Site Home > Multiple fields & archaeological practice > Contemporary survey practice
Key Pages
Christopher Witmore |In 1953, at the suggestion of Carl Blegen, William A. McDonald began exploring the region around Bronze Age Pylos on the southwestern peninsula of the Peloponnesus, Greece. The original focus of this surface exploration was on the topographical remains of Bronze Age settlements associated with the palace at Pylos. From this inception, however, UMME would undergo a radical transformation over the course of two decades. This transformation amounted to ‘a major reorientation and expansion in goals and methods’ that for M. Fotiades would lead ‘Greek archaeology into its modern phase’ (Fotiades 1995, 59). Indeed, by the late 1960’s the project had taken on an interdisciplinary scope through the incorporation of a diverse body of researchers from fields such as agricultural economics, anthropology, cartography, civil engineering, geography, geology, metallurgy, and soil chemistry (McDonald 1972, 5). Through this interdisciplinary collaboration UMME began generating and attempting to address new questions on a regional scale. Furthermore, there was an explicit and repeated emphasis on the use of scientific method—amounting to an empiricist/objectivist approach—in addressing particular questions. The regional methods enlisted by UMME involved a combination of extensive surface search, aerial photography (military derived), mapping, and test sondages at ‘selected habitation sites.’ Because UMME was regarded by many as the landmark project that spawned the ‘New Wave’ in landscape survey it is important to accentuate its connections to wider disciplinary developments that were occurring at the time.
McDonald was acutely aware of the broader trends toward regional survey in archaeology taking place outside of Greece. In the final publication he had asserted explicit links to the some of the key projects behind the development of the New Archaeology. Beyond the three interdisciplinary surveys he dealt with in some detail—Robert Braidwood’s Prehistoric Project, Robert M. Adam’s Diyala Basin Archaeological Project, and William T. Sander’s Teotihuacán Valley Project—McDonald made special mention of, among others, Kent Flannery’s Oaxaca Valley Project, Gordon Willey’s Virú Valley Project, and Richard MacNeish’s Tehuacan Project (1972, 13-17). For McDonald, these survey projects epitomized regional scale interdisciplinary undertakings with their focus on collaborative research in addressing questions of paleoecology, settlement patterns, agricultural practice, and demography. Furthermore, they were also exemplary in their supposed adaptation of the epistemology and methodologies associated with the natural sciences.
McDonald’s willingness to forge intellectual links with major projects outside of Greece was counterbalanced with his acknowledgement of the older traditions within. In contrast to the rhetorical strategies utilized by proponents of the New Archaeology in the United States (Wylie 2002, 29), a decisive break between early topographical field exploration and large-scale regional surveys was never formulated as such in the Mediterranean. While working against the grain of a very conservative tradition in Classical Archaeology, the figures behind the new scientific innovations of regional survey, such as McDonald, also connected their regional approaches to a ‘the long-established tradition of archaeological field exploration in the Aegean area’ (1972, 13). What is more, McDonald explicitly associated UMME with what are widely considered to be canonical examples of regional ‘field’ practice in classical scholarship—the topographical tradition beginning with William Martin Leake and the large-scale French Expédition Scientifique de Morée (McDonald 1972, 10). In Chapter 2 I will discuss these links in more detail. However, my purpose here is not to address every aspect of this canonical intellectual history, for it has been dealt with to a satisfactory degree elsewhere (McDonald 1972; Schon 2002), rather I am interested in how the taken for granted dichotomous schemes of modernist thought come through in how survey practitioners describe what it is they actually do on the ground.
In the language used to depict the UMME mission and research agendas we encounter many of the fundamentally modern, dichotomous distributions of entities—nature | culture, data | interpretation, the field | the archaeological home bases, and so on. For example, McDonald portrays the mission of UMME as one of pulling together a diverse body of scientists who recognize ‘the importance of obtaining all relevant information about the natural environment as well as the cultural features of the target region’ (McDonald 1972, 16; my emphasis). More specifically, McDonald along with Richard Hope Simpson, describe part of the role of the research undertaken by UMME as ‘to coordinate and systematize all the available data and put forward some hypotheses based on all the relevant evidence we have been able to gather’ (1972, 121; my emphasis). For McDonald and Hope Simpson, ‘even when (or if) archaeologists are better trained in the fundamentals of science and technology,’ in attaining this research agenda archaeologists ‘will need the expert’s help in the field as well as in the laboratory’ (McDonald and Hope Simpson 1972, 122; my emphasis). While these dualistic schemes often come through in the language used by members of UMME they also serve as a conceptual base for the scientific methods they employ.
