Just in time for International Women’s Day, Mar. 8, a recent study—reported this morning by Popular Science—has found that some of the most respected authors in literary history don’t exactly treat men and women equally. Well, at least not when it comes to pronouns.
The survey was conducted by Ben Schmidt, a graduate student at Princeton who studies intellectual history (and the brilliant mind behind Prochronisms, a site that examines the time-period-appropriateness of language used in movies and TV). Schmidt’s blog “Sapping Attention” focuses on “text mining and digital humanities”: he scans texts for elements that can tell us something about the context, rather than just the content, of the work. For his most recent post, Schmidt examined works by some of the most well-respected “canonic” authors—the top 100 whose work is in the Bookworm database of 19th-century literature—to determine how many times they used male versus female pronouns, cross-referenced against the author’s gender.
What did he find? Charles Haddon Spurgeon, a 19th-century British preacher, used nearly 20 male pronouns for each female one, the most unbalanced ratio on the list. On the other end of the spectrum, Jane Austen had about three female per male. But the chart—nicely presented at Schmidt’s site—is not even on both sides. Shakespeare is presented as the average, but his work contains about three male per female. Austen is one of only four authors to use more female than male. Almost every writer who refers to females more frequently is a woman. To generalize, women wrote about women and men, but there were more male writers and they mostly wrote about men—which means women’s stories are disproportionately excluded from the literature we respect the most.
In addition, the women who do write are often restricted to less well-respected library shelves. When Schmidt breaks it down by genre, we see that the pronouns can indicate more than just to whom the author is referring:
…that Louisa May Alcott is a juvenile writer and James Fenimore Cooper a literary writer presumably has as much to do with gender as with content…. That highlights one of the most important things to keep in mind when doing this sort of work: that it’s a convenient abstraction that gender and classification might be two different things, but in fact gender is so important that we have to assume it bleeds over into other form of classifications. This is one of the great problems with using established libraries (because they are “biased”) but also one of the opportunities, because we should be able to see some of the ways bias is instantiated into categories.
But Schmidt’s sources aren’t modern—and the canon can always evolve. Who knows what the same study on an International Women’s Day of the future might show?