Blog Post

Feminist Text Mining and "Big Data"

Feminist Text Mining and "Big Data"

        I was surprised that a feminist digital humanist, Lisa Rhody, would be interested in text mining of “big data” because of the problematic assumptions behind text mining itself. Rhody brings up one of them: that text mining of “big data” currently relies upon positivistic claims of objectivity. The Western traditions of scientific positivism and objectivity have historically neutralized the unneutral. Most feminists would agree that there are no universal “truths” because all studies and interpretations of data and information are mediated through social location. In response, Rhody suggests that this is precisely the reason that feminists should “dig:” so that they can fundamentally transform text mining of “big data” into a socially conscious activity aware of social location (e.g. by reversing the erasure of “small” words like pronouns that help convey information about social location).

        Yet, this begs the question: should text mining of “big data” be used for feminist purposes? This reveals a second problematic assumption behind text mining. “Big data” necessarily requires the flattening of texts. Regardless of tools that incorporate some aspects of social location, the practice of textually analyzing “big data” necessitates the erasure of some aspects of localized nuance and specificity. As a study, this would likely corrupt feminist textual analysis by encouraging universalist interpretations of literature. Then again, Rhody provides a compelling response by arguing that feminist literary critics should use computational text analysis to “embrace grand technical and social challenges through theorization and praxis, incorporating moments of productive discomfort.” In other words, the study of a litany of texts does not necessitate the erasure of social location. In fact, it can be provide for a moment of departure from the search for neat, objective truth. This was a source of reflection for me because it forced me to challenge the assumption that “big data” is equivalent to social erasure. 

Reference: Lisa Rhody, "Why I Dig: Feminist Approaches to Text Analysis," http://dhdebates.gc.cuny.edu/debates/text/97

109

No comments