Thursday, December 3, 2009

Gender and POV

So apparently I am not done with my NaNo work because it seems to still be sucking up a lot of my brain power.

Tonight's procrastination exercise (posing as pseudo-experiment) was with "The Gender Genie," a fantastic little gizmo that analyzes passages of text and predicts whether the author is male or female based on the words used.  My NaNo work started with three different points of view (two men and a woman) that would alternate by chapter, so I thought it would be interesting to see whether my male points of view sounded like men, and whether my female point of view sounded like a female. 

To run the test, I pulled 500-550 word samples of dialogue and internal reflection from each POV, then put the entire POV into the system.  I also ran all four of my previous blog posts through individually, as well as a memo I wrote for work (non-fiction) and the U.S. Constitution, just to see what the Genie would say.  The Genie actually has different settings for fiction, non-fiction, and blog entries, so I wanted to test out each setting. 

First off, all of my characters, when looked at overall, came off as women.  Not promising since, as I mentioned, two were men.  The only one of the samples I ran for the NaNo work that came back as male was the main character's internal relection, so at least I've got something going for me.

More interesting were the results from my blog, my memo, and the Constitution.  My memo came back as male, and so did my two of my four blog posts as well as all posts lumped together.  This surprised me because, as you may have noticed from my picture, I'm a chick.  Apparently, my own inner dialogue is more male than one of my male POVs?  How disturbed should Hubby be about this?

But the best was the U.S. Constitution.  I put the text in, both with and without the signatures at the end, and both times it came back as being written by a FEMALE.  This makes me smile.  :)

Now I'm curious as to whether certain styles of text are more likely to be male or female.  Is it that more descriptive text tends to be female and more action-oriented text tends to be male?  Or is it really the nouns versus relationships idea that the New York Times article that inspired the Gender Genie says?  I'm going to pick up a few samples at the local library and run them through to see whether this is true, and also whether I'm unusual with my fiction-as-a-woman, non-fiction-as-a-man. 

If you try this for yourself, I'm curious to see the results!

1 comment:

  1. Alissa,

    Thanks for joining my blog. :-) This Gender Genie thing sounds totally fascinating. I will have to experiment with that and see what results I get.

    ReplyDelete