I was super excited yesterday by this tweet 

 The Woodchipper topic modeling demos contrast Austen and Byron’s Don Juan: 

The explanation of the usages  of Woodchipper, created by the folks last year at Corpora Camp, validates my methodology of modeling to generate hypotheses that then require a return to the text to explore more fully. 

I also woke up to find that Jonathan Goodwin just whipped up some topic modeling of Signs.  His abilities put my efforts to work with JSTOR to shame, but I’m so happy to see that what I want is possible.  Now I just need to learn how to manipulate the data from JSTOR to use with Mallet. 

Meanwhile, yesterday I found yet another feminist art manifesto. Rita Mae Brown A Manifesto for the Feminist Artist appeared  the Furies in 1972 

Most frequent word list 


Art certainly not an unexpected result, but people and male are interesting.  Looking at this bottom four words, I’m definitely expecting to find a document that is about how men control women and screw them over in the art world.  I try to use the keywords in context for women, but get an extreme error message ok so on to male, which works, revealing only four usages, so I can read them all, and they are all negative.

The collocate clusters* for women and male, which confirm my expectations.  Women is surrounded by rape, control, revolt, and sex although also artists. 

In terms of my hypotheses yes, women is in the top 10.  Yes  the presence of male indicates that document should focus on critique of sexism/women’s oppression, which it does.   Presence of “control” furthers that line. 

In terms of my questions, especially the one reached yesterday 

 do documents that claim the label “manifesto” look a particular way when viewed with digital history tools?    I’m not sure yet. 

*I finally had some time to play around with the additional documentation for the tools.  Collocate clusters explained very nicely under #4