I was pretty excited to read S.T. Stewart, P.M.N. Hitchings, P. Bikoulis and E.B. Banning, “Novel survey methods shed light on prehistoric exploration in Cyprus,” in Antiquity 91 (2017) over lunch yesterday. First off, it had the words method, survey, and Cyprus in its title, which always hit me in the “feels.” Secondly, it deals with survey efficiency across complex landscapes on the island, and this reflects a challenge that we’ve faced on the Western Argolid Regional Project (WARP) over the past three field seasons.
Like any good article, it sent me through an emotional rollercoaster.
At first blush, I was horrified at the prospects of using predictive modeling to exclude certain units within a survey zone from intensive pedestrian survey in the name of efficiency. This felt like exactly the kind of technology-driven Taylorism that my recent scholarship has railed against. A slow archaeology embraces the kind of contingency, chance, and randomness that defies (at best) and taunts (at worst) efforts at efficiency. At its most absurd, this is discovering the most significant and time-consuming finds on the final scheduled day of field work.
The authors describe how they excluded units from survey because more recent geomorphological processes would obscure or destroy the late Pleistocene and early Holocene landscapes that would contain artifacts from the earliest periods of occupation on the island. In practice, survey archaeologists know that nothing is ever that simple. Modern, mechanized activity in the landscape is capable of removing meters of sediment to expose earlier paleosols, reworking the water flow across the landscape to erode through more recent deposits and revealing scarps and sites long buried, and even moving soils from elsewhere are depositing them and their accompanying artifacts in unexpected places. While all these contingencies require critical consideration when used to construct settlement patterns across the landscape, they can also provide unexpected windows into the past.
Finally, intensive survey is about more than just looking for artifacts. By ignoring units that are unlikely to produce artifacts, they deprive themselves the opportunity to engage the landscape in a wholistic way. While it is fair enough observation that contemporary landscapes are different from earlier landforms, engaging the landscape compels the survey archaeology to recognize the diachronicity of all survey assemblages. An exposed late Pleistocene horizon is no less a component of the modern landscape than the earliest levels of the so-called Anthropocene.
In short, the urge to efficiency in their survey methods felt like a lost opportunity (at best) that risked insulating the archaeologist from the full context of even the earliest artifacts within a dynamic modern space. This modern space is where we as archaeologists encounter the landscape and produce our understanding the fine strands that connect our world to the ancient.
Then, I took a deep breath. What Stewart and her colleagues proposed is actually pretty cool. They created two models. One was a general model of landscapes in the Tremithos River Valley and the other was a more specific model based on their daily work in the field. This latter model was particularly interesting because it was iterative. Each day this specific model was updated with data from the field revealing the potential and power of a sophisticated GIS and data-management system.
More than that, my colleagues and I have argued in print that intensity matters in producing analytically meaningful survey assemblages. A system that takes into consideration data collected on the fly and allows the archaeologists to know where added intensity is likely to produce the most meaningful results – and if this system bore fruit – is exactly the kind of targeted and variable intensification that my colleagues and I have recommended in survey practice. So whatever efficiency is gained by using models, for me the gain is really in intensification.
Finally, sometime about 4:30 pm yesterday while I was on the second mile of my run, I realized that Stewart and her colleagues were probably not wrong in their approach. Having spent the last three seasons trudging through cobble-strewn fields along the banks of the Inachos river and finding nearly nothing (and learning as we went that some of these units did not preserve much of the ancient surface), I am acutely aware of the needs to treat the landscape with systematic efficiency. From sampling and collection strategies to field tactics, intensive pedestrian archaeology is inseparable from modern, industrial practices that extended from auto manufacturing the organization of universities. If industrial production can be designed around predictive models and machines that learn, then intensive survey will invariably absorb these same impulses and trend toward increased efficiency in the kind of archaeological knowledge that it produces. In fact, check out the first 100 or so pages of Pyla-Koutsopetria I: Archaeological Survey of an Ancient Coastal Town (2014); we lay out explicitly how our sampling strategies balanced intensification and efficiency.
In a very practical sense, much of our conversation over the past three on WARP is how to approach our landscape efficiently. We had the luxury of increasing the number of field teams every year and retaining our amazing group of team leader who develop more efficient field practices each season. As a result, we surveyed larger and larger amounts of territory each day and each week. This allowed us take risks and survey areas where we though it unlikely that we’d find significant artifacts scatters. This included steep, densely-vegetated, hill slopes as well as the valley bottom near the river where erosion and sedimentation conspired to obscure ancient surfaces.
Knowing what we know now about the geomorphology of the Inachos valley and the artifactual landscape (that is in hindsight), we probably would have deployed our survey teams differently. At the same time, walking the valley bottoms did prompt us to think more carefully about both modern land use and fragmentation as well as routes both along and across the Inachos river. These were important considerations as our survey was diachronic and all parts of the landscape could contribute to our larger arguments.
By the time, I was done processing this short article, I had come full circle. It’s a fine article and characteristic of the discourse in intensive pedestrian survey and reflective of both practical challenges and opportunities facing field work in the digital age.