viz. - data visualization

Militant (Feminist) Grammarians

Jenn Shapland — Wed, 05 Mar 2014 20:49:35 +0000

Cropped from image below

You know you’re a huge nerd when multiple people from various corners of your life all forward you the same link, and that link is a bunch of diagrammed sentences.

This snazzy, minimalist new print from "renowned" infographic artists Pop Chart Lab satisfies the demands of everyone's favorite niche demographic (all those grammar-fiends/”classic-literature”-snobs/data-visualization-enthusiasts/fans-of-quality-design in your life) to a T. But before you place your order, let’s take a closer look at what this “Diagrammatical Dissertation” actually visualizes.

Image via Esquire

Apart from the anti-feminist overtones of the Reed-Kellogg diagram schemata itself—which just reflect the gendered taxonomization of the English language (from whence do we derive concepts like objectification and subordination, exactly, if not from the rules that structure everything we say?)—this poster perpetuates all the standard inequalities we’ve come to expect from lists of “Notable Novels.” Of the 25 opening sentences the poster presents, four were written by female authors (a whopping 16%). One black author makes the cut, Toni Morrison. 13 have male subjects or refer to male characters; only 4 of the sentences contain female characters or narrators (and in one of these the female appears as “fire of my loins” w/r/t the male narrator, so that probably shouldn't even count). Only two women are their sentence's subjects; seven male subjects voice their own sentences. It barely needs to be said that none of these sentences come from openly queer writers.

Cropped from image above

I know what you’re thinking. "This isn’t news, yo! It’s just a reflection of literary history. More men historically wrote and published novels, and more novels have been published about men, therefore any chart would demonstrate this inequity! And besides, they're talking about classics. When reading or penning a classic, race and gender magically disappear. Right?!"

Wrong. Here's (one of) the (many) problem(s), provided for us in the poster maker's own humble description of his or her (but in all likelihood his) product:

This chart diagrams 25 famous opening lines from revered works of fiction according to the dictates of the classic Reed-Kellogg system. From Cervantes to Faulkner to Pynchon, each sentence has been painstakingly curated and diagrammed by PCL's research team, parsing classical prose by parts of speech and offering a partitioned, color-coded picto-grammatical representation of some of the most famous first words in literary history. Whether you’re a book buff, an English teacher, or a hard-line grammarian, this diagrammatical dissertation has something for the aesthete in all of us.

Gross. This chart offers its list as a set of “famous,” “revered,” “curated,” “most famous in literary history” works for “the aesthete in all of us.” All of whom, might I ask? And talk about range! It covers "Cervantes to Faulkner to Pynchon"—and all the straight euro/anglo male novelists in between. The problem with using words like "notable" and "classic" to describe a set of authors that is 96% male and/or European by origin in 2014 is (duh) that it simply perpetuates the idea that some literature is more important than other kinds; and that most of the "important" books EVER have been written by or about (straight) men. To which I say, uh, nope.

Like the Buzzfeed quizzes I mentioned a few weeks back, curation has ossifying powers. Every time I see a new list of “best books,” which is what this actually amounts to, the same ugly old narrative about who sets the cultural bar rears its representational head. And it isn’t any surprise, is it? Just take a look at the latest VIDA statistics: we’re STILL publishing an average of 75% male authors in the most read magazines in the US for nonfiction and fiction ("Drumroll for the 75%ers: The Atlantic, London Review of Books, New Republic, The Nation, New York Review of Books (actually holding steady at 80% men for four years) and New Yorker"). Mainstream book reviews have always skewed overwhelmingly toward books written by men. In words from Cate Blanchett's controversial acceptance speech at last week’s Oscars, “The world is round, people.”

Image via VIDA count 2013

And maybe, just maybe, if this much-tweeted-and-"liked" and temporarily back-ordered diagram had included a more varied group of authors and texts, it wouldn’t be basically all direct objects and strings of subordinate clauses—that one was for you, feminist grammar nerds. Is it any surprise that the only sentence on the chart that neither takes an object nor subordinates a clause is the opening to Beloved, “124 was spiteful”? Subject noun + copula + subject complement = simply exquisite.

A Workshop from a Visionary about Data Visualization

james.wiedner — Thu, 02 May 2013 18:51:52 +0000

Source: Mark Andrew Goetz

About 2 months ago Austin was lucky enough to be among the handful of cities selected as a stop on a one-day presentation and workshop featuring Mr. Edward Tufte. What's more: The DWRL agreed that covering the cost of admission for a few of their staffers to be money well spent (thanks, Will Burdette).

For the uninitiated, Mr. Tufte is the granddaddy of all things related to visual representations of large amounts of data, complicated concepts, historical trends, and- quite literally- just about anything else you could think of. Hailed as “The Leonardo Di Vinci of Data," by the New York Times. Tufte was synthesizing massive amounts of information into beautiful visuals before the term “big data” had even pushed “the cloud” out of the way as the buzzword(s) of the moment.

