Departing from a social constructionist understanding of gender, this thesis examines how the female entrepreneur is constructed in research articles about women’s entrepreneurship. It finds that even if the texts celebrate women’s entrepreneurship, they do it in such a way as to recreate women’s secondary position in society.

Building on Foucault’s theory of discourse, the thesis analyzes the discursive practices by which this result was achieved. These practices include certain assumptions that are taken for granted about women, men, business, work, and family. One of these assumptions is that men and women must be different. Despite research results to the contrary, many texts insist that the genders are different and construct three kinds of arguments in support of this. One is making a mountain out of a molehill, i.e. stressing small differences while ignoring similarities. Another is the self-selected woman, which proclaims women entrepreneurs as unusual women. The third is called the good mother and consists of molding an alternative, feminine model of entrepreneurship while leaving the dominant model intact. These arguments reproduce the idea of essential gender differences and the idea of the woman as the weaker sex.

The discursive practices also include certain ontological and epistemological assumptions, which are questioned in the thesis. In addition, they contain disciplinary regulations as well as writing and publishing practices that reinforce the discourse. The practices and the ensuing research results are moreover dependent on the particular context in which the articles are produced. This means that their results and conclusions cannot be transferred to other contexts uncritically.

By discussing these practices, the thesis opens the way for alternative ways of theorizing and researching women’s entrepreneurship. Suggestions for alternative research practices include the addition of institutional aspects to the research agenda, such as labor market structure, family policy, and legislation. The thesis also suggests a shift in epistemological position – from gender as something that is given, to gender as something that is produced.