Tag Archives: GIS

Archaeological futures III

[This post is a continuation of two other blog posts: part one and part two.]

I think that a basic structure of the talk I’m giving at Smith College is coming into shape. I’ll start with a brief description of the accelerating sophistication of archaeological methods since I began as a student in the mid-1990s, focusing first on the proliferation of archaeological sciences and their integration into Mediterranean archaeology and the proliferation of data produced by archaeological projects. (I’ll hopefully use data from Corinth Excavations to get a quantitative sense of the increase in data produced).

This leads to the issue of data and digitization. One of my big take-aways from the Mobilizing the Past for a Digital Future book was a more clear sense of how this proliferation of data has encouraged archaeologists to go digital, for a couple of different reasons, two of which I’ll highlight here: (1) The quantities of data are so immense that digitizing them (or better, creating them digitally to begin with) is an elegant solution to basic storage and dissemination needs, and (2) the types of data are so divergent, largely because of the increase in specialist and scientific studies in archaeological projects, that integrating these data are a significant hurdle, and digital integration is again a good solution.

A couple of personal anecdotes come to mind here. The archaeological survey that I was trained on, EKAS, started fieldwork with a “fully functional GIS” – that was exciting in 1999. When I was planning a survey in 2011 (WARP), one of the first things that I did was to think about the GIS and how we could integrate all of our data within it. Although I haven’t exactly been successful at pulling in non-archaeological data to the GIS framework, it still remains the basis for much of our analysis. When I write papers about WARP – as I was this week for a paper I’m giving tonight – I usually write on my desktop and keep my laptop reserved for ArcGIS (I run GIS on my laptop because it’s that indispensable to me). When I started excavating at Corinth in 2004, on the other hand, Corinth Excavations still used hard-bound notebooks, although they had become increasingly form-driven (we used a stamp). To figure out what had been excavated in my area prior to 2004, I had to flip through two or three different hard-bound notebooks. They proved to be not very well-written or illustrated, and I found myself increasingly frustrated by the fact that I was “digging blind.” (Actually that is why I was put there, to figure out what was going on and to “clean up” the interpretive knots created by past supervisors who hadn’t done such a good job).

One of the best arguments for digitization and born-digital data is (in my opinion) to give supervisors more interpretive tools in the field, by allowing them to call up previous years’ work easily on their computers or tablets in the field (additionally reference works and scholarly literature in the field can be extremely useful). This is a way to bring everything that we can to bear on solving interpretive problems in the field. I stress this point because if archaeology is anything, it is that: thinking through problems in the field, at the trowel’s edge (to use the excavation metaphor). Excavators are not just collecting data – they are interpreting as they go, and their interpretations shape their data. To do better archaeology, we need better interpretations more than anything else, and digital tools have allowed us to stretch our capabilities to pull up information of various kinds, especially outside of the library.

In that respect, digital technologies have the ability to help us bridge the field/library divide that is such an important structuring device of archaeological discourse, so much so that it even made it into an Indiana Jones movie (note: Vere Gordon Childe didn’t spend most of his time in the field). On the other hand, there can be a tendency, at least in some writing on “digital archaeology,” to emphasize the importance of data collection in a way that seems to separate it from interpretation. The word “data” appears 1619 times in the Mobilizing the Past book, “interpret*” only 164 times, and the general tone of much of the discussion focuses on interpreting data that have already been collected, thus reifying the field/library divide as it is being transformed into a field/computer lab divide. The section entitled “Interpretation” in the introduction to the Mobilizing the Past book is less than two pages long and focuses on a discussion of Bill Caraher’s “slow archaeology” work – in the same chapter, we get more words devoted to Apple’s famous “1984” commercial. Likewise Roosevelt and his colleagues briefly discuss the importance of interpretation trench-side in their introduction, but at no point is their born-digital system described as improving interpretation, nor do they talk about interpretation very much at all. In fact, I only see one contribution to interpretation in the entire article: an assertion that technologies that model space in three dimensions would aid in interpretation. This is how they describe their system (emphasis mine):

Conceived and designed before excavation commenced, the system was ‘‘born digital’’ (Austin 2014: 14), operating with integrated databases and aiming for the production of high-quality data with manifold improvements in accuracy and efficiency.

I don’t want to linger too long here, since I’ve already made my feelings clear. This is not a criticism of the tools or practices of digital archaeology, but rather of its discourses. But discourses are important: they reveal what we value and what we don’t.

What I want to say, then, is that we’re barely realizing the potential of our new digital tools, but that to make them work for us, we need to avoid this obsession with high-quality data, efficiency and accuracy. All those things are wonderful, but they are all a means to an end: productive interpretations and analyses of research questions. It’s true that high-quality data aren’t just produced for us but for future generations of archaeologists, but I think it’s a mistake to think that if we just produce the highest-quality data then one day someone might make interpretive hay of them. Our primary goal has to be productive interpretations in the here and now.

“Digital archaeology,” if that is even a thing, then, is at once an opportunity and a challenge. The opportunities are obvious, and they’ve encouraged even the most skeptical of us to become digital archaeologists in a very meaningful way. The opportunities are made very clear in the Mobilizing the Past book and in the article of Roosevelt et alii. The challenge is to unthink data and methodological sophistication as ends in themselves.