In Tufte’s 2006 book, Beautiful Evidence, Tufte discusses at length an entirely new means of data visualization he called Sparklines (which, according to his workshop, he created in the mid-1990s). Sparklines can depict thousands of measurements in graphics that are generally small enough to fit within a longer piece of regular-sized text (like the one you’re presently enthralled by...yeah, right).

Source: Edwardtuft.com

As the above illustrates, sparklines are “dense, word-sized line graphs that can show, in less than an inch, the last 100 readings of a patient's glucose, a year's worth of stock activity or the win-loss record of a baseball team.”

So, Tufte’s Sparklines came along at a point when a new way to visualize the unprecedented amount of data people were being exposed to was sorely needed. Evidence of this can be seen in the proliferation of the handy little graphs that Tufte came up with. Sparklines have been incorporated into Excel for Windows and Mac. Google introduced sparklines into their API data products in 2007. As of this writing, a google search for “sparklines” returns nearly twice as many results as a search for “steve jobs” (106,000,000 vs 46,000,000; I chose “steve jobs” for the comparison because Tufte did the same thing in 2011, at which time he had merely twice as many returns as “steve jobs”).

And it’s not just massive corporations and data aggregation sites that are finding utility and compelling aesthetics in Sparklines. Just click here #sparktweets the next time that you’re going to go bouncing around twitter. Sparktweets are simple, Unicode sparklines, and so can be included in the body of a tweet). There’s even a Twitter Sparkline Generator out there!

Source: Data Driven Consulting

However, even though Tufte has embraced technology in creating his visual representations, to describe Tufte as a zeitgeist with an unequivocal belief that “newer is better” would be to miss his message completely. Although his presentation on Sparklines was an amazing demonstration of how we can make sense of seemingly infinite amounts of data in the digital realm, there was a larger theme underlying Tufte’s presentation that might temper any overexuberance for technological progress one would otherwise have. A point he kept driving at was that the latest isn’t always the greatest, and that new technologies should only be implemented when they actually better lend themselves to a particular function. He displayed a recent picture of NASA’s Mission Control room in Houston (not in a PowerPoint, of course), pointing out that the technology employed when sending people into space consisted largely of Windows 95-based computers and landline telephones (remember those?). Tufte said that the reason NASA was still employing these older technologies wasn’t because they couldn’t afford to implement the Next Big Thing whenever they so desired. Rather, the technologies of Windows 95 and landlines, because of their stability among things, were determined to be the best technology for the particular function they were employing it for.

Towards the end of his presentation, Tufte rendered the bibliophiles in the room awestruck when he pulled out an original copy of Galileo’s “Istoria e dimostrazioni intorno alle macchie solari” (“History and Demonstrations Concerning Sunspots and their Properties”). He left us all awestruck when he demonstrated the ways in which this 400-year-old book contained better examples of data visualization than anything we had seen on his projector or on our own laptops: to illustrate the movements of the earth that he had observed, he created a small movable, fold-out, 3-D cutout of the rotation of the earth around the sun, as he had observed it.

The computer screen, for all of the praise we give it, Tufte, hasn’t really taken us that far forward in our abilities related to visual representation than the book that preceded it. Further, the book had not really relieved us of the principle shortcoming of the scroll before it…or the slab of stone before that. Why? With all of them, you’re still only writing on and looking at a flat surface.

Tufte gave another shout-out to the actual, paper and ink book that we seem to be so desperate to abandon by showing us a book that was even older than Galileo…about 200 years older. Tufte said that, despite having been printed on the medium of paper that we presently see as unquestionably inferior to the digital medium, the pages of Euclid’s Elements STILL had better “resolution” than last season’s Dell Laptops.

The workshop didn’t end there, many more interesting facts, graphics, and visual demonstrations followed. I’m sure he had enough material to stay up there all day. Instead of doing that, however demonstrated first-hand one of his fundamental techniques for successfully conveying your message: Always end early. I enjoyed the hell out of the presentation, but ending unexpectedly was really nice!

(As a concluding aside, I did stay long enough to meet him after the lecture and to give him a DWRL t-shirt…which, I might add, he seemed to be genuinely interested in.)

How USA Really Voted on November 6

Chris Ortiz y Prentice — Sat, 17 Nov 2012 15:10:56 +0000

Image Credit: IDV Solutions

What a wonderful map! This IS the popular vote on November 6, 2012. John Nelson gave us this map, and we thank him for it. It's called a "pointillist map:" one blue dot for every 100 votes for President Obama, randomly distributed in the county in which the votes were cast. One red dot for every 100 votes for Mr. Romney. You've heard of purple states? Well here's our purple country. Click the link on the image credit to find a large and hi-def version of this map. Then meet me back here, won't you?

I'll be candid. There's an irrational part of me that wants the result of an election to match how much blue or red there is on the map. I know that's not how it works. This time, the state-level electoral college map came out pretty evenly red and blue. But take a look at the county-level map:

Image Credit: Wikipedia

As usual, it looks like a sea of red with a few islands of blue, and yet, as we all know, President Obama was elected for four more years. I realize that it's a question of population density not geographical space, but now, at long last and thanks to Mr. Nelson, I can see that.

