The descriptive text for the Wordle service says that “Wordle is a toy for generating ‘word clouds’ from text that you provide.” The emphasis on “toy” is my own, because while Wordle may be a toy in that it is interesting, fun, and generates play, it is certainly not “something of little value” or something useful only as a “diversion, rather than for serious practical use” (that’s me, showing off my mad dictionary skillz). Instead, I firmly believe that Wordles are the gateway drug to getting students interested in textual analysis. In a word cloud, the most frequently used words are the largest words in the display, while words used less frequently are displayed in a smaller font. It’s so very simple, yet so very useful.
Believe me, I think talking about concordances and collocates is cool, and I’m still waiting for the day I can throw “weighted centroid” into a conversation without people looking at me like I’ve grown another head, but the truth of the matter is that if I were to send my composition or lit survey students off to the Text Analysis Portal for Research (TAPoR) they’d lead some sort of revolt, I am sure. But put them in front of Wordle with their own essays or their own texts for reading, and they “get” it. What they “get” is that word clouds can:
- help introduce a topic. For instance, who do you think is significant in Heart of Darkness—if you said “Kurtz” then you’d be on to something.
- help students discover key words and ideas they might not otherwise have noticed (or support ideas they might have).
- help students reflect on their own writing and word choices.
Last year, I had the good fortune of leading a few class periods of Dr. Donna Campbell’s upper-division undergraduate course in American Literature 1855-1915. We spent some time with the Whitman Archive and Dickinson Archive, but the tool students really wanted to use was Wordle.
One student generated the Wordle that you see in this blog post—it is is of Walt Whitman’s “Song of Myself” from the 1891-1892 “deathbed edition” of Leaves of Grass. The student said it was really creepy (but in a good way) when the three words “one” and “shall” and “know” popped up on the screen and just stared back at her. “It was like a message from the grave,” I think were her words to me. Another student wanted to see if the text of Whitman’s Drum-Taps ‘could describe any war, not just the Civil War,” and he thinks they can. Another student took a portion of Emily Dickinson’s poems believed to be written around 1862, and saw a prevalence of words like “go” and “away” as well as “wished”—and the student then talked about those words in relation to the preconceived image that students often have of Miss Emily.
After students are hooked on Wordles, the next logical step would be to have them read Geoffrey Rockwell’s “What is Textual Analysis” and see if I can’t interest them in stepping their interest up a notch.
Wordles can be used in many different classrooms—anywhere there’s text to be analyzed. Wordles in the American History classroom? Knock yourself out with Wordles of all the Inaugural Addresses of U.S. Presidents from Washington to Obama. Teaching Visual Rhetoric? You can have a field day with Wordles, as there are many display options: fonts, layouts, color schemes are all modifiable.
I’ve used Wordles for a few years, think they’re ultra-cool and useful in the classroom, yet I meet people all the time who haven’t heard of them or made the connection between colorful jumbles of words and a viable pedagogical tool. But it sure can be!
30 Comments
Nice post, Julie! I love Wordles for all the reasons you describe. In classes the past few days (classes for preservice Language Arts teachers), we’ve used Wordles with both their formal writing (graded essays) and their informal writing (their blog posts). Students were amazed at how much they learned about their own writing by seeing the work displayed in such an unusual way. As a class, we were able to brainstorm about how Wordles could be used with students of all ages and their writing. Try a Wordle with 10 or 15 words, the type of writing a child might produce, and see how it displays the work. Very cool.
Billie (and Kenneth too?) — there are a lot of people using Wordles in the K-12 classroom, so teaching the teachers-to-be how to use the tool (and why) seems like a good thing to me!
Julie,
Thanks so much for this post. We use Wordle frequently but I am always amazed at the things creative people come up with. The next Teaching Methods class I teach will utilize this. What a GREAT idea!
What a bang-on idea. I’ve used Wordles for fun with my own fictional writing but I never had the clever idea to apply that in class. I teach in history and this would be great for some of the textual analysis we’re doing in the senior seminar on early modern gender as well as another series of seminars I run on medieval chronicles.
I’m going to throw up a post for my seniors about the joy of Wordling. Thanks for the inspiration!
Let us know what happens!
