First some context: one of my main interests is in centers of manufacturing along US borders, specifically along the border of Texas and Mexico. Growing up in the Rio Grande Valley in South Texas, I experienced what I believe was a shift in economy. Up until the early to mid nineties, the Valley’s economy was largely based on agriculture–when Florida’s citrus is out of season, South Texas is responsible for almost all of the US’s citrus (especially grapefruit). The enactment of NAFTA largely changed this: the lower Rio Grande Valley began to work with Mexico and their maquilas–manufacturing centers that provide cheap labor for major companies. At the same time, industrial centers and exporting/importing centers began to pop up along the border.

So here’s the idea I have: creating a visualization that shows the growth/decline of industrial centers along both the US-Mexico border, US-Canada border, and in the midwest. My theory is that you’ll find an inverse relationship in growth between the US-Mexico border and the midwest specifically.

I’m looking to you all for suggestions. This would largely be displayed using some sort of graphic–maybe dots that grow or shrink depending on the year you select.

What do you think?

Information is beautiful!

For my data set I wanted to look at something a little different. Watch the video (spoiler, IT MOVES) and see how the guy created a visualization from Van Gogh’s “Starry Night.” This is a highly-packed, visual representation of a ton of data: Van Gogh’s style, his influences, the brushstrokes, the color, the scene, etc. By breaking it down into these moving pieces of the “big picture” we can investigate the complexity of each piece of data as an important, calculated, designed piece. Try to focus on one brushstroke, the motion of the data set plays with the fixed nature of the data (i.e. the picture as a static object) by putting the flow of the picture on display through these tiny moving loops. Instead of the painting as a static whole, we can really get into the complex intermingling of the data as it flows together and apart in creating the blues and yellows and browns of the image. And of the image, a fully realized scene.

Why this is super cool, though, is how the guy starts moving the data around. Here the fixed nature of the data is suddenly pushed and pulled against the back-end brushstroke loops, resulting in brief redirections of the flow before falling back into place. We’re still in the rigid structure of the interplay of these bits of data, but we’re able to affect the connections between them by dragging our fingers around. When he pulls that yellow from the star out, or messes with the flow of the background, we see how these reconfigurings have an effect on the whole of the image, and the way slight changes to Van Gogh’s individual data nodes can result in a different “big picture.”

I like the Lima book specifically because his included data visualizations are so impossible, unusable, un-break-downable. We get that pulled back image and we get the sense of chaos, or order, the data conveys, but we need to get down into the visualizations and see the bits of context. We need to get our hands dirty, by pushing the brushstrokes and tracing the lines. The book helped me start thinking about what I want those broad connections to look like, or the different ways I could start re-arranging what I’m thinking of in 2 or 3 dimensions. But the little nodes and the data they contain is where we need to be concerned–in what we can do with the big picture and how we can make sense of that tangled web of data when it comes down to individual subjects, instances, and contexts.

Looking through Visual Complexity, I feel that network diagrams are at their best when they show relationships between things in detail. Zoomed out network images, while often pretty, seem sort of like big fancy blobs to me. So, I find the diagrams on 102-103 more helpful than the ones on 100-101, because the ones on 102-103 label nodes (perhaps the zoomed in versions of the others do so as well.)

Visual Complexity is a beautiful book that has many useful things to say about visualizing data–I think it’s strongest when it sticks to relating strategies for visualization. I found a few of the claims made in the early framework part of the book to be highly problematic–I grant that this is not meant to be a scholarly work that attends to fine-grained distinctions, but that doesn’t mean it’s allowed to paint broadly misleading strokes: “But it was during the Industrial Revolution that many of our hierarchical conceptions of society were widely put into practice” (57). Really? Feudalism, anyone? I also find problematic the suggestion on pages 59ff that decentralized networks are (completely?) egalitarian. Although networks arguably do allow for greater participation, it’s a fallacy that networks are inherently egalitarian or democratic, which Lima seems to imply. My criticism is perhaps somewhat tangential to Visual Rhetoric, but at the same time it’s worth considering as part of the framework that undergirds Lima’s argument for network mapping.

As I re-read Lima’s Visual Complexity this week, I couldn’t help but think about Tufte and the relationship between data and aesthetics. And thought about this site (and the book it supports): Information is Beautiful. Then, I open up my Sunday New York Times and find this substantial and smart introduction to Big Data. In the Gray Lady.

