Geography Tag

Alva Noe on dance, knowledge, and sync objects

We are a tool-using animal. We make tools to extend our bodies, such as a rake for reaching, but also tools to extend our minds, such as notations for performing calculations. Tools solve old problems, but they also enable us to frame new problems. When we talk through a problem — in the laboratory, say, or at the kitchen table — we use linguistic tools (language) to think about our problem, or about our lives. But this exercise of our tool-using know-how also enhances and develops our mastery. To use a tool effectively requires know-how, and it affords deeper understanding.

Scientists — who seek knowledge — frequently make models of phenomena in the domain that interest them. Working with the model, thinking about the model, enables them to think about what is modeled. The weather bureau models a storm system in the hopes of predicting the storm’s behavior. Models are a very sophisticated kind of tool.

Synchronous Objects is model of One Flat Thing, reproduced. It is an instrument for thinking about this choreographic structure. one flat thing, revisited is work of art, to be sure, but it presents itself to its audience as a complex phenomenon, that is to say, as an event; the work is compelling and absorbing, but it is also, like many dances, and like life itself and the environments we occupy, very difficult to understand; that is, it is nearly impossible to command a clear view of it. Synchronous Objects, then, is an instrument designed to model the dance — along several distinct dimensions of its complexity (e.g. cuing relationships, counterpoint, kinds of actions). If it is effective, it will help the audience (and also the dancers and the choreographer) command a clear surview of the complex whole.

For one thing to model another, it must, at least notionally, exhibit something like the complexity, or the possibilities, of that which it models. A map, for example, must stand in a pointwise relation to the city it describes. This means it is always possible to get lost in our maps, or in our models. But it also means that our models become objects of inquiry in their own right. This is a common transformation in the history of thought. We devise a formal system in mathematics, for example, in order to see what follows from our theorems; we then make the system itself and the question of its soundness, completeness, consistency, etc, our object of investigation. If Synchronous Objects objects is successful, then it is likely to command our interest and attention in its own right. And because no tool or rule is self-explanatory, it will raise problems about how we are to understand it. And so the process of learning, questioning and learning more, moves forward. When it comes to understanding, there is no equivalent to the atomic bomb. So Synchronous Objects objects, this instrument, really, is a gesture towards a conversation that will, that must, go on. It will go on differently than it would have.

Forsythe’s work, if I understand it, very frequently takes these questions about knowledge, investigation, modeling as its very own concerns. one flat thing, revisited itself in a way models, or plays with the idea of modeling, Scott’s failed journey to the North Pole, and so the effort to model the dance itself is a way of engaging directly with the dance. But there is more. We are, after all, a tool-using animal and what art enables us to do — alone with philosophy perhaps — is to try to frame a clear surview of our lives, of our work to bring the world itself into focus for ourselves, and, in particular, of the role of tools (pictures, gestures, stories, models) in our lives. If this is right, the work of one flat thing, revisited, and the work of Synchronous Objects are one.
—Alva Noe


Ola Ahlqvist on Dance and Geography

variance_d1NORAH: Ola, let’s start with a discussion of your involvement in the project. I’ve had a long term interest in geography because of my previous work in environmental science but for many people the connection between dance and geography might seem surprising.

OLA: For me too, in fact my biggest surprise was how familiar I felt with some of the things that choreography and this particular dance contain. A lot of analysis and visualization work we have done has used ideas from geographical analysis and spatial theory, and it was exciting to see that they were interesting to you as ways to look at and understand this dance. What intrigues me most at this point is how much similarity there is between choreography and daily life as a spatial phenomenon (which is what we study a lot as geographers). The spatial dynamics and interactions are guided by the environment, there are some rules, but also a fair amount of improvisation in both “worlds”.

In terms of how these collaborations come about, my way into the project was mainly as advisor to a doctoral student, Hyowon Ban who got to know you all through an animation course at ACCAD. I could see a way in to working together because you had generated data from the choreographic structures in the dance. That data is spatial (something of interest to geographers) and also qualitative in that you know what the dancers are doing. The equivalent in geography would be having GPS data (tracking where someone goes) combined with GIS data  (telling you who they are and something about what they are doing).

