I came across
this tool today and gave it a few tries. It predicted that Gertrude Stein, Arthur Conan Doyle and I were male based on our writing and that Dan Savage was female. Based on a
paper by Shlomo Argamon et al. suggesting that female writers use more pronouns and male writers more noun specifiers, the Gender Genie analyses the functional words in a passage of writing, assigning different weights to different keywords, and makes a prediction about the gender of the writer. The standard effects of socialization and gender performativity come into play here: women tend to compliment, apologize and talk about relationships; men's speech focuses more on objects and events than the speaker's relationship to the listener.
Assumptions are made that sex/gender (the words are used interchangeably in the paper) is natural, non-transgressive and public, and that gender is a binary. The paper is "progressive" insofar as it asserts that women's use of pronouns is a method of conversational control rather than the effect of subordination. Men's speech is discussed throughout the paper as default: women use relational and referential strategies to "personalize" discourse, while men simply relay new information. There are some interesting discussions of specific texts. Non-fiction examples suggest that although men and women both write in the first person, men use passive constructions and general statements to make grammatical person invisible while women more readily assert it.
Please note that the Gender Genie was based on research by Argamon, Koppel and others but not actually created by them. So, is the Genie more than an interesting conversation piece? Is it important to know the author's gender when reading or discussing a text? Is this tool useful for actually predicting a person's gender when it is unknown, or for discussing gender performance in writing when the author's gender is known, or both?
[Note: This post has a female score of 140 and a male score of 607].