Artificial intelligence read an astonishing 3.5 million books and found that women tend to be described in terms of their physical appearance, while men – in terms of their behavior.
A whopping 11 billion words were extracted from English language nonfiction and fiction books published between 1900 to 2008 and then run through an AI machine to search for gender-related statistical trends.
The research was done by the computer science department at the University of Copenhagen. The dataset was based on Google Ngram – an online search engine that charts word frequencies from a wide collection of printed sources published between 1500 and 2008.
Excerpts selected for the research included gender-specific nouns (such as sister/brother, actor/actress), as well as adjectives used to describe them (such as “sexy” or “just”).
The adjectives were later divided into categories based on sentiment (positive, negative, or neutral) and their semantic categories – behavior, body, feeling, and mind.
As it turns out, positive adjectives used most often to describe women include complimentary “beautiful”, “gorgeous”, “sexy”, “classy”, and pejorative “untreated”, “underweight”, and “shrewish”. Those defining men, on the other hand, include “just”, “sound”, “righteous”, and “unsuitable”, “unreliable”, “brutish”, accordingly.
As visible, females are usually described in terms of body and appearance, while males are defined by adjectives that refer to their behavior and personal qualities.
Moreover, negative verbs associated with body and appearance appear some five times as often for female figures as male ones.
But are the results relevant? Isabelle Augenstein of the University of Copenhagen explains: “If the language we use to describe men and women differs in employee recommendations, for example, it will influence who is offered a job when companies use IT systems to sort through job applications.”
While the research is not groundbreaking – gender language and bias is quite a popular topic in linguistic studies, and linguists have long suspected such results – it is the biggest one up to date.
Via World Economic Forum.
• • •
Don’t miss other lists and posts: