Zombie GIS

Ask most people what is GIS and they’ll not be able to tell you. Ask most people in a GIS class (such as the Intro GIS class I am teaching this spring), in a geography or allied field, and they’ll likely tell you a bunch of things that would have heads nodding at Esri: it is software, data, visualization, mapping and analytic capabilities for spatially referenced data. (Here’s Esri’s own answer.) If I ask most of our graduate students in our department (famously grounded in social theory as it is) they’ll likely tell me it is something like ArcGIS: Big GIS, costing thousands of dollars, hard to learn, for specialists only.

All of these, in different ways, are wrong to the point of dangerousness. Some are out of date. Some are irrelevant to the realities on the ground (hem-hem, including those of our grad students). All of them, it seems to me, refer to something that may as well be considered dead, but won’t die: zombie GIS. So who killed it?

To answer that question, a clarification is in order. I am not saying that the things we used to call “GIS” are no longer relevant, critical, or worth pursuing as a career. Quite the opposite in fact: GIS has gone viral. It has been injected into the body politic and infected nearly everything from the economy, to politics, to everyday activities such as going to the shops. In doing so, however, it can no longer be called GIS. Much like the walking dead, it is not what it once was, but doesn’t seem to know it: call it “zombie GIS” a phrase I confess to having just made up. (Fans of zombies have sometimes studied them using GIS, as these great projects illustrate. However, that’s not what I’m saying–I’m saying GIS is a zombie.)

So what has it become? Putting aside whether it has evolved or simply been replaced by other developments, here are some new key terms I think we need to come to grips with:

–software-sorting of geography, or, code/space
–geolocational (privacy | tracking | surveillance)
–geopolitical digital economy
–cyber materialities
–geolocatable knowledge economy
–Big Data biographies, the “spatial self”
–Internet of Things, smart city
–“lively” Big Data
–algorithms and calculability
–governmentality of “data derivatives,” data shadow, data-image, metadata

And so on. It would be interesting to put together an “Intro GIS” course syllabus that delved into these topics; a couple of weeks on the genealogy of the algorithm, say, followed by practical exercises in tracing how value is produced, extracted and enhanced as a geopolitical assemblage, capped off by a project creating geoweb-based mappings, CartoCSS and mobile-collected data. Even if the Chair of my department wouldn’t balk at such a proposal, perhaps it would be pointed out that this isn’t GIS. Exactly. But such a notion sustains something that’s dead; hence a zombie.

Zombies are interesting to many writers because they’re neither one thing nor the other. They occupy a liminal, in-between state, and as such you don’t have to close off possibilities by choosing one thing (life) over the other (dead). In fact, the one thing may allow the other. (Think dialectics.) Thus for the Slovenian philosopher Zizek, zombies are paradoxically both creatures of habit and yet illustrative of the freedoms we can attain due to those constraints. Like when we speak by learning rules of language, this then allows us to write poetry (or blog entries). Zizek says:

What this means is that what Hegel says about habits has to be applied to zombies: at the most elementary level of our human identity, we are all zombies, and our “higher” and “free” human activities can only take place insofar as they are founded on the reliable functioning of our zombie-habits: being-a-zombie is a zero-level of humanity, the inhuman/mechanical core of humanity.

If this is true, then perhaps the “rote” and “habitual” aspects of zombie GIS are what have prompted the list of developments I give above! So perhaps we might not have CartoDB if ArcGIS had any kind of real web capabilities.

(Another reason for the attractiveness of zombies to critical theorists: it echoes the ghost-talk of the famous opening line of the Communist Manifesto: “there’s a spectre haunting Europe…” which in turn feeds off Shakespeare. Derrida made so much of this it became a subdiscipline: “hauntology.”)

So what are these things in the list doing, and in what sense do they attract our attention? One place to begin is with assemblage, a term I’m using fairly straightforwardly here to indicate a mess of intersecting interests, practices, policies, epistemologies and so on. Fortunately Rob Kitchin offers us a look at a data assemblage in his new book The Data Revolution:


As you can see, there’s quite a mix of things here. the best article I’ve found so far to guide through these issues is Jason Dittmer’s “Geopolitical Assemblages and Complexity” paper. Drawing on assemblage theory (Deleuze, DeLanda, Latour), Dittmer argues that we need a more “materialist” approach, specifically that of assemblages. These are conceived as “wholes characterised by relations of exteriority.”

