools etc. By their integration these actors can be considered informal policy-makers. The observation of an increasing ‘intermingling’ of previously supposed ‘homogeneous’ actors is object of the concept of educational governance. Theoretically, it starts out from the assumption that the difference decreases between those who steer and those who are steered , i.e. that the difference between policy-makers and their public addressees / audience (Mayntz, 2005). This would require that in such constellations actors are able to find a joint language or rather a commonly shared knowledge they may refer to. These actor networks might be understood as transactive knowledge systems, then (Brauner, 2006). Within these, meta-knowledge (i.we. knowledge about one’s own or others’ knowledge) and object-knowledge (i.e. concrete own declarative knowledge) needs to be coordinated in order to initiate or run policy measures. It remains unclear, though, how this knowledge is generated which is meant to be suitable for solving a common policy issue: How do heterogeneous actors in the educational system perceive and proceed a policy implementation, and how do they generate the knowledge necessary to answer this, i.e. how do they translate the knowledge incorporated in a policy request? Which knowledge sources are applied? What differences can be observed between several heterogeneous actors, and why? How does this interfere with the development of concrete policy measures? How is a policy issue ‘refracted’ by different constellations of heterogeneous actors? The paper aims to address these issues by a qualitative empirical study. It refers to the implementation of the UN-Decade in Germany, exemplary. The considerations are theoretically embedded into the aforementioned concept of educational governance on the one hand, and in knowledge sociological approaches (Keller et als., 2006), on the other.

The analysis is oriented towards the field of qualitative social sciences and refers to a programme of a knowledge sociological informed discourse analysis (Keller, 2008). It aims to reconstruct and trace social processes of knowledge generation and use. Methodologically, the analysis is based on the grounded-theory-methodology. The analysis is based on publicly available written material, particularly on protocols and interviews of working groups contributing to the UN-Decade ‘Education for Sustainable Development’ in Germany.

The results show that some constellations reproduce traditional patterns of top-down implementation while others, compared to these, rather pursue 'experimental' paths. Because of the comparative description of the varying styles of knowledge/knowing context-specific statements concerning the translation of the innovation request towards action are expected. These will be deliberated against the aforementioned concept of educational governance. It is intended to explore the meaning of the form of knowledge translation towards action co-ordination within constellations of heterogeneous actors. The alienability of the results shall be discussed with regard to other European educational policy-processes.