Let me say this another way, the data you created from the dance (One Flat Thing, reproduced) is in many ways similar to what we would expect from looking at individual activity patterns in a real-world situation, but that type of data is still very hard to acquire. Because we have such a rich and contained data set I believe we can use this as a model for the real world or as a surrogate for many types of questions and method development. We’re just starting this research so the specifics aren’t clear but the potential is. Some of the things I’d like to see in further looking into this data is to include the narrative descriptions from dancers or other people involved in this project. As geographers we often try to mix quantitative methods, such as the spatial data we use in the Movement Density object, with qualitative analyses from interviews and texts. Again I think that this data give us a good opportunity to come to grips with some of the complexities of such analyses. It is almost as if we have a controlled environment for qualitative inquiry, similar to what the ‘hard science’ tradition has capitalized on for more quantitative hypothesis evaluations.

NORAH: How do you think the research we’re doing in Synchronous Objects is relevant to geography specifically?

OLA: I see it mainly as a channel for communicating the idea of using geographic theory in another context. After working on this project I now realize that there is much more depth and things going on in a dance than I ever imagined. This could impact other geographers as well. I am learning new things each time I sit down with the other people in this project and hopefully that experience will be the same when people see the website. And for me it was the use of space and dynamic, spatial interaction between actors that struck a chord, but others may find completely different dimensions, and that’s what is so exciting.

Also, the ideas that currently emerge are intriguing directions of research from a geographic perspective. We have identified some new research ideas that probably will generate publishable results both as stand-alone geographic research and probably also as cross-disciplinary contributions to the literature. The Movement Density object we use a spatial summary statistic called density surface that is a common tool in geographic visualization. In this case, it helps us explore patterns of where the dancers spend most of their time. As we follow the dance we can see how certain areas are used more than others by the dancers. After a while, hot-spots, or places that were most used by the dancers, show up as intense, brown-red and places with little activity remain in blue shades. The density surface can also be illustrated as a topographic landscape where the count of points is used as an elevation value, creating a landscape of mountains, peaks, and valleys. In the topographic map the hot-spots, are found as mountain tops or ridges, and the deep valleys and flatlands represent little or no dancer activity. You can see all this in our explanatory video on the Movement Density object and I talk about it there too. Because we have such a contained data set we can really use this as a model for the real world.

Another direct outcome of this research is that my student Hyowon completed her dissertation using her work on this project and it helped her get a job as a faculty member at another university. I’m excited to see what else emerges as we continue to collaborate.

–Ola Ahlqvist, assistant professor, Department of Geography


Choreographic Objects e-Symposium


As we go live with our Beta version of the site we’re also kicking off the blog with contributions from the participants in our symposium held here in Columbus, OH at the Wexner Center for the Arts from 3-5pm on April 1. We’re thinking of these posts as our “e-symposium” for audiences outside of Ohio and we invite your comments and questions.

You can also view the live web-cast of the event at by clicking here.

Here is some more info about the symposium:
As an American working internationally for the last thirty years, William Forsythe is recognized as one of the world’s foremost choreographers. His work is celebrated for reorienting the practice of ballet from its identification with classical repertoire into a dynamic 21st-century art form. In recent years and with the establishment of The Forsythe Company, he has extended his choreographic thinking into new forms such as digital media productions, installations for gallery and public spaces, video, and publications. For Forsythe, these projects are part of a larger sphere of interest he terms “choreographic objects.” The idea of a choreographic object allows for the transformation of a dance from one manifestation (the performance on stage) into an array of other possibilities (such as information, animation or installation). This transformation requires cross-disciplinary collaboration and makes space for new connections between art and science, practice and theory.

In this symposium, Forsythe and his collaborators will discuss how the idea of “choreographic objects” took form in Synchronous Objects for One Flat Thing, reproduced, a new interactive web project created collaboratively by Forsythe with Ohio State’s Maria Palazzi (Advanced Computing Center for the Arts and Design) and Norah Zuniga Shaw (Department of Dance) and an interdisciplinary team of collaborators from across the arts and sciences. To celebrate the launch of the project, invited outside experts will contextualize the project in terms of its relevance to current trends in the philosophy of cognition and architecture. Participants will respond to posts and comments on the blog.

