derBlog

Once again, this December the Center for Computational Science at Tulane hosted a “Computational Art Show.” This is the second year in a row that they’ve hosted this event. Last year, I found out the day of the one-day affair and was able to arrange my schedule so that I could attend. This year, I didn’t find out till the week after, so I can’t even give a first-hand account of the show. It’s a shame really. This is exactly the sort of thing that I’d love to see more of at Tulane…but not in its current formation. In fact, it seems to do more dis-service to the ideals of interdisciplinarity and collaboration that I think it should represent than anything.

As I say, I didn’t get to see the actual show, so all I have to go on is a short blurb with video posted by the Tulane NewWave and a web page that has images and descriptions of several of the pieces in the show. While I agree with the basic assertion concerning the need for data visualization proffered by director of the Center (and de facto curator for the show), Ricardo Cortez  (“The computer codes give you back a bunch of numbers and you want to visualize what’s happening. … You can’t do that by looking at numbers or data. You have to look at figures or pictures.”), his further comments make it very difficult to know how to approach the work. Here is an excerpt from the NewWave notice:

In preparing work to be displayed in the art show, researchers are given the liberty to add an aesthetic element to the presentations, says Cortez, who also is a Duren Professor in the mathematics department.

“I think it’s freeing for scientists to do that because for once you don’t have to be accurate scientifically,” he says. “If my aim is to produce an appealing picture, I can do that.”

I know several data visualization professionals who would choke on their morning Wheaties if they read this quote. Right off the bat, any usefulness of the work as visualization is diminished. If the images aren’t accurate, then their usefulness in understanding the data is compromised. In other words, they can’t be trusted (then again, no image can be, but that’s another essay). I am all for taking liberties in the name of aesthetics, but if the purpose is to understand the data better, then those aesthetic choices must be in service of the data. I’m just not sure when aesthetics became an excuse for a lack of accuracy.

Of course, the flip-side of this discussion would be to meet these images as images, not as a representation of any particular data. Before doing so though, I’ll point out that I do not think this is what the students involved intended. If you read the descriptions of individual projects, it is clear that these are meant to be considered in relation to their data sets not as individual artifacts. If, however, we agree that the images have been relieved of the burden of burden of scientific accuracy, then what are we left with? Unfortunately, not much.

One of my main aesthetic criticisms of this work is the over-reliance on the default chromatic “rainbow” color palette. I say “default” because without really knowing what applications they’re using, I can’t say for sure, but it feels like a standard output setting for statistical analysis software. Anyone who has ever skimmed an Edward Tufte book would think twice before using this color palette. It might be a good for quickly understanding variations in data (which we’ve already established we aren’t interested in, right?) but as an aesthetic choice, it leaves something to be desired.

Really, the problem in the default output of a statistical analysis program is that it puts the emphasis on the numbers…or rather some overwrought visual understanding of the numbers. There is no subtlety here! But that brings us back to the data again. There is some underlying assumption that the beauty is in the data itself and so it doesn’t need the aid of aesthetic choices. Of course, that viewpoint is antithetical to Cortez’ description of the show which leaves us in a bit of a conundrum.

Of the images posted on the show’s web page, the one I’ve posted at the top of the page here by Priya Shilpa Boindala is the only one I find visually interesting in its own right. It could be a digital print or a pen and ink drawing, and it doesn’t matter to me that its represents the  “collective dynamics of microorganisms in viscous flow” though that fact adds some interest…if only I knew I could trust it as a representation of that fact… Most of the others seem to have very little intervention on the part of the students, and  the one image that clearly has been manipulated could have used some editorial guidance.

In the end, I wish that Cortez and his associates would have reached beyond their own domain to engage in a dialogue with the Art Department perhaps. We may not be data visualization experts, but we do have a bit of experience in the aesthetics department. I, for one, am quite sure that I am often too focused on my own work and would welcome the chance to collaborate with mathematicians and engineers. The possibilities of collaboration might just turn an event that is seen as lagniappe for one department into a richer experience for our students and ourselves as artists and researchers across a much larger section of the university.