Let me end with another personal anecdote, which reveals that I am as guilty of this as anyone. When I started working on EKAS, I was extremely suspicious of defining “sites” in the field. I still am. What I wanted to do was to plug all of the artifactual evidence into the GIS and produce maps that would help us to understand artifactual distributions better and to define “sites” (if that’s we wanted to do) in the computer lab. I still think that’s a better way of doing things than declaring “ooh, there’s lots of stuff here” and defining a site on the fly. The on-the-fly way of doing things is data-poor and isn’t meant to require a lot of thought. You might feel rushed to make a decision as you pace back and forth trying to decide if it’s “really” a site while your undergraduate field walkers look at you impatiently, waiting to get on with it. The GIS-y way of doing it is data-rich and involves more thinking. In fact, that’s the value of the digital way of doing it: you can display and comprehend more evidence than is easily comprehensible when you’re in the field. So on the one hand, there is a separation made between fieldwork and analysis, which is something I’ve argued against. On the other, that separation gives you more information and more time to contemplate your decision-making. It gives you a little more information, and a bit of space to think.

There is no magic bullet

The most recent issue of the Bulletin of the Institute of Classical Studies has a great little cluster of three articles about digital approaches to the Mediterranean world that I read over the last couple of days with great interest. I was particularly interested in Elaine Sullivan’s advocacy for 4D GIS visualizations, which, she argues, “afford new potential for the examination of now-altered ancient landscapes from a human viewpoint, specifically for exploring aspects of experience that changed through time and space” (71).

I don’t know nearly enough about Egyptian archaeology to evaluate Sullivan’s paper as a contribution to that field; I’m more interested in her contention that certain technologies (specifically 4D GIS) are better positioned to contribute to (in her terms) humanistic and qualitative analyses of the ancient world than others (specifically traditional 2D GIS). Sullivan’s rhetoric is measured and her discussion is thoughtful. I do worry a little bit about the idea that certain technologies in and of themselves are good for one thing and not for another, at least as it is usually expressed, since it seems to me overly simplistic.

Let me explain. Sullivan begins her discussion with an exposition of the limitations of traditional, two-dimensional GIS. It’s a criticism that all archaeologists (I presume) know well, since it’s been out there since the late 1990s. As Sullivan explains,

With its powerful aggregation and layering tools, GIS offers numerous avenues to approach ancient landscapes quantitatively. However, the limitations of GIS systems have led to serious critiques that question its larger potential for archaeology. In the 1980s and 1990s, post-processual archaeologists rejected GIS as reductionist, positivistic, and lacking engagement with cultural and social factors. Landscape was theorized as more complex than an environmental stage onto which human actors were dropped – and the human-environmental relationship was redefined as dialectic. Archaeologists investigating how embodied humans would have experienced and interpreted specific cultural places still in many cases see GIS as antithetical to exploring this relationship, creating false objectivity in what were subjective spaces. Current theorists emphasize that landscape studies must include the ‘material, cognitive and symbolic’ aspects of this dialectic.

Indeed, traditional GIS lacks many features providing the type of contextual information vital to approaching humanistic research questions. The platform works primarily in a two-dimensional coordinate system, which lacks the qualitative aspects that reflect the inhabited human world. People do not engage with the world from an overhead, omniscient viewpoint, but from the perspective of a single viewer. Cartesian space does not replicate human sense of scale, physical relationships between people and things, or aspects of ‘local distinctiveness’ that create cultural meaning in specific places. Also, human movement through space and the changing perception of spaces through time cannot be duplicated in traditional GIS. As one leading scholar in the field succinctly stated: ‘GIS are currently ill-equipped to deal with space as it surrounds an individual.’

I’ve always found this criticism of GIS a little bit – okay, a lot – misplaced. Sure, GIS represents the world in particular ways that do not correspond to the ways that humans experience the world, but that’s hardly surprising. GIS doesn’t even represent the surface of the earth accurately, after all, since it’s constrained by the limitations of cartography. It’s essentially a mapping program (at least as it’s used by archaeologists), and I don’t think that anyone worries that maps are too positivistic; they’re ways of displaying simply a complex reality. Certainly there is a problem if one thinks that GIS represents reality unproblematically, but does any normal person actually think this? If we get a sense of false objectivity from GIS, it is because we are not doing enough thinking.

Moreover, if we’re going to wring our hands about GIS, then it seems to me that we need to wring our hands about everything else. We have to worry about maps of all kinds, representations of all kinds, rulers and compasses, survey tapes and GPS units. This criticism of GIS feels a little bit to me like the lack of recognition that there is a difference between a model of reality and reality.


Or, put more famously, “all models are wrong, but some are useful.”

So sure, it’s true that human beings do not live in a digital panopticon, but none of the tools at our disposal for the study of the ancient world really capture the lived lives of ancient people. Maybe 1% of all ancients could read and write but we don’t seem to be too concerned that our primary mode of communicating is the academic text.

I never really understood why GIS was the target of this post-processual assault (I say this as someone who uses GIS quite a bit and whose theoretical framework is essentially post-processual). The only thing that I can think of is that the problem is the faith that many practitioners have in their tools. If you think that there is a technological magic bullet, then any technology that falls short is a false idol that must be denounced.

That seems like exactly the wrong approach to me. This isn’t me picking on Sullivan; she’s really using this pre-existing critical discourse about traditional GIS to pivot towards her discussion of 4D GIS. But her treatment of 4D GIS is balanced; she doesn’t claim that it will solve all of our problems, merely that it is a useful tool. (I have more to say about that, but I think in another blog post).

Ultimately it’s our faith in various technologies that is at fault here more than anything else. We need to remember that no technology will (on its own) allow us to really understand the lived experiences of anyone. If we aren’t mindful of this fact, we’ll just end up jumping from bandwagon technology to bandwagon technology.


Kriging the artifact densities from the Western Argolid Regional Project.