What Do We Say When We Talk about Women and Science? Framing Problems and Solutions in the EU and US.

Around the world, there is a steady output of reports decrying the gender imbalance among scientists in particular. However, even what counts as "science" is a contested concept and the construction of the social problem of “women’s underrepresentation” even more so. In this paper, we combine quantitative and qualitative content analysis of the discourse defining what the problem is understood to be, to identify variation in frames for intervention offered by different actors to different audiences. We construct a topic model (using LDA) of the reports from between 2000 and 2015 from the US and EU’s task forces and study groups to see the similarities and differences in the words chosen to describe the situation. We also subject a smaller set of 25 full reports (and another 25 executive summaries) to a qualitative analysis of the kinds of problems constituted by gender disparity and for whom these are problems, the approaches suggested for remedying them, and who precisely is to do something to fix them. The quantitative topic model emphasizes how specific terms cluster together. The qualitative model provides a more integrative look at the types of argumentation deployed in when these concepts are invoked, and how problems and solutions are connected in individual documents. We find that EU discourse is more overtly political and addresses educational systems as part of a broader policy commitment, while the US approach is more closely tied to the internal practices of individual universities and integrates gender with other forms of diversity. The role of academic capitalism in the framing of the problem lies more in the style of argumentation than in the choice of specific frames for either problems or solutions.