I began this post by perusing the Project Gutenberg website. After lots of meandering, I decided to try "reading" some important historical literature that I've been meaning to get to. Specifically, I endeavored to soak-in the The Canterbury Tales, The Epic of Gilgamesh, Beowulf and The Odyssey. Below are the results, which I'm calling "Project Gilga-mash."
Starting out, I did not have a clear outcome in mind beyond learning more about these seminal works. Chaucer's was the first of the poems that I came across on Project Gutenberg. After loading it up on Wordle, there were some interesting immediate observations. First, the word God seems to occur just about as often as words like ye, thy, eke, etc... This is probably not too shocking, considering that the characters in this story are a group of religious pilgrims. Other prominent words like, wise, good, right, love and heart help us further understand that this is intended to be a moralizing narrative. As is to be expected, the words man and men are used much more frequently than the word woman. However, the word wife appears to be the preferred female designation. Also, interestingly, the word Chaucer is on the list with moderate frequency. I don't remember the author being a character in this story, so I probably neglected to remove all of the footnotes and acknowledgements.
What most-epic of epic of poems would be complete without the Old English story of Beowulf? I read a good chunk of this in 9th grade English class, and I watched the rest in the recent CGI animated movie. I also enjoyed the novel Grendel by John Gardner. I grew to like that character a lot, and I was a bit dismayed that I could not find the name Grendel anywhere in the word cloud. I understand that I'm compounding a lot of texts, so maybe he was just forced underground (a little Beowulf joke, sort of). The word Beowulf is a bit hard to find, too. He's mentioned less frequently than Gilgamesh, but about the same as Chaucer. God is still doing pretty well, overall.
I thought that I was finished with my incredibly-epic poem, but I had totally forgotten about Homer's Odyssey and Iliad! I decided to add just the Odyssey, so as not to overly-Homerize the word cloud. Immediately, the name Ulysses jumped out at me. He is clearly the winner of the character name mentions. Gilgamesh is still holding on for dear life, but Beowulf and Chaucer appear to have been buried under a Homerific pile of text. In the end, man did slightly better than God. Wife also won out over woman and mother. I didn't realize until the final word cloud that the gender-neutral pronoun one is used about as equally as the gender-specific pronoun man. Maybe this was a more common way to refer to individuals of either gender.
After uploading all the texts into Voyant, I was surprised to see that the pronoun one is, in fact, used more frequently than the pronoun man. Man is at its highest usage with 26.58 per 10,000 words in the Canterbury Tales, where one is at its highest usage with 43.31 per 10,000 words in the Odyssey.

And, here's a DocuBurst with just Beowulf. It kept crashing when I tried uploading the full Project Gilga-mash text. Look, there's Grendel!






Loved Grendel!!! My son had to read it in high school and everything he read, I read. I love the idea of the Gilga-mash. And I love the texts you chose. I'm curious what would emerge if the common words (the old English words "thy," "ye," etc. that are akin to today's "the" and "of" and the like) were stopped as Wordle does with contemporary texts. Still, an interesting mash-up and a fun conglomeration of texts.
ReplyDeleteAri, what an epic posting! :) One thing I'm noticing about myself: I definitely respond better to the graphs and word counts in Voyant. The word cloud does less for me. Maybe because I have always tended to be a text-based rather than visual learner.
ReplyDeleteWhat an ambitious project! Like others in the group, it sounds like you came across the usual difficulties of finding / editing a 'corpus' in Project Gutenberg. You can also edit the stop word list to tweak the outcomes of different common words. I really like the DocuBurst with Beowulf. Like the rest of us, you've discovered this tool (and others) can be quite finicky about actually visualizing a text, but once it works, the results are always intriguing! (I really like the title for your project, btw!)
ReplyDeleteSuch a creative collection. I noticed that your word frequency graph actually has the name of the work on the x axis. How did you do that? I must have missed that. :-) I just put all the texts into one word document. Thanks and I really like your title too.
ReplyDelete