Frustratingly little is said in the UMME final publication of how they reached their conclusions from the ‘evidence’ (refer to Hodder 1999 for a similar critique of archaeology more generally). Beyond assertions of having formulated hypotheses on the basis of solid ‘facts’ much of the ‘raw data’ is presented so that it may speak for itself (more will be said of the rhetorical strategies of scientific ‘texts’ at a later point in the dissertation). This notion of explanation put forth by UMME asserts a distance between the data and the hypotheses that are derived to explain that data (Figure 1.1). This distance creates a distinction between practice and knowledge. While practice is what archaeologist do, whether ceramic sampling or mapping in the ‘field,’ knowledge is whatever information is mobilized in the course of this activity (Latour 1988). Such separations become important for those who wish to act at a distance and are therefore fundamental to the classic positivistic model of archaeological practice, which enabled practitioners to mimic the detached outlook of the natural sciences (Shanks and Tilley 1992, 29-45; also Wylie 2002, 61). Still, this is at best a partial contextualization because scientific forms of documentation incorporate a number of different modes for presenting an argument. Nevertheless, in laying out the classic critique I will blindly ignore, at least for now, the two-dimensional plans, maps, diagrams, illustrations, and photographs and will continue to focus solely on what these practitioners say. While UMME’s distinct scientific approaches and research agendas tied it into the wider transformations occurring across the discipline at that time, they also marked it as a watershed in the archaeology of Greece. To be sure, many other projects would soon follow its lead in the Mediterranean. Here I will focus on a few of these key survey projects arising in the wake of UMME in what has come to be known as the ‘New Wave’ of regional survey (Snodgrass 1990, 119).
The factors behind the emergence of the ‘New Wave’ of landscape survey in the Mediterranean and specifically Greece are complex and multiple (Alcock, Cherry, and Davis 1994, 137). On a general level the ‘New Wave’ is often connected with the rise of the New Archaeology and its emphasis on the meticulous and systematic study of regions (Cherry 1983; after Binford 1964, 426). But the rhetoric behind such broad disciplinary movements often obscures the finer details of such change. In other ways the ‘New Wave’ is also contextualized as a reaction to UMME (Cherry 1983; Schon 2002; and others). More specifically, the ‘New Wave’ is often connected with a number of specific factors by its practitioners: 1) dissatisfactions with traditional excavation centered practice in urban areas of Greece (Cherry 1983; Snodgrass 1987; and others); 2) more general differences between survey and excavation, such as relative costs (Cherry 1983, 383; Bintliff and Snodgrass 1985), a rhetoric of excavation as destructive in comparison to survey which is argued to be a low impact and repeatable form of practice (Cherry 1983, 383-384; Snodgrass 1987; etc.), and restrictions on excavation permits in Greece (Cherry 1983, 378); 3) crises brought on by rapid development, mechanization in agricultural practice, and more general environmental change (Bintliff and Snodgrass 1985, 123-125; Jameson 1976; McDonald 1972, 13; Snodgrass 1987, 93-131); 4) the different questions that might be addressed by survey ‘data’ and its strengths—multi-period, regional scope, inter-disciplinary aspects, and so on (Barker 1995, 1-2; Cherry 1983; Jameson, Runnels, and van Andel 1994, 11; and others). While these factors contributed the proliferation of regional survey approaches, much remained to be done regarding methodology.
With UMME systematic surface reconnaissance amounted to the use of military derived aerial photography as a guide to extensive coverage. Areas deemed fruitful for surface search were then visited on the ground. In this way, UMME was able to cover around 3800 sq. km of the Peloponnesus. However, the number of sites it produced per square kilometer (1 for every 13 sq. km) was extremely low by ‘New Wave’ standards (Alcock 1993). In contrast practitioners behind the ‘New Wave’ placed increased emphasis on intensive and controlled pedestrian survey. Pedestrian survey or ‘field walking’ is a technique of ground reconnaissance where individuals walk across landscapes in regular intervals collecting artifacts from the surface in order to establish patterns of occupation and land-use. Intensity often varies depending on the presence or absence of surface materials with distances between pedestrians ranging from 5 or 10 to 15 or 20 meters. With this technique much smaller areas of a region were covered (20 sq. km with the Keos Survey (Cherry, Davis, and Mantzourani 1991) and 44 sq. km with the Argolid Exploration Project (AEP) (Jameson, Runnels and van Andel 1994)), while much higher percentages of archaeological sites per square kilometer (7.5 per sq. km with the AEP) were located.