Mr. Nelson tells us he was inspired to make this kind of map by his advisor, a Professor Kirk Goldsberry. Here's a pointillist map of the 2012 presidential election Professor Goldsberry did of the Dallas Fort Worth Area:

Image Credit: Kirk Goldsberry/KK Outlet

The top map shows red and blue dots for Mr. Romney and President Obama respectively. The bottom map shows voters by ethnicity. (Can you guess? Try and then click the link to find out.) What a revelation! Of course, pointillist maps are only one of the new mapping techniques to show election data, but they are a powerful one. Looking at John Nelson's map I find myself thinking: so this is who we are.

Image Credit: IDV Solutions

One thing we seem to be is country and city. Do you notice how there is a ring of red around the purple-blue cities? That seems to hold true around the nation.

Colorful Geographies of Beliefs

Calliope — Fri, 16 Nov 2012 19:31:22 +0000

Image Credit: MIT Technology Review

This electoral map, created by Princeton mathematician Robert J. Vanderbei, uses a spectrum of colors between blue and red to represent the ratio per county of Democrat to Republican votes. The height of the verticals indicate the number of votes in each county. Vanderbei's representation of the U.S. votes by region accounts for nuances in the data that other red-and-blue-state maps miss: the political dividedness of certain counties, the intensity of partisanship in others, and centers of strong voter turn out. From a visual standpoint, the map is eye-catching because it is purple. Purple is not a color usually associated with political belief. But other data crunchers, looking to complicate our picture of national voting trends, have unveiled maps this year with a similar palette. See my fellow viz. contributer Chris Ortiz y Prentice's post for an electoral map that also reveals (through pointillism instead of 3-dimensional modelling) the nation's purplish complexion.

It might be mere coincidence that Chris and I both decided to write about visualizing ideological regionalism; but it's possible that our posts register an increasing need to redraw and redefine assumptions about voter demographics. That said, I'll leave the actual work of redefinition up to political analysts and turn to the far more obscure aims of this entry: to discuss the rhetorical role of color in images that chart belief systems and controversial policies.

The distribution of color in Vanderbei's map implies that political allegiances are not as ubiquitous in some regions--like the South and Midwest--as one might suspect. It even gives the impression (which may be misguided) that Americans are politically moderate, since variegation in color usually expresses in-betweeness instead of extremes. But let's talk about the map's predominant hue for a moment. Purple is not a color that Republicans, Democrats, or Greens, for that matter, would claim, probably because it doesn't have any patriotic or environmental associations. The main things that come to my mind when I think of purple are grape juice, Lisa Frank, and Skittles. I also have the unreasonable expectation that anyone who is wearing a purple shirt should be friendly and easy to approach. I'm pretty sure that everyone has their own set of responses to this and other colors--some responses may be arbitrary or psychological, and some culturally-determined. And yet we rely on colors to codify politically important groups and ideologies. The fact that colors have both a particular and universal significance for most people is what makes assigning them to maps like Vanderbei's a rhetorically interesting (and potentially dicey) practice.

Image Credit: The Pew Forum

I looked at some graphics by the Pew Research Center for its project on Religion and Public Life to see which colors are presently being applied to other sensitive attitudes and beliefs. The creators of the chart above, which explains current state policies on gay marriage, chose an earthy palette of green, yellow, tan and white to represent the four main legal positions: states that allow it, states with constitutional bans, states with statutory bans, and those with no legal stance. It's conceivable why green was chosen to mark the states that have or will shortly legalize gay marriage. Green is typically the color of environmentally-friendly and/or socially-conscious causes and gay rights is an issue that fits with this progressive platform. It's interesting that the none of the other colors are linked to political parties, though. The color used to denote the most restrictive of the four positions (states with constitutional bans on gay marriage) is a sallow yellow shade. Conscious color choice? You decide.

The think tank's news story about the rising number of Americans who do not identify with any religion is striking for its lack of color. The article includes illustrations and graphs (one depicts the ratio of religious to non-religious individuals in the U.S. with a map-shaped group of gray and black figures) but they are all monochromatic and colorless. Does the term that the editors use to refer to this growing group--the 'nones'--necessitate such a color scheme? I wonder if non-religious people identify with the possible valences of this term and the corresponding color coding (anonymity, absence, indefiniteness)?

Image Credit: The Pew Forum

Lastly, we have one of the only sources I found floating around on the Internet that specifies which Christian denomination has the leading number of adherents in each U.S. county. (The map was updated in 2010, but I couldn't access it for free online). I'm taken by this map because of the way it differs from the one at the beginning of the blog post. The people who built this visualization opted to portray the adherents of each religion as unified and undifferentiated by using flat colors. And the data they collected indicates that most churches hold sway in contained areas of the map--the Baptists in the South, Lutherans in the midwest, and Mormons in and surrounding Utah. Interestingly, the map-makers assigned the two most dominant church bodies, the Baptist and Catholic religions, colors that correspond to the main U.S. political parties. It's worth reflecting whether one can look at a map like this without seeing the color-coded territories as translating to votes for liberal and conservative candidates.

Image Credit: valpo.edu