I remember trying to decide between Lima and McCandless’ texts, and I erred on the side of Big Data and rhetoric. I don’t think that’s much of a stretch for us. McCandless is more Tufte-like in his insistence on the aesthetic, whereas I read Lima as much more rhetorical. And Lima’s visualizations are, I think, more powerful and less pretty precisely because they wrestle with issues of information articulation. Like Horn* before him (see Horn, Visual Language, 1999), there is power especially in seeing visualizations as a language—see Lima’s chapter 5 titled “Syntax of a New Language”. Lyotard’s différend allows us to be a bit more precise and assert that visualizations allow us a means of transcending the current limits of language. See Liu’s Auto- différend page if you aren’t familiar. Or perhaps visualizations provide a route, a heuristic, from inability to articulate through to our drafted and preliminary articulations, from ineffability and a sense there is “something” there, to language, to explanation, and on to persuasion. So for me, for us, following Lima, we are grasping at a “Rhetoric of a New Language” as Lima’s chapter articulates its syntax.

So in asking you to work with datasets, I am not asking you to come up with fully formed projects or ideas for Tuesday, but to search through these links I’ve provided and set out on your own to explore data with these new visualization tools&em;rhetorical, technical, and aesthetic—and to think about problems or questions for which you think there may be an answer that does not yet have an idiom (in Lyotard’s terms)—that we do not yet have a language to express. How can (how might) we use these visualizations tools to offer a “new” or emergent means of expression and persuasion not previously open to us? So Data Visualization is in this way a means of rhetorical invention.

With any luck, this line of argumentation leaves anyone who was uneasy about exploring Big Data, datasets, and visualizations a little less uneasy and a bit more confident that my obsession with visuals remains visual rhetoric. The challenge persists: what arguments are (our and others’) visuals making possible that are not otherwise possible?

Periodic Table of Visualization Techniques (And why use a scientific metaphor?)

I’ll kick things off with some geographic/mapping examples, unless you’re all so excited you push my stuff to the side. (I can hope.) IBM’s eBook (linked below) offers a solid introduction to Big Data. My recommendation is to spend lots of time with Lima, especially Chapter 5, and flip through these links and find ones that speak to you, your interests, and your current obsessions towards understanding various assertions of dataviz and visual rhetoric as invention. Once you’re okay with that, I’m curious about the kinds of arguments you want to make, and the data you need to do it. That’s what I mean about finding datasets. Pracjus et al remind us of the importance of whitespace – Helvetica. Scott comes from advertising and argues for a theory of visual rhetoric, which connects it to my interest in Lima as well as to Handa. And if you run Ray and Charles Eames in the background while reading Lima, all the better: Charles & Ray Eames: The Architect and the Painter . For those who may have missed it, Kristen Moore offered this link as a powerful dataset for our consideration: The Adjunct Project collecting adjunct-related data from the instructors themselves: crowdsourcing and user-centered…and prosumer. Let’s see how we can rise to her challenge.

Big Data:
Define Big Data at Wikipedia.
McKinsey Report on Big Data
O’Reilly Defines BD
ZDNet is next,
then Information Week
Us Census Data is Big Data,
that IBM and
EMC want to play with.

IBM has dedicated itself to Big Data,
releasing an IBM eBook
and they market with Big Data.
IBM has link to visualization tools
as well as a DataViz community portal called ManyEyes.
And business journalists are paying attention.

TED has committed itself to BigData:
TEDxUofM – Jameson Toole – Big Data for Tomorrow
big DATA, BIG stories (TEDx event at University of Vermont) see esp. TEDxUVM 2011 – Peter Dodds – Big Data and the Science of Complexity

Hans Rosling is a DataViz node of his own. We watched parts of the 2007 piece on poverty. He supports Gapminder and while he kept a blog, the last post was 2007. More recent posts are gapminder news.

Worldmapper and Worldometer are two favorites.

I’ve sent email in preparation for our Feb 7 class: Check out the post attached to “Links.” But since this is the public face of the class, this is the first and only post like this I plan to make.

So log on and post what you think should be public. You can see I’ve added parts of your “historic visualizations” as banner images for the site. One way we can knit together this and the dropbox technology is that you can add materials to the public folder and use that link (or you can link from your career account).

How can we take the best of our classroom discussion, examples, and projects and create a display space here? I think all your maps, historical data visualizations, and memes could go here, along with some explanatory text.

Finally, some of you brought to my attention that having everyone as administrators could create some problems, like the entire site disappearing without a trace. So I have set everyone’s role to “editor” which gives everyone significant access to content without overriding ability to destroy. So to avoid accidental destruction, I’ve rolled back that level of access.

-M