It’s always of concern when a journal article opens with a vague statement. So when What does Google Earth mean for the Social Sciences? opens a claim that cave painting can be equated to maps, then one has to wonder about the explantory power of the word “map”. Is it a representation of geography, of ecology, of political boundary? I’ve never seen a cave painting that could be classified as a map, unless we’re going to so broaden the definition in both symbolic and analytical ways so that we might as well include any image of any thing. I realize it’s a throwaway intro remark, but the lack of specificity is worrisome. The article points out that hierarchically invested geographies have theoretical consistency in urban, political and environmental fields, and then goes on to describe Google Earth. But when it comes to Google Earth, all I can focus on is the simplicity of the tool itself, which seems to enforce a methodological simplicity on its users. Sure, you can build kmls till the cows come home, but the “lack of analytical functionality in Google Earth” means one can do little more than display data.
The difference between discrete objects and continuous fields underlies Google Earth’s orientation toward an audience interested in looking at a map that is a static spatial reference point for various phenomena to be placed upon. The map exists, as a classic map, underneath the data, and is considered an external, agreed-upon norm for the purpose of discussing the really interesting subjects, which exist on top of the map. Whereas in GIS the raster or continuous field data is as easy to manipulate and analyze as the point and polygon data, in Google Earth it is the points and polygons where any manipulation will occur. Doesn’t sound so bad, really, until we start to think beyond the surface of the Earth being simply topographical. Land use change, environmental change and all those transportation and urban networks represented by Google Earth have an effect on what is presented as static data. Deforestation areas, available as polygons, or sites of past flood levels, available as points, in reality overlap these and raster data areas, either periodically or permanently, and Google Earth reinforces a worldview wherein this is not the case. You’re using a static tool consisting of overlays to represent an acknowledged dynamic socio-political/socio-environmental object whose various (arbitrarily mediated) fields are in constant interaction and do not sit neatly upon one another, without mixing (Are there miscegeny laws against the interaction of different data from different disciplines, did nobody tell me?).
This doesn’t mean Google Earth is trash, of course. You can still examine the horizontal and vertical spatial relationships of point and polygon data. And you can see how census laws get circumvented by certain coastal California cities. But, ultimately, and as with most consumer products, its assumptions reinforce an ontology that may hinder knowledge generation by unnecessarily limiting the scope of questions asked or relationships examined. Maybe that’s why the current crop of spatially-informed peer projects is so, well, bad. Wikimapia seems little more than an application waiting for a meme bomb; there’s still no Wikimaps or Wikiatlas (and why would you even propose a Wikiteer when Wikipedia already contains its wealth of geo-referenced information in gazetteer format?). I’m sure I’ve mentioned that Hypercities is terrible, but it bears repeating. Open Street Map seems to have the community support, but rather pedestrian (if you’ll pardon the pun) aims. I realize not everyone has the time or inclination to pick up a copy of ArcGIS and start creating hillshades, but isn’t there a middle ground in the manipulation and collection of spatial data between the superficial and the highly technical?
What we need is a serious, peer-collaborative, spatially-oriented agent-based model. It’s not as crazy as you think. First off, people are familiar with modeling, they just don’t realize it. People just call them games, especially the sandbox games. Given a space to play in and describe the interactive, dynamic processes that define our existence, I think we’d do wonders for improving general human knowledge (Which can’t be bad).
It should start small, with established principles, by creating a commons-based model of the Earth’s environmental systems. Scientists have been building these for years, and creating a scaled-down (But not dumbed down) visually rich version that anyone can play with (And, more importantly, contribute to, which would require a MediaWiki version of some scripting system to design the logic for your particular actor or process) and then slowly integrate geologic processes (Imagine a world where every child has access to an animated explanation of the formation of the Andes), economic processes, migration, political change, you name it. Since I already described Web 3.0 as the Hybrid World, I’m going to call the commons-based peer collaborative agent-based modeling Web 4.0, or maybe Web4 or Web4D or Axxiles. No, if it has to be a Greek, it should be Antikythera. The hardest part is getting the name right, after that everything else should simply fall into place.
Just to update, because you should never write anything about spatial matters without first looking at Very Spatial and Vector One, take a look at this procedurally-generated city and tell me we can’t build a beautiful model of world systems and processes with a low barrier to entry.