/der

I’ve been meaning to write this post for over a month…ever since SIGGRAPH. I’ve titled it like a doctoral thesis, big words and colons, more to satirize the endemic navel-gazing of modern academia than anything, but the tension between making (knowledge, in this case) and teaching has certainly been at the forefront of my thoughts in the last several months. Questions of what an “instructional technologist” should be doing, wrestling with my own identity as a teacher and a trainer (two different things, I think), not to mention as an artist…all have made for an interesting ongoing–unfinished–inner dialogue.

One central question has been the place of technology in education. For years, I’ve explained my position as “teaching teachers to teach better with technology.” Aside from the obvious alliterative enjoyment of such a turn of phrase, what it really came down to was helping faculty members to communicate better with their students…or at least to communicate less badly…using technology (chalk to computers). Breaking it down further, it often came down to trying to rescue really bad PowerPoint presentations. However, at some point, even this had been done. I’d reached a level of audience saturation. This is common among all instructional technology groups at institutions of higher learning.

Of course, thanks to Moore’s Law…and Mr. Bill…there’s always another version of PowerPoint or some other new technology (blogging? Twitter? etc.) to keep us going. What we have going could be described as the technology-academia complex. We believe that we have to keep up with the ways in which our students communicate or somehow we’ll be left behind. In my opinion, the crisis is not in how we deliver our teaching as much as it is in what we are delivering and how we’re selling it, but that’s for another time…

Now, back to SIGGRAPH and the real reason for this post. Among the many mind-blowing things I saw at SIGGRAPH this year was a presentation on a project out of Ohio State University called “Synchronous Objects for One Flat thing, reproduced.” One Flat Thing was a dance choreographed by William Forsythe back in 2000. Working with Forsythe and a host of others, Maria Palazzi and Norah Zuniga Shaw, both from OSU, created “Synchronous Objects,” a project that takes the dance as data then analyzes and visualizes it in an effort to “reveal the interlocking systems of organization in the choreography.”

What this team has built though is more than just an analysis of a form; they’ve built a new way of understanding choreography, not just a visualization of data but new knowledge, new pathways of thinking. In other words, this is not just about teaching the dance (pedagogy) but about the production of new forms and new understandings (epistemology). And none of it could have been done without digital technologies.

I’ve come back to Synchronous Objects several times since SIGGRAPH. I think it offers a model for what “instructional technology” should be: an instrument of instruction. This gives instructional technology the latitude to be both pedagogical in the sense that it can be used as a tool for teaching, but it also, and I think more importantly, allows it to be a constructor of new knowledge from within that pedagogy.

As educators, we sometimes forget that the classroom can be a collaborative, productive environment. The sciences do this better than the humanities. University labs produce new knowledge. There is room for the production of new knowledge in the humanities as well. “Synchronous Objects” demonstrates this. What kind of change in academic culture would it take for such a goal to become widespread though? Is it the administration? the faculty? the students? all of the above?

And what role can instructional technology play? Certainly, it is not about better PowerPoint presos. Or at least, that is but a small part of the answer. Tulane has been involved in a couple of projects that begin to broach the possibilities although mostly from the perspective of a University service. As long as a project is defined in terms of a client relationship, the possibilities will always be compromised because it will be seen as a matter of content and delivery. Instead of synthesis, it will always focus on translation. “Synchronous Objects” isn’t about recording, display or demonstration. It is about revelation, a remarkable “Eureka!” of a moment where interdisciplinary collaboration has produced new ways of seeing. This one is going in the bookmark file, to be revisited often as a reminder of the possibilities.

/david

So, I’ve been interested lately in trying to create some visualizations, specifically to hang in the front of the ILC’s offices here in the Tulane’s library. After spending some time wandering through Tulane’s website, I found some data to play with. The problem is that the data has already really been parsed, so the questions have already been asked and answered. There’s no way to cross-reference data, for instance.

I’m starting to ask around though and have some promising leads to more granular/raw data that might prove more interesting. In the meantime, I went ahead and created the map above. It shows geographically where the freshman class of 2008 came from. I’m not really satisfied with it, but I think it works as a proof of concept. I think being able to unroll a poster in front of someone might show them the concreteness of the project. We’ll see…

In the mean time, tell me what you think of this poster. And let me know what kind of questions you’d like to ask of student population data. I’ll try to post a picture of the poster itself soon….if I can remember to bring in my camera…

/der