Saraceno, Latour, Networks, Complexity and Visual Complexity

As I discussed in an earlier post,  I am in awe of the work of artist and architect, Tomas Saraceno.

My goal here is to discuss Saraceno's work in greater detail, largely by focusing on Latour's article on Saraceno, which, by definition, moves to a larger discussion of complexity art, visual complexity and complexity science in relation to the topic of globalization.


Saraceno's work impresses me for several reasons:

First, it is visually very powerful.  The time involved in creating his networks or his geometrical shapes, for example, is incredible.  They are also very mathematically appealing.

Second, like many artists today that I think of as doing some type of complexity art--from Benjamin Edwards to Julie Mehretu to Matthew Ritchie--he demonstrates through his novel visual language how macroscopic systems emerge through microscopic relations, in the form of what can be called generative art. In Saraceno's case, this is done through his creation of networks, assembled one link at a time, or through the connecting up of various geometrical forms.

Third, his work visually and metaphorically addresses--sometimes directly and other times indirectly--the intertwined issues of cities and sustainability, architecture and ecosystems, globalization and complexity and, ultimately, our complex interconnectedness with one another, the earth and the universe.  Click, for an example, here.

Fourth, his work indirectly deals visually with the topic of complexity and complex systems.  I have not read enough about or gone through enough interviews with Saraceno to know how interested he is in complexity.  But, his emphasis on visualizing networks (even in the form of spider webs) certainly allows for easy exegesis to the field of complexity science.

In fact, it is the exegetical ease of Saraceno's work that seems to have led Bruno Latour, the famous French scholar, to write his article SOME EXPERIMENTS IN ART AND POLITICS--originally, as best I can tell, published in the visual arts journal, e-flux.

Saraceno (or those working with him) must find in Latour's article a useful "written" narrative because they have it connected, at least on the internet, with several of Saraceno's shows, including his recent installation--CLICK HERE--at the Metropolitan Museum of Art in New York City.

I thought a few comments on the article, therefore, worthy, given Latour's reputation (e.g., his development of actor-network theory and his role in the sociology of science and postmodernism writ large) and the relevance of Latour's argument to the work many of us are doing at the intersection of complexity science and complexity art, particularly in connection to issues of globalization.


Reading Latour's article took me back to my 'graduate student' days when I first became involved in the sociology of science.  I always found Latour's ideas provocative; they really made me think, even though I went more in the direction of Michel Foucault, Richard Rorty and, ultimately, complexity science.

Here is my take on the article--best, however, to read it, as it is only a few pages. (SOME EXPERIMENTS IN ART AND POLITICS)

First, it must be pointed out, the article is not about Saraceno.  Instead, it uses Sarceno's work, along with a few others, to make a point about how art can help to create a new politics.

As I see it, Latour sees in Saraceno's art the ability to conflate science, philosophy and art to create room for a new form of 21st century politics, one Latour sees grounded in composition.  As best I can figure, for Latour, compositions--like those of Saraceno--are powerful because, in their acts of conflation, they flatten hierarchies, particularly those hierarchies getting in the way of us understanding globalization--a term Latour finds empty.

For Latour, the term "globalization" is empty because, in his mind, those articulating its theoretical frame (such as complexity scientists, environmentalists, etc) have yet to make a number of important moves to get to what, for a lack of better word, is some type of compositional, post-human politics.  Such a compositional, post-human politics, on the one hand, blurs the 'system' boundaries between social systems and ecosystems, human systems and animal systems, the sciences and the humanities; while, on the other hand, integrates the notions of habitation and connectedness, the local and the global.

To make his argument, Latour breaks his essay into three sections.  The first section deals with what he calls the 'networks versus spheres' debate--a debate that, to my knowledge, Latour derives from his 'reading' (translation through transportation) of the work of Peter Sloterdijkm.  It is this section where he talks about Saraceno's work.

The second section deals with the sciences versus humanities debate--again, Latour has a very specific view of this hierachy via his re-enactment of Einstein and Bergon's famous debate.  Here Latour is concerned with the ability of science to handle the nuances of subjectivity and the blurring between society and nature.  And, the third deals (via Donna Haraway and Isabelle Stengers) with an example of a post-human approach to politics, one that tries to enact an example of composition--and one that, it must be made clear, has nothing to do with Saraceno.

Because my post is focused on Saraceno, my comments will focus on Latour's first section: networks versus spheres.

Let me say immediately that, overall, I think I get Latour's point.  Dealing with the massively complex social problems facing us at the beginning of the 21st century requires new ways of thinking, many grounded in composition, collaboration, combination, assemblage, merging and juxtaposing differences.