The proliferation of survey projects was perhaps most dramatic in the late 1970’s and early 1980’s. At a colloquium dedicated to the new profusion of archaeological surveys in the Mediterranean area held in Athens in June of 1981, over sixty different regional surveys were represented (Keller and Rupp 1983). Of these, perhaps the most conspicuous projects to contribute to ‘New Wave’ were the Melos Survey (Renfrew and Wagstaff 1982), the Cambridge and Bradford Boeotian Expedition (Bintliff and Snodgrass 1985), the Argolid Exploration Project (AEP) (Jameson, Runnels and van Andel 1994), the Keos Survey (Cherry, Davis, and Mantzourani 1991), and eventually the Pylos Regional Archaeological Project (PRAP) (Davis 1998). Together these projects built upon the pioneering efforts of UMME and maintained a common focus in addressing many of the key questions at the time (refer to Cherry 1983). While much of the literature on survey methodologies was dominated by discussions of site definition (Cherry 1983; Dunnel and Dancy 1983), sampling strategies (Cherry 1983), the nature of survey data (Cherry 1983), or the relative merits of survey verses excavation for addressing particular questions, the emergence of intensive regional survey work generated some critical and positive aspects such as experimentation and critical self-awareness regarding basic methodology that were seen to be lacking in contemporary excavation (Cherry 1983, 384). Still, Alcock, Cherry, and Davis argue that in choosing to promote survey there ‘is a tendency to down play the interpretive problems in handling survey data’ (1994, 137). More will be said of this subsequently.
The hypothetic-deductive reasoning in surface surveys (as with UMME) delegates ‘the field’ as the locus of ‘data’ collection. ‘Data’ form the basis for the generation of hypotheses regarding the nature of past human activity (Cherry, Davis, and Mantzourani 1991, 16). But to claim the division between data and interpretation falls neatly along the same lines as the division between the field and the laboratories, archives or studies would be a mistake. Cherry, Davis, and Mantzourani, for example, recognize that ‘archaeological data do not necessarily come in neat, self-evident packages called sites, so that to claim one has found a site is an act of interpretation whose criteria should be made explicit’ (1991, 21). Nevertheless, they continue by stating ‘how highly variable the surface archaeological record can be and how little of it we can currently interpret’ (Cherry, Davis, and Mantzourani 1991, 21). The authors regard the archaeological record ‘as a distribution of systematically related artifacts across the landscape, peaking in intensity at the particular locations usually designated as ‘sites’’ (Cherry, Davis, and Mantzourani 1991, 21). Here the archaeological record is not simply an ambiguous stand-in for reality (Patrik 1985), because the authors have a priori factored out the confusing and chaotic background noise of the Greek countryside. I will return to this point for it is the basis for an accusation of self-contradiction aptly levied at empiricist/objectivist reasoning by Ian Hodder (1997; 1999, 81) in the next section. For now, it is important to point out that while Cherry, Davis, and Mantzourani recognize that the definition of ‘site’ ‘is always a relative matter’ (1991, 28) they fail to regard the deeper notion of the archaeological record as such.
Ultimately the authors of the Keos volume are weary of what they regard as ‘subjective decisions’ in the field. Such occurrences should be checked retrospectively through subsequent comparison in order to avoid anything-goes-on-the-ground relativism. Cherry, Davis, and Mantzourani provide three ways by which this can be accomplished: ‘(1) by the comparison of off- and on-site densities; (2) by examining the tract-by-tract data for density peaks not defined at the time as sites; and (3) by later revisits to every site, armed with information about the density, spatial extent, and chronological range of the material collected from it’ (1991, 28). They want to maintain a clear separation between subjectivity and objectivity in their practice. It is such a firm maintenance of this separation that will provide grist for the mill of the interpretive turn. But this is more than simply a conceptual problem in ‘the field,’ because it is also embodied in the format of the final publications.
The very structure of the publication volumes produced by many of these projects replicates the divisions between data and interpretation or the field and the supposed contexts of interpretation in their layout (Hodder 1999; Lucas 2001b for similar critique). In this way, emphasis is placed upon internalizing the ‘data’ within the final publication in order to facilitate comparability. The highly praised UMME volume (Boardman 1974) took the lead with the presentation of ‘raw data’ in the form of 2 ‘Registers’ and several pocket maps. The authors maintained these as firmly distinct from the attempts ‘to understand and explain’ data in separate chapters (McDonald and Hope Simpson 1972, 129). With the Keos volume the survey data are arranged in their own catalogue—’The Gazetteer of Archaeological Sites.’ In the gazetteer the data are presented as a numerical list of both ‘site’ and ‘off-site’ artifact distributions. The textual component of these ‘full accounts’ of the artifact distributions contain site location (given as a grid-coordinate on the British Army map 1:50,000 Aegean series) size, and description (‘current land use, overall visibility, architectural features, and the location of artifacts of various dates’) chronological date, ceramic and lithic summaries, as well as find descriptions (catalog number, typology, corresponding illustration (intermittently presented), ‘measurements in m, physical description, date, and references’) and figure numbers for maps, plans, and photographs deemed significant for publication in a separate section, which follows this list (Cherry, Davis, and Mantzourani 1991, 69-70).