We look forward to hearing from you!
—Norah Zuniga Shaw


Noel Cressie on Dance and Statistics


NORAH: One of the goals of our project is to produce research resources (objects) that can deepen the possibilities for interdisciplinary connections focusing on dance. This will evolve as a wider public engages with the site. But we began testing the possibilities for collaboration by reaching out to a number of different people here on the Ohio State campus. How would you describe your involvement in these explorations?

NOEL: Statistical Counterpoint had to happen. The source video of Forsythe’s “One Flat Thing, reproduced” had been quantified – into cues, movement material (themes and improvisations), dancers, stage locations, times, and so forth. The Department of Dance, ACCAD, and William Forsythe had transfigured the dance piece into many 17-dimensional vectors indexed by space and time. As soon as I saw some of your Objects in progress at an ACCAD open house a year ago, I felt that there was something I could contribute.

NORAH: Yes. And your involvement made sense to me because our focus has been on looking at the spatial and temporal patterns in the choreography and finding new ways to visualize them. These are the kinds of things you’re concerned with in your discipline as well, although not usually focusing on dance.

NOEL: Right. I’m a statistician with research interests in spatio-temporal modeling, applied mainly to the environmental sciences. I had worked with forest ecologists in the mid 1990s; we had come up with a way to visualize forest health using a spatial dataset collected at monitoring sites spread throughout New England. We found a way to link regression plots with spatial maps. Then, by “painting” or “brushing” points of interest on the plot showing the regression (the “multivariate view”), we could see where those monitoring sites were located on the map (the “spatial view”). The dance data were more complex, many more variables, and time added another dimension. But, the same principles, of conditioning first (brushing) and then looking at how the dependencies change with different brushings, remained.

Ola Ahlqvist from the Department of Geography, and his student Hyowon Ban, were already working on a prototype Object using Geographic Information Systems (now called the Movement Density object), and I suggested a collaboration that would take a statistical approach to visualization of the data from “One Flat Thing, reproduced”. My idea was that the need for highly multivariate views could be solved using Parallel Coordinate (PC) plots, and the extra complexity of the time dimension could be incorporated by either adding it to the PC plot or by creating a temporal view. It was an interesting road to follow with some potholes and detours, but we got there.. We used the software called GGobi, which is in the public domain. It has more capabilities (such as 3-D rotation) than we show in our Object, and there are clearly other statistical objects in waiting.

NORAH: As the lead for the data creation on the project, I also found it really productive to have you and Ola around to discuss issues I was having with quantifying qualitative information, the lessons we were learning about how we would do data collection in future projects and quality assurance issues that are common areas of discussion in your professions. I remember at one point when I was explaining the many shifts in terminology and the evolution of the data you said something like “yes of course, this is a pilot project you can expect those kinds of issues’ and it was such a relief to hear. Even terms were useful. For example, Ola gave me the term “attribute data” for our data that describes the dance. On the other side, one of our hopes with this project is that these different disciplinary lenses into seeing structure in the dance might change the way it is viewed. Kind of train the eyes of the collaborators and audiences. Has this been true for you?

NOEL: I think the dance piece is quite stunning, speaking as someone who enjoys modern dance but knows very little about choreography and what it takes to turn it into a performance. I have learned from this project that choreography is very complex, and much of it must be intuitive because it looks too hard to build up from “atoms”. As a scientist, I’ve always believed that science can be artistic. This is an exciting project because I’m working with artists who believe that art can be scientific! I wanted our Object to be stunning in its own way, even as a prototype. I also wanted it to be useful to Bill [Forsythe], to inform him of the consequences of decisions he had made when he choreographed the dance. There is something illuminating and liberating about the spotlight that data and its statistical analysis can shine on things felt intuitively.

Bill also wanted Statistical Counterpoint to be something I could use as a teaching tool in my statistics classes. Statistics is the science of uncertainty. Quite a few people, particularly during election season, think that statistics’ role is to summarize collections of numbers for people to use or abuse! You’ve heard the words, “lies, damned lies, and statistics”…well the reality is that the statistics profession works hard at “the truth, the whole truth, and statistical inference”. We let the data “speak” that truth in myriad ways. Mathematics, probability theory, and computing all play a role in removing the unimportant, leaving behind the essential. Properly chosen statistical summaries can do that. Those summaries are usually most insightful when they can be visualized.

