Yesterday I caught a bit of the live stream of #dhtopic from MITH, and let me tell you I feel an ass even blogging about what I’m doing after that. The level of discussion about topic modeling is so vastly advanced. Watch the videos as soon as they are up. Be amazed and hopefully inspired, because, there’s really no turning back once you start (at least in my case).
So on I go with my project of modeling manifestos. This morning the Redstockings Manifesto, according to Jacqueline Rhodes, probably the most anthologized example of the genre from women’s liberation.
Fortunately for me, these manifestos all exist online. The Redstockings, or whomever comprises them at present, have a website. So I grab text, save as .txt and upload to the scrubber at Wheaon Lexomic. Doesn’t work, as in no scrubbed text generated. Hmmm, try again, doh, of course I’m doing this pre-caffeine, and daylight savings time kicked in last night, so I feel fortunate that I realize quickly that I saved the file as .txt but using word instead of text edit as I have in the past. Do again, try again. (damn which auto saved as .rtf, fixed that manually) Try again. Success!
Over to voyant. Hmmm I find my first possible OCR error when cf shows up as the most common word. Either the Redstockings were really into latinate abbreviations (which I do not recall) or something went amiss. Back to .txt file and no, my brain didn’t fail me. No cf in the text. Hmmm back to scrubber where my clearly eyes see that indeed my .txt file isn’t clean. Yikes I have to manually do something (haha this is a digitalhistory joke, as most scholars have to prep their sources way more than the lucky people who study the twentieth century ) Fortunately I pull out the scrubbed text, which is only 338 words, and just deleted the remnants of code. Sigh. As an oral historian I did way too much of different types of transcription back in the day. I have no desire to start transcribing documents again.
Back to voyant, where as I’ve now come to expect, the most frequent word list is topped by women. I’m starting to see a definite category existing of manifestos that speak or are about “women.” Thus far I’ve only had one manifestos that did not have women as the most frequent word, although in the case of Bitch Freeman is using Bitch syntactically where woman or women would go almost as a collocate). Could we use this come to at least a partial definition of feminist manifesto (something the literature reveals is hotly debated and which my article addresses)?
The second most frequent word is “male” which is also not unexpected. I’m starting to see a category of feminist manifesto aimed at attacking male privilege. In fact these result remind me of The SCUM manifesto, so I pull up the excel spreadsheet I’ve been using to export voyant data. (oh wait that’s right, yesterday as I finished up, I wrote over my excel file where I’d exported all this SIGH). Here’s hoping the voyant link still works as it offers the not very reassuring message when you generate the URL that it could be gone tomorrow). Sweet, I actually cut and pasted lists into my prior blog post. SCUM manifesto has “male” as 1st most common word.
More refined research question Is there a correlation between frequency of male and argument attacking male privilege? Hmmm let’s look back at Bitch, and yes, men is lower ( #8) which accords with the emphasis in Bitch on the limitations of femininity, as opposed to emphasizing the role of patriarchy (men) in perpetuating those limitations. OK on to Ukeles’ admitted unusual Manifesto for Maintenance Art where “man” came in at #5. SWEET my voyant link for Ukeles still works. So yes as Ukeles is also making an argument attacking male privilege, in the art world in her case, so far (on a n=2 I know!) the answer to this question seems to be yes.
Since this whole project was motivated by an article I’m writing about stretching the genre of feminist manifestos, what about the documents I use in that piece? Do they conform to these hypotheses? Yes. Women is the most common word in three, woman in one. However, the incidence of male is extremely low or non-existent. The CWLU statement of principles (1969) contains no male referents in the 50 most frequent words, nor does the FSW announcement (1973). More refined research question is there a sub category of manifesto that is programmatic identified by lack of attacks on male privilege? In other words these manifestos take sexism and women’s oppression as a given and are aimed at women who do the same already? If so I’d have expected to see a date trend here, with earlier documents like the CWLU still needing to explain women’s oppression before launching into rationale for program of action. The third source, a broadside announcing the Feminist Studio Workshop an alternative program in the arts, seems to conform. Man is the 14th most frequent word, although it is interesting to note that human shows up at #4. While the founders of the FSW were involved in early efforts to address sexism in the art world, the impetus behind the FSW was to turn their energy away from changing men into developing women.
Looking at the final source, Everwoman, a play from the first night of the founding conference of the CWLU, offered the most important insights from this exercise. Men is #5 most frequent word, which makes sense as I know the play was organized in order to convince socialist women of sexism in the New Left and hence the need for an autonomous women’s groups. Yikes, just realized that in fact my argument should be that the play, not recognized as a manifesto is the crucial context for the Statement of Principles, sometimes recognized as a manifesto.