Jodi Dean has an interesting post that’s skeptical of this premise: http://jdeanicite.typepad.com/i_cite/2009/01/tag-clouds.html.
(NB: I use word clouds myself.)
Right – I read the post when it made the rounds during discussions at either Computers & Writing or Digital Humanities or THATCamp or maybe it was just Twitter at the end of July. It is an interesting post. But my point isn’t that Wordles or word clouds in general are in any way a replacement for original thought. They’re simply visualizations that I have seen “reach” students and given them entry points into texts. Additionally good outcomes include students examining a data visualization and saying “but that’s the opposite of what’s happening, why is that so” and then continuing to investigate beyond the wordle, or saying “well gee, look at the effect you get when you change this from horizontal to vertical display” and so on. Wordles as a starting point, a gateway, not the be all and end all.
Julie I like your positioning of this kind of text analysis that it is a tool for teaching rather than scholarship. I had not thought of it that way.
I had a side conversation at THATCamp PNW with a fellow about text mining and history. I asked if there were any examples of text mining producing unique and original historical insights and analysis, rather than just reinforcing what we already knew. “Not really,” he said.
I agree entirely with you Julie, and I think we’d do well to embrace that reality instead of fighting or bemoaning it. Geoffrey Rockwell and I have been giving a lot of thought recently to how we might present analytic tools in a more gradual manner, so that the foreignness of conventional text analysis interfaces doesn’t overwhelm readers/users. To tell you honestly, I’ve become much more interested in providing tools for the general public, like bloggers, than to my recalcitrant literary colleagues (to unfairly generalize). In any case, I think one strategy is to showcase the current, most accessible level of analytic tool (a visual word cloud, for instance), but also present the next level of sophistication and complexity (which may differ according to the circumstances in which the tool is located): the next level may appear in proximity to the first tool, or be accessible directly from the first tool. So if Wordle is pot, what are shrooms?
Stéfan, I can honestly say that one of my goals for my academic career is to do whatever I can to, as you say, reduce the foreignness of conventional text analysis interfaces — basically, to figure out, build, and push the shrooms.
I have long used word counting software in my fiction writing courses, to give students a sense of words they might overuse (use as crutches) and to show them some unconscious patterns. This has usually been revelatory for them. They rarely get such a disinterested view of their own work. I, like, Billie above really do think students benefit from this. I ran my last book through one, and noticed how much I relied on “darkness,” for example, and “breathed,” I also mention necks a lot and how I overused “hobbled” in an attempt to not overuse “walked.” The book came in at 87,000 words, and I don’t think I’d have noticed these patterns with out some assistance.
I like your term “disinterested view” — that does make for a good selling point.
We did something like this at the outset of the semester with Heart of Darkness. Then, we forgot about it (in an Introduction to Literary Criticism course). But, this week, as we leave behind Deconstruction and move to Psychoanalysis and New Historicism, they were asking about this idea of word cloud and what to do with it. Everyone thinks it’s cool, but getting them to use it for research is another thing. Wish we had an entire online journal dedicated to using digital tools in pedagogical fora (ahem, DHQ).
Want to see something frightening or perhaps exciting with a Wordle? Put your student evaluations into it. I did… Besides seeing your name really big, some adjectives might start to stick out. I thought mine was quite fun. I’ll do more of them in the future. Just to show off, here’s mine: Wordle of some of my student evaluations!
Wordle is indeed a toy. You need to understand its limitations before you use it for serious work. I wrote on this topic last May in an article titled “Limits of Visualization: Wordle Misses Meaning.” You might find it interesting. http://intelligent-enterprise.informationweek.com/blog/archives/2009/05/limits_of_visua.html
As with the comment regarding Jodi Dean’s post, upthread, I’m pointing out that Wordles can be a pedagogical starting point — a gateway, not the be all and end all. Nowhere do I, or would I, advocate that a word cloud of any sort — colorful or otherwise, tied to a specific service or otherwise — be the singular object driving serious scholarly work.
I guess I’m confused by this distinction between teaching and research, as if we shouldn’t be giving the straight dope to students. If a tool is not useful for the scholarship of a discipline — and I think it is an interesting and problematic pedagogical choice to use Wordle if in fact it may distort such scholarship (as I think Wordle does, not only for the reasons Jodi Dean and Seth Grime point out but, if you did want to take vocabulary frequency out of context, font size is a problematic and optical-illusion-ish way of conveying quantitative data, and I suspect the random position of words give unplanned cues to cognition).