NORAH: Can you talk a little about the Statistical Counterpoint object and some of your initial findings? I know this is just a beginning but I think you’ve found some nice early insights.

NOEL: There are two objects within one and you see them both in the explanatory video. The first is the “Statistical Counterpoint: Cues.” It is a simple object that shows the power of linked visualizations. By brushing on Dancer, we can see dancers’ roles in Cue Giving and Cue Receiving during the dance. We can also see where each dancer gave and received cues and when they did so. Here, a reduced dataset was carved out of the original one; this demonstrates what statisticians call marginalization. Further, brushing is a way to visualize what we call conditioning. In this object, we condition (brush) on Dancer.

Three basic views of the dance are shown. The first is a multivariate view based on the PC plot, the second is a spatial view based on the location of the dancers on stage, and the third is a temporal view showing, for example, when a cue was given in the dance. At the beginning of the eplanatory vidoe for our object, we show all lines and dots (in green), without the brushing capability turned on. The parallel coordinates on the multivariate view are Cue Giving (the left-hand vertical axis),
Dancer (the middle vertical axis), and Cue Receiving (the right-hand vertical axis). Lines represent relations between dancers and their cueing activity. Dots on the spatial and temporal views are linked to the lines shown in the PC plot. If you watch the explanatory video we made you can notice that the density of lines for Cue Giving is noticeably less than the density of lines for Cue Receiving. This makes sense since a cue given can be received by many dancers. Also, the spatial view showing the density of points on stage for Cue Giving is much less than that for Cue Receiving, reinforcing what we just observed on the PC plot. Second, we notice that
very few cues were given from behind all the tables, but there appears to be a fairly uniform distribution of locations where cues were received. Third, from gaps in the sequence of dots in the temporal view, there were times in the dance when no cues were given at all. These gaps appear to be longer in the middle of the dance than at the beginning or the end.

Then the brushing starts, and the lines and points light up according to each dancer’s role during the dance. We start with Amancio and proceed with the other 16 dancers in alphabetical order. We have a “Cue activity counter” at the top to quantify what is being observed -some dancers have noticeably more involvement in cueing than others. Sang’s count was the most at 86, and he gave roughly as many cues as he received. David’s count was the least at 17, all of which involved his
receiving cues (there were no brushed dots for him in the spatial view of Cue Giving).

NORAH: This is great. One thing to note is that for this version of the dance shot in 2005 Bill added a few dancers (usually it is for 14 dancers, this version is for 17) and David was one of the people he added in for just this video shoot. So that explains why David doesn’t give any cues.

NOEL: It would be interesting to share more of that information in future blog posts.

NORAH: Yes, that’s part of what this blog space is for, absolutely.

NOEL: Anyway, there is a lot more in these plots I’m sure, and this is just the simple object! In the “Statistical Counterpoint : All Systems” a more complex data set is included. It shows all the data, without marginalization. Familiarity with the simple object is helpful for understanding the complex one. In this one, time is now added as one of the parallel coordinates and shades of green have been chosen to reflect the amount of time elapsed during the dance; the darker the green, the more time has elapsed. Conditioning (brushing) is on Theme in this object.


“Statistical Counterpoint” is best viewed as a dynamic animation, where the dynamics are not necessarily derived from the time coordinate. The animation can be stopped, backed up, and studied. The power is in comparing the brushing patterns as the dynamics evolve.

NORAH: What do you think would be the next interesting directions to go with this prototype?

NOEL: My questions include: Can the Statistical Counterpoint Object be made interactive? Can we direct the software to answer specific questions, like at what times do certain dancers give and receive cues from each other? Or, do certain dancers perform certain themes more? You answer these questions in many of your other objects but I wonder if our statistical analysis could provide new information on patterns and interactions. Or, does the statistical dependence pattern, shown in the multivariate view, change in the middle of the dance versus at the beginning or at the end? I look forward to continuing our research and see the potential to publish some of the results.

NORAH: Thanks Noel.