Rewiring the Bechdel Test: Experiments in Computational Feminism

Convenors: Dr Anne Alexander, Director of Learning, and Ryan Heuser, Research Fellow in English and Digital Humanities, University of Cambridge.

Rewiring the Bechdel Test: Experiments in Computational Feminism

A Guided Project organised by Cambridge Digital Humanities

Bringing together researchers from across disciplinary and non-technical backgrounds to conduct a shared experiment of ‘computational feminism’. Specifically, we will work to model, analyse and reimagine the ‘Bechdel test’, a simple and popular measurement of gender inequality in fiction.

The test – named after cartoonist Alison Bechdel, who first formulated it in “The Rule” from her comic strip Dykes to Watch Out For – asks simple three questions of every work of fiction:

1. Does it include two named women?

2. Do they talk to one another?

3. About something besides a man?

The test has inspired much academic as well as popular discussion. It has also been the subject of digital and crowdsourcing-based projects. For years, the Bechdel Test Movie List has gathered 8,000 films along with pass/fail responses from thousands of viewers. Bechdel.io is an open-source script which scans movie scripts for evidence of female character dialogue.

Critiques and responses to the original Bechdel test have adapted this formula to investigate expressions of other forms of inequality and discrimination in cultural media, such as racism, homophobia or transphobia.

In this workshop we will begin by reviewing these previous studies and adaptations alongside the original test, asking what they capture, what they leave behind, and what new avenues remain unexplored.

We will assess whether and how the Bechdel test might also be rewired or computed for novels, short stories, fan fiction – even artworks, comic strips, or social networks, and explore the theoretical and practical implications of using Natural Language Processing (NLP) tools and methods for this kind of work. We will test the capacities of tools such as the Python text processing pipeline booknlp and genderize.io in our efforts to analyse digital culture at scale.

The specific research interests of the participants will guide the subsequent experiments, culminating in a final presentation workshop which will be open to the public.

Tickets:

We invite applications from early career researchers and others at the University of Cambridge and beyond to join a small project team for four one and a half-hour online sessions on 4, 6, 8 and 11 July, followed by a day's in-person workshop in Cambridge 13 July. Early career researchers from Cambridge will have priority for places on this course. However, we welcome applications from researchers from other institutions.

Participants will need to commit to joining the live sessions and to set aside at least 3-4 hours in addition to working on an individual or smaller group project during the course. We are interested in assembling an interdisciplinary group of researchers drawing on insights from across humanities, social science and technology disciplines. Prior knowledge of programming, computer science or Machine Learning is not required. The first four sessions will be online only, with the final workshop in person in Cambridge. Online attendance is also possible for the last workshop.

Places for this guided project are limited and prospective participants must complete the online questions to apply for a place. We will write and confirm your place by email.

Project session dates and times:

Monday, 4 July 13:00 - 14:30 (online via Zoom)

Wednesday, 6 July 13:00 - 14:30 (online via Zoom)

Friday, 8 July 13:00 - 14:30 (online via Zoom)

Monday, 11 July 13:00 - 14:30 (online via Zoom)

Wednesday, 13 July 9:30 - 12:30 and 13:30 - 16:30 (in person in Cambridge)

Image credit: Alison Bechdel