It’s great that Bruno Latour has come up in the scale debates so early in my reading. I haven’t had a good reason to think about him since I stopped being a science librarian. Even then, it was pretty casual thought.

The last time I did any serious thinking about him, I produced a piece that encouraged librarians to use the analogies of Actor-Network Theory
to get some work done. In the first week of the scale seminar, he was brought up in passing, and I was looking forward to see how his work has been used in geography. This week, a review article seems to completely miss ANT’s point.

The article says that discussions of networks “simplify/homogenize the role of geography as a form of knowledge/practice and as an institution.” Well, that’s what ANT was formed to do from the ground up: to take all factors in a given situation and to treat each of them as equal, individual actors in a network. The idea that an academic discipline should be privileged over a printed map or a feature of the landscape is anathema to the ontology that Latour has designed. To say (as Paasi does) that “Geography, boundaries and scales are not ‘intuitive fictions’ and their rejection/acceptance can hardly be a matter of the choice of a specific ontology” rejects using any of ANT.

For sure, there are other network theories that can be used instead. Graph theory has a host of methods that can be applied to geography (network mapping of any sort: whether commodity flows or transport networks) without directed graphs. Social networks are similar (and I assume don’t pre-date mathematical graph theory, so I imagine social science borrowed them from math across the board).

Privileging geography, boundaries, and scales as ‘inherently real’ is specifically what Latour and ANT reject. While Latour may do so playfully, he is serious when he builds a knowledge system that attempts to NOT privelege any player over any other. For Latour (and Science Studies, the discipline for which ANT was built) all knowledge is created, and none exists outside of a system. Therefore, geography, boundaries, and scales are all just nodes of a network to be analyzed: not meta-concepts that must be taken for granted as essential to the system. To say otherwise is really being an unreconstructed positivist (which I say only as one who has been accused of the same).

So yes, Latour argues that Pasteur ‘invented’ brewers yeast and bacteria. He also turns that language around and argues that it is just as valid to say that microbes discovered Pasteur. He’s playing with language folks: let’s not take this stuff too literally, but let’s try and understand the message. Pasteur, brewers yeast, and the French Academy are equal actors in the network. All affect the others.

I think, in the end, ‘network geographies’ are something quite different from Latour’s networks. Perhaps the terminology is too confusing. Latour’s networks are subsets of network and graph theory. When I talk about ICTs and networks, I’m going to try to be very careful to distinguish between computer networks and actor networks. I might have to invent a new word. Or maybe someone already has. Actually: using the ICT acronym goes a long way toward eliminating the use of internet and the hyper-generic ‘network.’ Maybe that’s why ICT got invented in the first place.

And for what it’s worth: geographers have been known to get basic facts about Latour wrong. For example: in the ‘Dictionary of Human Geography’ Latour is identified as an engineer, although his training is in philosophy and anthropology. He is emplyed as a sociologist. One of the first groups that he studied, however, was engineers. And, if I remember correctly, one of his earliest books is a Socratic dialog in which one of the speakers is a young engineer. I can understand the confusion, but a sloppy dictionary entry?

Two more down and two to go for this week. I’m finding a little bit of sloppy language. Or is it sloppy thinking?  Maybe it’s just sloppy examples:

Gentrification is typically localized to small areas of the inner city, and foreign direct investment occurs at the international scale.

That comes from the introduction to McMaster and Sheppard’s 2003 Scale and Geographic Inquiry. They use this example when explaining one of the typical definitions of scale (a level of operation for some phenomenon). Later in the chapter, to their credit, they wind up contradicting this ‘typical’ explanation when describing the production of scale ala Smith and Swyngedouw. Swyngedouw in particular argues that global economic forces are actually increasing the power of sub-national metropolitan areas at the expense of nation-states. One only has to look at Guangzhou to see how this blows away both ends of the earlier example. FDI has created a completely new local landscape.  Of course, it was the nation-state that allowed the FDI to occur, but largely development in China has been driven by urban-level (I almost said scale–silly me) political and corporate structures.

This explanation of Swyngedouw’s ‘glocalization’ makes me wonder what happens to the urban form (heck, what happens in any arena) when a small town becomes the headquarters for a global company. How does Bentonville, Arkansas function as a global city?  I don’t expect to find examples like that in China.  The capitals of even the smallest provinces have multi-million populations.
Andrew Herod’s “Scale: The Local and the Global” opens with another annoying example. I can only hope he was doing it self-consciously tongue-in-cheek. He mentions the “French farmer who attacked a McDonald’s” as an example of “localist tendencies among those who have sought to defend traditional ways of life.” Granted, the man might be a farmer, but the staged attach on the McDonald’s was hardly a local act. Consider for a moment using ‘moving within a small geographic sphere’ as a definition for local. If we can, then Jose Bove has never been local. He lived in the SF Bay area while his parents were doing research at UC Berkeley.  In addition to Northern California, he has protested with Greenpeace in the Pacific, against the WTO in Seattle, and is now a candidate for president of France.

Granted, my operating definition of local is totally contrived. Why should ‘local’ mean untraveled?  I might even be guilty of classism by hinting that the son of elite researchers by definition can’t be local.  Was I conflating ‘local’ with uneducated as well as untraveled?  By my argument I am almost as global a citizen as Bove. Herod goes on to make a great connection between Marxist materialism and scale, stating that the local is just as produced as the global: each is a contrivance. The re isn’t necessarily a ‘natural’ scale to any social, political, or economic phenomenon.

Kevin Cox is brought into play here, who argues that actors live inside of one set of spaces of dependence while they come together and “construct associations with other actors” in spaces of engagement.  Some argue that this is a binary relationship between global and local, but I think I agree with those who argue that it’s not quite so restrictive.

Speaking of binaries:  Gibson-Graham has a set of binary relationships that are listed under ‘discourses of the global and local.’ Latour comes up here, and I swear I didn’t know he was going to appear on the scene when I wrote yesterday’s post.  The binary for Latour’s networks is not between local and global, but between shorter and longer (or simple and complex, if you will) networks. Regardless of which set of assumptions we want to play with, both articles agree that some scales become more important (read as: ‘written about’) because they feature prominently in the discourse.  We do research and write about them because the culture points us towards them.