Last week, I read and thoroughly enjoyed the massive multi-author article, “A guide to good practice in Mediterranean surface survey projects” in the most recent issue of the Journal of Greek Archaeology.
Not only does it represent the collaboration of a number of key figures in Mediterranean survey, but it also reflects their vast range of experiences in different situations, regions, and landscapes. It’s aim is to produce a sweeping guide to “good practice(s)” in the Mediterranean which both recognizes the limits to best practice rhetoric in archaeology and offers a series of well-considered recommendations for archaeologists starting or engaged in survey projects in the Mediterranean.
In the spirit of good practices, I thought that I would think with the article as part of an shared conversation as opposed to respond to it specifically. Here are five things I think that I thought after reading this article:
1. Institutions, Questions, Methods, Scale
One of the remarkable things about this article is the wide range of experiences among the authors. Their experiences in the Aegean, Near East, Italy, and elsewhere ensures that they recognize the wide range of institutions and traditions present across the Mediterranean. In fact, the Mediterranean is a famously fractured landscape defined by different topographies, different national archaeological laws, different institutions, and different professional discourses. Surveys that cross thousands of miles in the Near East are simply not possible in the Aegean for not only reasons of archaeological policy, but also owing to the landscape and the kinds of research questions Aegean archaeologists tend to ask. These questions, in turn, reflect the academic and institutional history of archaeological work in these places.
Issues of scale and intensity likewise reflect the kinds of questions that survey projects seek to answer as well as the kind of permits issued and the access to resources: funding, labor, infrastructure, et c. Intensive pedestrian surveys have invariably developed in different ways in different contexts. As a result the authors are wise thinking in terms of “good practices” rather than “best practices.”
2. Landscapes and Sites.
The article also got me thinking a bit about the tension between the concepts of landscape and sites. The latter typically reflects an interest in settlement patterns, functions, and hierarchies that demand careful highly detailed documentation of concentrations of artifacts in the field. The classic technique used for surveys with an interest in sites is gridded collection (and corresponding “off-site” scatters). And while we’ve moved away from “dots on the map,” it seems like many surveys still think in terms of sites and want proceed with the assumption that surface scatters represent archaeologically delimited places of human activity.
At the same time, there are surveys with an interest in landscapes which tend to be less distracted by sites and more interested in the distribution of artifacts throughout the survey area. Distributional survey often seeks to answer questions that require an a large regional assemblage of material to consider interregional connectivity, diachronic changes in material culture, and presence/absence of certain periods across large areas. As a rule, these projects give less weight to any particular cluster of material on the surface and use intensive sampling of the environment as a way to identify larger spatial or regional trends.
Projects that focus on landscapes also tend to privilege the integrated character of survey areas and often seek to weave together geological, environmental, ethnographic, and historical data to reveal the diachronic complexity of lived space. While projects with an interest in sites have started to do this as well, of course, but there remains a tension between surveys focused on sites and those focused on landscapes. This tension is most visible as surveys with an interest in landscape, at their most ambitious, look to phenomenology as a way to think about the survey universe as lived space in the past and in the present.
The influence of phenomenology in the Mediterranean survey is pretty small and most Mediterranean surveys tend to be fundamentally processual in their methods and analysis. Even then, a landscape approach reflects a different view of the survey universe. My thinking on this matter has recently benefited from some conversations with my colleague David Pettegrew who introduced me to the idea that siteless surveys are susceptible to “deconstruction.” The refusal to distinguish between on-site and off-site scatters has produced assemblages of artifacts that are susceptible to being pulled apart in various ways to reveal landscapes associated with different historical periods, different methods, and different processes.
3. Time and Chronology
Good practice also explores the tension between time and chronology in intensive survey analysis. Unlike stratigraphic excavation which can create its own context for relative dating, survey relies on existing chronologies and ranges of absolute dates for the analysis of surface assemblages. In most cases, however, the ranges of dates are still quite broad meaning that while artifacts might be contemporary in the technical sense of sharing a period, they are not simultaneous in the sense of existing at the same time in the survey area or sharing depositional or functional worlds. In other words, artifacts of Classical date on the Isthmus did not all enter the landscape at the same time or relate to one another, but may well represent trends that existed over nearly two centuries of activities.
This lack of precision in dating in survey pushes us to constantly assess the character of the sample collected from the surface or sites especially since the standard chronologies used in excavation and surveys tend to be irregular in duration. For example, the Classical period might represented fewer than 200 years while the Late Roman period might represent twice that and the Final Neolithic over one thousand of years.
Time and chronology get all the more complicated when considering that an object like a fortification wall or terrace might be visible and functional for a thousand years or more whereas this is less likely for an amphora sherd. At the same time, it is not uncommon for a 2000 year old roof tile to be used as chinking in a 19th century church or as part of an early 20th century oven.
In short, we’ve only just started to think about how survey requires a fundamentally different view of time than we might rely upon to make sense of an excavation.
I still get excited by formation processes and intensive survey. To my mind, this is the place where questions of time and scale intersect. It seems to me that the main difference between site-based survey and siteless survey rests in our confidence in our understanding of the formation processes that shape surface assemblages. A confident approach argues that sites reflect past human activities which took place in a particular space, at a particular time, and for a particular purpose or function. A less confident approach argues that while we can’t necessarily isolate the processes that create a surface assemblage in any one place, we know that the artifacts that are visible in our survey universe represent past human activities broadly. In other words, the range of human and natural formation processes will tend to obfuscate any patterns that consistently reflect the distribution of functional spaces where activities associated with a particular moment or period in the past took place. Instead, we get a muddled and residual record of past activities that took place in the survey universe that will inconsistently resist our efforts to isolate particular depositional processes. This inconsistency means that we either need to only work at the smallest possible scale, which like excavation, can hope to isolate formation processes that produced a particular surface strata or at a scale sufficiently large to make the need to understand the variables at any one site inconsequential to diachronic and regional level research questions.
This lack of confidence characteristic of siteless survey may appear to compromise the interpretative potential of intensive survey which continues to lean heavily on research questions derived from excavation and historical narratives that focus on sites, individuals, and events. At the same time, we’d argue that siteless survey pushes us to read the landscape in ways that make questions about whether artifacts entered a particular surface assemblage (or survey unit) as a result of manuring or low-intensity activities less important than what these artifacts tell us about past activities in the survey area a whole. This also means that we don’t have to get bound up in the difference between a single episode of high intensity activity (e.g. a high-density military camp occupied for a few weeks) or a long period of low intensity activity (e.g. a season settlement) that might produce similar scatters.
As this article show, understanding the limits of what we can say about formation processes and the surface assemblages remains central to understanding what arguments we can make at a regional scale.
Finally, I’m getting more and more interested in field practices. That is procedure at the most basic level. I’ve been mulling over publishing a series of intensive survey manuals.
For example, last month, the Western Argolid Regional Survey (WARP) published its survey manual. I have a copy of the EKAS survey manual as well that I’m kind of itching to publish. A few years ago, I pulled together a little gaggle of archaeological manuals that were available online and put them up here.
I’d love to add some more to that list or talk with anyone interested in publishing their intensive survey manual perhaps in a volume with several manuals presented side-by-side.