However, I find that, for the most part, Latour has constructed a bit of a straw person by not telling us who or what group he is specifically discussing.  One cannot make dismissive statements about an entire field without a bit of evidence.  For example, unlike Latour, I think that many complexity scientists, specifically those working in globalization--such as Manuel Castells and John Urry--are already doing a variety of the creative compositions Latour seeks to promote.  And, as far back as 1996, Fritjof Capra (in his book, Web of Life) was making Latour's argument through his discussion of deep ecology (Arne Naess) and ecofeminism (Fran├žoise d'Eaubonne), plus without all the Sloterdijk metaphorical language.  Another example is the embodied mind work of Francisco Varela.  Or, how about the cutting-edge work on the sociability of bacteria being done by Eshel Ben-Jacob?  There is also an increasing network of artists, some of whom I have discussed on my blog (e.g., visual complexity, Benjamin Edwards to Julie Mehretu to Matthew Ritchie), who are blurring distinctions between society (socio-systems) and ecosystems, humans and animals, science and the humanities.  Still, to defend Latour a bit, he is right that certain forms of ideological or economic or environmental globalization, as Joseph Stigliz often critiques, remain firmly entrenched in a modernist paradigm, as one sees throughout the global banking crisis or the global warming debate.

Despite these critiques, I very much like Latour's descriptions of Sarceno's work.  In Section 1, Latour focuses on Saraceno's show, Galaxies Forming along Filaments, Like Droplets along the Stands of a Spider's Web.  I have read these descriptions several times and I find more in them each time.

But, even here, when Latour is discussing Saraceno, I find myself once again struggling.  What I do not get are the concepts Latour uses to make this argument.  Specifically, I have a hard time with the concept of spheres and other related terms (e.g., envelopes, etc) that Latour borrows from the German philosopher Peter Sloterdijk.  By using Sloterdijk's work, Latour translates through transport a series of metaphors in which I easily get lost, as I search for their empirical utility outside the simpler points they seem to make, which do not require metaphor.  Said another way, metaphor works best when one cannot otherwise make a point.  But, complexity science, more specifically, network science, already has a very good and empirically high-yielding vocabulary connected to method that can easily be used to make Latour's point. 

For example, the main concept Latour takes from Sloterdijk into the vocabulary of network science is spheres.  For Latour,"The word “network” has become a ubiquitous designation for technical infrastructures, social relations, geopolitics, mafias, and, of course, our new life online. But networks, in the way they are usually drawn, have the great visual defect of being “anemic” and “anorexic,” in the words of philosopher Peter Sloterdijk, who has devised a philosophy of spheres and envelopes. Unlike networks, spheres are not anemic, not just points and links, but complex ecosystems in which forms of life define their “immunity” by devising protective walls and inventing elaborate systems of air conditioning." 

When Latour employs this type of metaphorical language I think to myself, 'hey, these spheres he talks about, they seem to be a type of weak-linked or strong-linked component; that is, they are networks within larger networks; with no hierarchical structure; where the inside of these networks is conditioned upon their connections to the outside.  So why not just say that?  Or, spheres seems to get at the idea that nodes are real things: people, animals, plants, buildings, social groups, organizations.  In other words, these nodes and their links are their own complex systems that, through their relationships, create the conditions in which they live, just as a small group of friends or animals or fish or plants do to help each other make it through the daily grind of life.  Again, why not just say that?

Here, yet, is another example from Latour's argument.  In the next paragraph Latour states: "The two concepts of networks and spheres are clearly in contradistinction to one another: while networks are good at describing long-distance and unexpected connections starting from local points, spheres are useful for describing local, fragile, and complex “atmospheric conditions”—another of Sloterdijk’s terms. Networks are good at stressing edges and movements; spheres at highlighting envelopes and wombs."

This time, with this statement, Latour is just flat wrong.  For complexity scientists, networks are scientific maps that do not, unless tied to an actual set of geographical coordinates, have any inherent spatial meaning--in other words, and this is a very important point, local and global are not spatial concepts in networks; they are relational concepts.  Distance is entirely relational.  Furthermore, as I suggested above, networks do a very good job of highlighting boundaries and revealing the internal 'spheres, envelopes and wombs', as Latour, pace Sloterdijk, calls them.  Easy examples are plant networks or friendship networks, or, perhaps, the obesity networks studied recently by Christakis and Fowler, which they found reveal internal groups, cliques and communities within the larger network, that, over time, 'wall out' or 'wall in' healthy eating or exercise behaviors.

I can go on, but i will stop.  I think I have made my point.  The metaphorical language Latour employs, to me, is unceessary and, in my mind, confuses things.

Again, having said all that, I still like Latour's article.  And, I think he is on much more solid ground when he focuses on describing his experience of Saraceno's work.  In fact, in doing so, Latour admits that spheres are really just networks.  See, when you look at Saraceno's work, Latour is right: spheres are little more than the densely connected sub-networks where movement along its edges "might' slow down due to the density of connections; but these sub-networks are nonetheless nested.  That is an empirical point worthy of study.  Can, for example, a sphere be a type of network that has a certain density--some sort of tipping point or threshold level--that protects the nodes in it, whatever these nodes be (fish, people, social groups, etc) from outside influence, as in the case of contagion spread?  Can a sphere network (for example, some type of socioecological complex system such as a local fishing community and all the living things connected to make it up) act as a wall, an immune system against wider movements or trends?  Or, conversely, as we see in poverty traps--see, for example, Bowles, Durlauf and Hoff--can such a density threshold 'wall-off' a community from socioeconomic well-being and improvement?  Or, is density really able to slow globalization or other processes down, or is this just a myth?

Critique aside, these are the types of things that Latour and Saraceno's work inspires me to think about.  And, at the end of the day, that is what good art (written or designed, compositional or not) does.