There are two things which are useful about this in terms of the list above. First, it implies the necessity of not just looking at the component parts. Second, Dittmer says that what’s important is not so much object with properties (how GIS sees the world, a basically Aristotelian logic) but parts with capacities to do work in the world (“actants”). This is sometimes known as critical realism.

As Thrift pointed out in his book of collected writings on capitalism published ten years ago, the modern knowledge economy can be traced back to the late eighteenth century. This coincides with what Foucault called the rise of the “disciplines” organized sets of knowledge with rules on what counted as data, ways of explaining (whether correct by today’s standards or not) and practices of disavowel (eg of the abnormal). For Thrift, what’s interesting about capitalism today is that these knowledges have been bent back to apply to capitalism itself. (Ian Hacking has described this “looping” effect in other contexts of knowledge about people.) So we get business schools and Masters degrees in finance; topics which are not traditionally academic subjects. (GEOINT Certificates may be understood in this context as the nearest equivalent in geography.)

This turn to materialism could be read as a turn away from discourse, but I don’t think it’s as simple as that. Let’s say that it creates a space for a new discursive materialism. This is important because one often hears people talking about “cyber” in the same way (that’s it’s non-material). But cyberwarfare, for example, has real material effects (computers explode, freeze or send hacked info to the attackers). So, this theoretical perspective is useful for “geospatial” (as a lot of people call what used to be called GIS).

I am not an assemblage fanboy. I don’t really like Latour, I’ve only recently started reading DeLanda and I don’t particularly “get” Deleuze–although I hope to read Thousand Plateaus more thoroughly for a seminar I’ve proposed with my colleague Jeff Peters (on “Big Data Narratives”).

However, it seems to offer some useful approaches, and, if you can cut through the (relatively sparse) thickets of French social theory speak, some suggestions for research projects. The one weakness of it is avoiding what Thrift calls “slab theory” or the temptation to use assemblages to study everything, at all scales. If the geoweb (to take an example relevant here) is political, economic, legal, geopolitical, performative, embodied, etc., well–you can see the task becomes very daunting and complex (see Dittmer again on complexity).

One of the few people working in this vein is Agnieszka Leszczynski (eg., “Situating the Geoweb in Political Economy,” Progress in Human Geography 2012)Leszczynski argues that the geoweb (the collection of geospatial services and technologies online) is more than just new forms of technology (web 2.0) but actually what she calls “technoscientific capitalism” which if not an assemblage is certainly a cross-cutting point of view. Or again in her most recent article, she insists:

There is a pressing need, therefore, for theoretical, empirical, and conceptual apparatuses for apprehending and evaluating the implications of and extent to which networked location-aware devices and spatial content have assumed pervasive presences in individuals’ daily lives, and of the material effects of associated spatial big data-based productions of living.

This is the notion that reality is always the product of the myriad intersections and mutual constitution of technology, society, and space relations that are themselves the products, or effects, of mediation.

Notice the everyday level of her focus on “techno-socio-spatial-relations” (p. 4). According to her, understanding these developments as media(ted) gives:

a position from which to not only engage with these presences as inherently material, but also to divest ourselves of GIS as an epistemology for ‘reading’ emergent technologies by designating how particular software, hardware, and information artifacts are being brought together in ways that are more-than, and other-than, GIS.

I think this is right for reasons I’ve been saying above. Zombie GIS, undead as it is, cannot provide the epistemological ground for apprehending what we’re interested in. Nevertheless, out of habit, GIS lumbers on!

3 responses to “Zombie GIS

  1. Okay, that’s your next book project sorted – can’t wait for full length ‘Zombie GIS’ to hit the shelves.

  2. Pingback: Lyzi Diamond on what to learn first about mapping | Open Geography

  3. Pingback: There is no matter here – the materiality of the Internet | Lines on Land

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.