Katy Huff (one of our Software Carpentry instructors) has a great birds-of-a-feather talk on attracting diversity in the scientific python community up. She frames as an “intersectionality problem”, the notion of when you have two or more competing diversity problems (say, women in science, or women in science that code) to a double compounding of the problems.

Huff goes on to further explain: “The feature by which you’re both a rainbow and a unicorn, have all the features of social discrimination or social privilege because of your rainbow-ness. And also the same feature because of unicorn-ness but adding these two in a linear combination isn’t the same as having the experience of rainbow unicorn.”

Wikipedia describes intersectionality as “the study of intersections between forms or systems of oppression, domination or discrimination. An example is black feminism, which argues that the experience of being a black female cannot be understood in terms of being black, and of being female, considered independently, but must include the interactions, which frequently reinforce each other.

Take science for example, where there are roughly around 20% women (though certain disciplines are higher and lower). Computing in industry has something like 20% women. SciPy, Huff says, doesn’t have 20% women. These problems compound one another to ill-effect. And it’s not just women, but other minorities and groups in scientific computing.

I really like Huff’s take home recommendation to the audience. Lower the barrier to entry, but not the standards, because as Huff aptly puts it, “that’s offensive.” And as one of the panelist suggests, be careful about micro-aggressions. But that’s a separate post.

Have a watch.