A new study out of USC Viterbi School of Engineering analyzed 3,000 English language books contained in the Gutenberg Project, which are books that are in the public domain. This included novels of all genres as well as short stories and poetry. They used natural language processing, an AI technology, to determine whether the characters were male or female — this process is not yet able to identity “non-dichotomous” characters, but the researchers expected to not find non-binary characters in the public domain books used as the sample.
They found that male characters are four times as likely to be found in literature than female characters, though this difference shrinks when looking at only books by women authors. Mayank Kejriwal, the research lead on this project, noted that “Gender bias is very real, and when we see females four times less in literature, it has a subliminal impact on people consuming the culture.”
They also noted that female characters were associated with adjectives like “‘weak,’ ‘amiable,’ ‘pretty’,” while male characters were described with words like “‘leadership,’ ‘power,’ ‘strength’ and ‘politics.’”
You can find out more about the study at the USC Viterbi website, or you can read the study in full.
Find more news and stories of interest from the book world in Breaking in Books.
Source : 4x More Male Characters Than Female Characters in Literature, AI Study Finds