Indeed the separation is replicated with the AEP volumes. In A Greek Countryside, for example, interpretation is confined to a chronologically organized narrative concerned with long-term change from the prehistoric era to the 19th century. The survey ‘data’ are presented in an appendix as a ‘Register of Sites.’ This format differs from that of the Keos volume because plans and photographs are occasionally inserted directly into the textual descriptions organized as a numerical list of sites. Moreover, with the subsequent additions of Artifact and assemblage: The finds from a regional survey of the Southern Argolid, Greece and Contingent countryside: Settlement, economy and land use in the Southern Argolid since 1700, the multi-volume format of the AEP includes separate volumes for presentation of the ‘data’ and their ‘interpretation.’ Therefore, more expansive pottery and material assemblages are catalogued within their own volumes (Runnels, Pullen, and Langdon 1995). Beyond the issue of a separation between data and interpretation these publications give priority to textual descriptions. Detailed illustrations, plans, maps and photographs are not published with consistency. Some sites are simply presented in textual form while others may have an accompanying map, illustration or photograph. However, when two-dimensional plans and photographs are presented they seem to be treated as transparent windows to the objects of study because they are regarded as ‘raw data’ (Shanks 1997). Moreover, this treatment often tends to close the very materiality of those objects off from further reiterative practice (Webmoor 2005). (The issue of iteration and reiteration will be discussed in detail in Chapters 3, 4 and 5.)
Other critiques can and have been levied against the approaches utilized by such projects—top-down hierarchical issues and the effacement of input by skilled labor as well as the demarcation and depersonalization their activities as archaeologists in contemporary Greece (Fotiades 1993). At the most general level the epistemology brought to bear in systematic surface survey resembled that of the accepted mode of field practice, excavation. In this way, many of the most fundamental, taken for granted aspects of how these practitioners simplified their practice were left unquestioned. Therefore, the paths of knowledge production used in regional survey fell into the familiar conceptual ruts blazed by excavation. A thoroughly modernist ‘New Wave,’ of survey archaeology remained unaware of the more embedded and pervasive aspects of dualist (often regarded as Cartesian) thought which were beginning to be interrogated around them (refer here to Hodder 1999, 22-29).
Given the pervasiveness of modernist thought in archaeology (Thomas 2004b) it is hardly surprising that such divisions of entities are to be found in the very language used by the practitioners of regional survey in Greece. So with the ‘New Wave’ of surface survey one finds the familiar dualities in their empiricist/objectivist language—data and scientific explanation or historical interpretation, the field and the supposed contexts of knowledge production (laboratories, archives, studies), and so on. If we were to focus solely on the language of survey epistemology, as I have intentionally done thus far, then it might appear that survey methodologies employ a linear process, similar to excavation, involving the collection of artifacts from field-walking, materials-processing, analysis, and synthesis. In this uni-directional process analogous to the Henry Ford model of factory production (Shanks and Tilley 1992) interpretation takes place after the survey, both conceptually and temporally.
The pervasive dualisms and seeming linearity of scientific epistemology have fallen under increasing criticism in recent years and this is certainly warranted and necessary in the tuning processes behind the development of disciplinary methodologies. Indeed, much of the impetus for these critiques has come from within the realm of scientific methodologies. However, the contextualization of the empiricist/objectivist approach associated with the ‘New Wave’ presented here should not be taken as applying to the whole of scientific field practice in archaeology. Admittedly, the approaches I have outlined are a very traditional and yet they serve as basis for the next turn. Many of the critics of these ‘scientific’ methodologies by utilizing the highly sophisticated and reflexive art of interpretation (the core practice of the humanities (Gumbrecht 2004, 52)), have made significant headway in accentuating and interrogating how these pervasive schemes stand in for the separation of the material world and the world of ideas. In the next section I wish to contextualize postprocessual challenge to the scientific modes of archaeological practice—what I describe as the interpretive turn in fieldwork.
Return to: The notion of fieldwork in contemporary archaeological thinking
Forward to: Theory into practice—the interpretive turn in fieldwork
Multiple fields and archaeological practice Home