Experiments

Sometimes I make stuff on my computer. If I’m feeling brave, I’ll rope my students into making stuff with me. They always surprise me with the beautiful things they create. My stuff, on the other hand, is never rarely pretty, logical, or useful.  Often, it doesn’t even work properly, or at least not the way I thought it would when I set out to make it.  In this respect the stuff I make tends to fail both aesthetically and functionally. That’s ok.  I mostly make stuff to learn how stuff gets made or to learn more about how texts get read.  If you are interested in making stuff, too, for whatever reason, here are a couple of places you might enjoy: 1: DiRT (Digital Resource Tools) and 2: Digital Toy Chest for Humanists.   At any rate, here’s my stuff! (If it’s password protected it’s still in the oven.)

Text Palettes

I’ve been working on a few projects that put digital technology in the service of close reading literary texts. This first one is called “Text Palettes,” a simple Python script that reads a text for its colors and then outputs a simple pie chart of the color distribution. It is limited and reductive, but in some ways this is the point. I will be refining this and the others for my “Literature by the Numbers” class next semester and am interested to see how it will work as a tool for close reading in the classroom.

This is a text palette of Edgar Allan Poe’s “The Black Cat.” Not surprisingly, the color black dominates.
black_cat

python resources

I am learning python to work with texts.  I’m not a programmer, but since I want to be able to do simple things that a computer can do way faster than I can (like, for example, count words), I am picking up some new skillzzzzz.  Here’s what I’ve been using:

tutorials
https://www.codecademy.com/learn/python
(+) free, thorough, and straightforward.  The forums are helpful and when you get stuck there are several solutions on github.
(-) sometimes there are concept jumps that aren’t explained, but the Q&A forums were super helpful for this.  This course might benefit from extra breakout lessons/projects that illustrate key concepts.  They seem to have a few of these if you pay for the pro upgrade, but even these don’t seem to be very thorough.

https://www.udacity.com (Intro to Computer Science)
(+) free, thorough, and you make a web crawler! teacher is great and does a great job breaking things down
(-) no complaints. Although sometimes it’s both overly broad and a bit *too* thorough, and you have to code using their online interpreter, which sometimes doesn’t seem to be running.

http://learnpythonthehardway.org/book/
(+) free, thorough, and all done through the command line and TextWrangler (also free).  Plus, the author is kind of mean and hilarious in a great, angry gym coach kind of way.
(-) no complaints–not enough text-based examples for me personally, but that’s not the purpose of the book.

http://nealcaren.github.io/python-tutorials/
(+) free, a great list of projects to try
(-) geared more towards social scientists than humanists

text analysis
http://nealcaren.web.unc.edu/an-introduction-to-text-analysis-with-python-part-1/
(+) free, quick, fantastic.  Exactly what I was looking for.  While slogging through the Python courses listed above, I despaired that I would never learn what I wanted to learn.  I feared I would need to learn the whole dang language to do some simple text analysis.  This changed everything! I learned in ten minutes how to do sentiment analysis from the command line.  Neither of the other tutorials really touched on this. THANK YOU.
(-) None.

http://nbviewer.ipython.org/url/norvig.com/ipython/How%20to%20Do%20Things%20with%20Words.ipynb
(+) free, quick, and this:

(-) maybe a bit too quick.

beginner projects I want to try but haven’t yet had time:
https://www.reddit.com/r/beginnerprojects
https://www.thinkful.com/learn/intro-to-python-tutorial/#Conclusion-and-Resources

Five mini programming projects for the Python beginner


solutions to the above:

Other useful resources:
Python for beginners:
http://www.pythonforbeginners.com/files/reading-and-writing-files-in-python

Practice Python:
http://www.practicepython.org/

“Axolotl” in Voyant

Very pleased by the ease of Voyant as a tool for text analysis!
I used Julio Cortázar’s “Axolotl” and was excited that the tool allows you to make a word cloud, easily apply stop words (and edit these if you need to), and visualize frequency by simply checking a box, *all on the same page*. (In the screenshot below, I decided to check words that suggested transparency (aquarium, tank, glass) to see where and how often these appeared in the text.)   I will definitely use this in my classes this year–am looking forward to it after piecemealing it with Wordle …

voyant-axolotl

Elegy for a Dead World

As soon as I saw this beautiful game I wanted to play it in my classes. Fortunately,  Steam was generous to give us enough keys to try.   I was personally more interested in getting lost in the visuals than I was in writing, but my students loved it and produced excellent work.  This game is gorgeous, but I recommend it for a more practical reason as well: versatility.  We played in in my Creative Writing SF class, but I now want to try it with Literature and the Environment, Literary Theory, and—if I ever get to teach it—a class on Experimental Narratives. It could work for any course that emphasizes composition, rhetoric, and creative expression.  A big thumb’s up.

Continue reading Elegy for a Dead World

Kafka in CATMA

I’ve been wanting to do more with markup but don’t have the time I would need to do it in XML.  So I heard about CATMA (Computer Aided Text Markup and Analysis) and have been playing around with it.  It allows you to load a text and tag it up according to your own logic.  I picked Kafka’s “Metamorphosis” because I am curious about the way transformation works in the story. The title of the story suggests change in general and Ovid’s poem, in particular.  But Ovid’s poem is filled with marvelous transitions and transformative sexual encounters, all contained within a highly mobile text that leaps from character to character with facility.  In Kafka’s story, on the other hand, the change is painfully singular and ironically (paradoxically? satirically?) a result of societal and personal stagnation. The story’s form, too, is excruciatingly confined (we remain trapped with Samsa in his bedroom for the duration).  But even so, a transformation has occurred, and I am curious about how the words about change that appear in the text might be linked to the way that Samsa’s roach-body gets revealed, in parts, leg by little leg.  So I went through the first section of the story and tagged all word-level instances of change, bodies, and body parts.  I don’t have any conclusions yet, but it was cool to see what my tags revealed in terms of frequency and distribution. (Lots of little legs.)

Robot Evans

Robot Evans is the name of my Avatar in Second Life.  I didn’t “make” Robot Evans so much as I customized her from the template.  With that said, Robot Evans made me aware of both the limitations and potential of these pre-packaged (but malleable) bodies and helped me think through some ideas I have about self representation online.  Many thanks to my friends in Second Life for the stylish tuxedo and top hat.

Here is an interview with her, conducted in-world.

Dragon Balls!

I made the game DRAGON BALLS! in Scratch to accompany my husband’s awesome web comic, ODIN AND FRIENDS, which is also my favorite web comic, and not just because it’s by my husband. You can blast some of Odin’s enemies here:

Notes: Scratch was fun and easy to get started with, but I was useless when it came to the math involved in timing the speed of fire, so I used the pieces of another scratch game to help. Also, all of the images in the game come from ODIN AND FRIENDS or other projects. In other words, I assembled pieces of other people’s stuff to make DRAGON BALLS! work.

baby

It’s no wonder that Robert Coover’s “Babysitter” is often invoked as a hypertext precursor.  The story is filled with recursive text chunks, overlapping timelines, and shifting centers of perspective, all of which destabilize a clear linear progression and  make the story a convincing precursor to hypertext fiction.  But what if you turned the text of Coover’s story itself into hypertext?  What if you linked to different characters within the story and followed these links?  I conjectured that, if you did so, you might find that hypertext, in this case, would make a more linear story out of what is a nonlinear print-based text.  To test this conjecture, I took a section of Coover’s text and linked all the instances of the word/character “baby.”  You can check it out here to see how it reads.