No one is arguing that “we shouldn’t be giving the straight dope to students.” If you don’t want to use Wordle in your teaching, then don’t. If you want to use Wordle and let your students know what its limitations are, then do so. If you have a different and better tool for visualization in the classroom, then use that one (and let us know what it is). If you don’t want to use any visualization tools at all, that’s fine.
However, if you’re looking for a forum in which people argue about who knows the most about this or that tool, then take it somewhere else. ProfHacker ain’t the place for such exchanges.
We’re the place for “Here’s what we’ve found useful. Maybe you’ll find it useful, too. If not, that’s okay…”
Dig?
WTF, George?
I apologize for my rather curt and rude comment…
…I’ve got more to say about using Worldle and similar tools in a future blog post, but ScreenSteps–the software I used to begin composing that blog post–appears to have eaten it.
In a short post of this nature, which is purely about introducing a tool to others who may then make their own decisions about it, obviously I am not including the entire context of how I’ve used it in the classroom. For instance, I do spend time explaining what it does, what it does not do, and where it fits in the greater scheme of textual analysis tools (see Stéfan’s comment above). As I’ve said, not only do I not think that Wordle is the be all or end all of tools for students, but my students are aware of the inherent rhetorical issues of it as well.
The nigh-obligatory Wordle of ProfHacker itself.
Apparently we’re all about two things: students, and chicken.
Is that all ~200 posts? I didn’t think we talked about chicken THAT much…
Well, that just says something, doesn’t it? We need more beef. Or fish. Or something.
And I don’t see Google on there at all. There are some obvious problems with this wordle.
chicken
Okay, now do one for each author so we can see who’s responsible for all of that chicken.
Or maybe come up with some kind of color-coding system where one author is blue, one is red, one is yellow (etc) and then each word’s color will reflect how many times a particular author used a particular word…
Don’t need a Wordle to tell you Billie is responsible for all the chicken. And George, there are approximately 87 billion other things I’d want to do before a color-coded word cloud of author posts!
Let’s not become chicken haters people. We are bigger than that, better than that. (And George, I’m with Julie: I have no time to color code anything. I’m too busy finding ways to write “chicken” on this website.)
I think Wordles are pretty nifty, especially for helping my history students grasp the subtle and not-so-subtle emphasis in Wikipedia entries. Here’s one from the Wikipedia’s entry on the American Civil War: http://www.wordle.net/show/wrdl/1260366/The_American_Civil_War_According_to_the_Wikipedia
I’ve been working recently on this tutorial about Wordles and poetry:
“How to use Wordle.net to help analyze and compare poems“
7 Trackbacks
[...] This post was mentioned on Twitter by Ryan Cordell, Karin Dalziel. Karin Dalziel said: RT @jcmeloni: new @ProfHacker post by me: "Wordles, or the gateway drug to textual analysis" http://bit.ly/4BJCMy [...]
[...] don’t mind being a pusher for this kind of drug. Read about Wordles, or the gateway drug to textual analysis at [...]
Social comments and analytics for this post…
This post was mentioned on Twitter by ryancordell: RT @nirak: RT @jcmeloni: new @ProfHacker post by me: “Wordles, or the gateway drug to textual analysis” http://bit.ly/4BJCMy...
[...] groups, Julie identifies wordles as the “gateway drug to textual analysis” (one of our most-commented posts [chicken]); and Natalie explained how to use Library of Congress subject headings to improve [...]
[...] heard a lot about Wordle lately and after seeing Prof Hacker’s recent post about I thought I would create a Wordle of my Master’s Thesis from last year. A [...]
[...] appears to be being employed in some interesting ways by some interesting people. For example, in Wordles, or the gateway drug to textual analysis (ProfHacker, 21 October 2009) and Using Wordle in the classroom (ProfHacker, 13 November 2009) both [...]
[...] appears to be being employed in some interesting ways by some interesting people. For example, in Wordles, or the gateway drug to textual analysis (ProfHacker, 21 October 2009) and Using Wordle in the classroom (ProfHacker, 13 November 2009) both [...]