Notes on computer maps for state murder (well, broadly speaking) and geospatial software history
A question that one of my non-geographer committee members immediately had upon reviewing my dissertation proposal was “where is the Cold War in all this?” He’s a computer historian, and history of computing has a very well-defined military-industrial complex lineage. Maps and geospatial analysis have a military lineage too (and an even longer lineage than computers) so it would stand to reason that the development of GIS and geospatial software would also have defense origins.
The computerization of American warfare (sort of disastrously applied in Vietnam but also championed by Reagan with his “Star Wars” program) shaped the context of critical geographer critiques of GIS’ expansion and normalization within geography. One of the better-known versions of this critique appears in Neil Smith’s “History and philosophy of geography: real wars, theory wars” from 1992. The essay opens with a recap of a trade magazine article about the use of GIS in the Gulf War as a way to illustrate the “deliriously detached” way that academic GIS evangelists tended to overlook the technology’s military ties and applications. Smith doesn’t name the specific companies who contributed GIS tech to Gulf War operations (his citation for both, an article in an out of print and seemingly never digitized trade publication, has seemingly been lost to the vagaries of archiving) but he does mention a specific product:
…a California company supplied the ’ruggedized DOS-based’ software called Map, Operator, Maintenance Station (MOMS); the pilots slotted the resulting portable GIS (Geographic Information Systems) packages into cockpit computers, and the ’turkey-shoot’, as a USA general called it, was on.
It’s not that hard to figure out that the company behind MOMS (incredibly dumb acronym, incidentally) was San Diego-based Horizons Technology Inc. Horizons Technology employee Rick Pope presented a paper about MOMS at a 1989 conference which comes up in academic database searches. The presentation explains that MOMS was developed on contract to the US Marine Corps for use in the AV-8B Harrier aircraft. But Horizons Technology doesn’t have a huge historical footprint. A 1991 LA Times article states they were founded in 1977 and primarily worked as a defense contractor—more than 90% of their business was in defense contracting. MOMS wasn’t the only geospatial tech Horizons built—during the Gulf War, they also worked on software for a “hand-held flight management computer used in Chinook and Blackhawk helicopters that, among other things, “reconciled the often conflicting coordinates of the various mapping systems used in the region during the war”—but it doesn’t seem like this became their niche specialty and certainly this work didn’t get civilianized. Given this was written in the still early years of GIS I understand the impulse to categorize the tech as such, but the difference between “military contractor did two GIS softwares and then sort of didn’t do anything else” and “GIS software company did a military contract” seems worth distinguishing.
While I think that it was fair for geographers in the 1990s to be wary of uncritical approaches to GIS that didn’t acknowledge how it could also be a force multiplier for military atrocities, Smith’s assertion in 1992 that “the text and context of GIS is heavily underwritten by a military agenda” suggests a military origin story that the record doesn’t exactly bear out. Defense contractors made geospatial softwares for the military, yes, but those companies didn’t tend to civilianize those products. As best I can tell the only in-house military GIS software that’s been properly civilianized, GRASS, came out of the Army Corps of Engineers which, while no less fraught than any other military division, isn’t really the “we need maps for targeted murder” side of the military. There were also some in-house geospatial software projects that continue to exist but are probably never going to be productized and made available to the public, like the FalconView software product made for preflight mission planning for F-16s (actually a pretty similar use case to MOMS).
Meanwhile, geospatial software products that start in and serve civilian sectors tended to eventually seek out defense contracts after establishing themselves, because there’s a lot of money in defense contracting and not a lot of money in basically every other sector that uses geospatial data (except maybe oil and gas, which is secretive enough that it seems like a lot of their stuff gets built in-house anyway).
In Esri’s case, their big break as the main contractor on a defense project was in part an attempt to remedy the problem of defense agencies getting contractors to build all sorts of niche specific GIS products for their particular use-cases, establishing new and not remotely interoperable softwares and file formats in the process. The Digital Chart of the World (DCW) contract with the Defense Mapping Agency (DMA), while ostensibly about creating a 1:1,000,000 scale global base map using digitized DMA maps, was really more concerned with implementing a file format standard for vector geospatial data in defense agencies. The file format (VPF) didn’t really take off outside of uh, NATO country militaries and I imagine at this point has largely been superseded by shapefiles. (There is also a story about DCW data being used in the Gulf War, but the specifics of it are only vaguely mentioned in a final report about the project.)
Now, this isn’t to say that there were zero defense interests in GIS or that those interests were for totally benevolent and chill purposes. Poiker’s work on Triangulated Irregular Networks (TINs) was supported by the Office of Naval Research, and the old NCGIA Initiative webpage on TIN history speculates that one of the grant’s areas of focus was “almost certainly a slightly disguised version of the cruise missile guidance problem” (though, annoyingly, it doesn’t seem like anyone followed up to figure out if this was the case?). The Harvard Lab got ONR contracts in its post-Fisher years, and SYMAP was used for making visualizations of hamlet sentiment in Vietnam.
What exactly are we supposed to do with the knowledge of a now-civilian technology’s military ties? I think it can tell you some things about design decisions and maybe help you understand affordances and limitations, but I am wary of over-moralizing. I don’t love the “military origins of a civilianized technology makes it by default suspect/contaminates it” school of technology criticism; there’s something very Catholic and bound-by-original-sin about it and I get enough of that from my mom’s side of the family. Then again, people sort of tell on themselves when they go out of their way to conceal military ties of geospatial software—I don’t know that SYMAP’s use in the Vietnam War is “secret” so much as “wasn’t archived”, but it’s interesting that history was uncovered via the National Archives and isn’t really part of the “it all started here at Harvard” hagiography. In a more recent example, I know someone who worked on a software project for the National Geospatial Intelligence Agency that has since been open-sourced. On the one hand, it’s good open source software that serves a purpose, and turning military money into something open source is a kind of hack that has served other greater goods. On the other: it’s interesting that the repo makes no mention of that contract work or why NGA wanted the tool in the first place.
Basically in the case of civilian geospatial companies that break into/also do military work, the critique mostly lies in calling out a hypocrisy: a company that markets itself on the benevolent possibilities of geospatial going after military contracts is seen as a contradiction of those values. The cynical read on this is to cluck one’s tongue at the naivete of believing that any company has “values” beyond profit and shareholder value—no ethical consumption, blah blah blah. To me the big stakes and opportunity here are from a labor perspective. If someone who wants to do maps stuff gets into it because they want to help climate scientists or urban planners, it sucks to find out their work is helping the CIA instead. Especially if that worker drew some personal line about not working for defense contractors. I don’t know how much to trust this Reddit anecdote about efforts to suppress unrest within Esri around the company’s work with the IDF, but given the stories about basically every big tech company’s involvement in and seeming apathy to the Gaza genocide it sounds plausible.
The TLDR on all this, I guess, is: geospatial software definitely benefits the military, and on an infrastructure level (e.g., GPS) the military has benefitted geospatial software. But the development and proliferation of geospatial software can’t be tidily reduced to a military agenda any more than it can be reduced to a landscape architecture or urban planning or land use agenda.