Nearly 25 years ago, Anthony Snodgrass (then my colleague at the University of Cambridge) and I co-authored a short paper for the Cambridge review, on much the same subject as the one addressed in this panel discussion (Snodgrass and Cherry 1988). It was born out of sheer frustration at trying to explain to our non-archaeological colleagues why it was that we were not excavating sites, like all the other archaeologists they had ever met or seen on television, but rather were conducting large-scale regional surveys in Greece – Snodgrass (with John Bintliff) in Boeotia, myself (with Jack Davis) in the Nemea Valley near Mycenae. Our paper's title, ‘On not digging up the past’, was of course a playful allusion to one of the most famous of archaeological books, Sir Leonard Woolley's Digging up the past (1931). My general views, and the character of my own fieldwork activities, have not changed much since then: hence my present title, which likewise plays on another very well-known work, Sir Mortimer Wheeler's Still digging (1955).
The organizers of this forum have raised two main problems: first, the exponentially growing number of excavations, many of them (allegedly) merely adding to the mountain of data on types of site, periods, or regions we already know well; and second, the glaringly obvious fact that many of them are not being published adequately, if at all. They suggest that stiffer obligations and sanctions for failures to publish in a timely manner are perhaps a blessing in disguise. And by asking, ‘how can we go on excavating new sites?’ (editorial, this issue, p. 1), they raise the worrying possibility of some voluntary or even mandatory moratorium on excavation, as well as, perhaps, a consequent downgrading of the defining role of excavation for the discipline of archaeology itself.
These are big issues, to be sure. Yet my first reaction is a sense of disciplinary amnesia, a forgetfulness that these questions have been asked many times before, perhaps especially since the 1960s and 1970s, the era that witnessed the dramatic growth of employment in rescue, contract and CRM archaeology in western Europe and North America, and the concomitant acceleration of archaeological data production. Almost three decades ago, for instance, the famous Cunliffe report on The publication of archaeological excavations (Cunliffe 1983) proposed, as a formal policy, much greater selectivity in both excavation and post-excavation activities; more particularly, in hopes of encouraging more rapid publication of archaeological work in Britain, it advocated a multi-level dissemination strategy whereby only the most synthetic and interpretive parts were to be formally and concisely ‘published’. The report did not have the desired effect. A comparable survey report by the Council for British Archaeology some 20 years later (Jones et al. 2001) likewise appears to have had little impact. Indeed, over the years there have been many conference symposia and round tables tackling much the same themes as the present panel discussion: to name just one instance, the session ‘To dig or not to dig?’ at the 1992 Archaeology in Britain conference. It is tempting to conclude that archaeology's publication problem is intractable, and that archaeology itself (in Björn Nilsson's words, p. 27) is ‘an unstoppable excavation machine’.
Here is a more specific example. As co-editor of the Journal of Mediterranean archaeology I helped organize a discussion forum in response to a paper submitted by two Israeli archaeologists, Raz Kletter and Alon De-Groot (2001). They were able to document that no less than 2,200 excavations had been carried out in the state of Israel during the decade 1989–98, and that the number of projects conducted annually was constantly rising and, in fact, had more than doubled during that decade (their figures, moreover, refer only to excavations, not surveys). Yet final publication – however that term is to be understood – lagged far behind, running at about the 20 per cent level. This, incidentally, is hardly aberrant. Renfrew and Bahn (2008, 573) cite figures showing that only 27 per cent of the digs funded by the National Science Foundation since 1950 have ever reached print, and as much as 60 per cent of modern excavations in Britain remain unpublished 10 years after completion; another study of biblical archaeology estimates that only 19 per cent of excavation projects in the Syro-Palestinian region in the 1980–89 period were fully published (Shanks 1996), and the ratio of published reports to excavation licenses in the Republic of Ireland from 1992 to 2004 appears to be far worse still (University College Dublin 2006). Kletter and De-Groot's proposal was a radical one: a temporary moratorium on new permits to excavate, whether on salvage or research projects, as the only means of restraining this burning desire to dig so much. Predictably, the several forum discussants did not take well to that idea, variously regarding it as an unfair collective punishment, as having no sound legal standing, or as being in any case unenforceable. Some saw the obvious solution lying with greater use of digital media and the Internet, whether for formal publication or for the archiving of data. But the intervening decade since this JMA forum was published has shown that ‘archaeology's dirty secret’ (Fagan 1995) actually remains stubbornly intransigent in the face of digital technologies (Watkinson 2008).
This forum's organizers refer to the publication crisis as ‘an additional complicating factor’ (editorial in this issue, p. 1) in answering the question ‘why excavate?’, but it is surely more than that. A site that has been dug in the public trust and (almost always, one way or another) at public expense, but that remains unpublished decades later, is really little different from one that has been looted: a part of the archaeological record has been destroyed for ever, with nothing to show for it, rendering the entire process both meaningless and ethically reprehensible. The shrivelling of state and local funding in the face of the current economic crisis is already forcing severe retrenchment in the furious pace of excavation, the closing of archaeological units, scaled-back university degree programmes and so on (Schlanger and Aitchinson 2010), but by the same token it will further limit both the means and the incentives for post-excavation analysis and publication.
Unfortunately, because the crisis is of global dimensions, it will impact archaeology not only in countries that already have an abundance of excavations and excavated data, but in poor ones too – ‘poor’ in the sense not only of economic underdevelopment and thus shortage of resources to devote to archaeology, but also in the sense of having an impoverished archaeological record as a result of insufficient fundamental field research. One of the obvious features of the statement for this forum, and of almost all those who participated in it, is a perspective largely rooted in experience in northern Europe, an area where archaeology has behaved like a glutton, unable to digest the huge amounts of data already produced, but always craving more. Archaeology in many other parts of the world, however, is a far leaner affair. Where resources are scarce and archaeological data still at a formative stage, excavation remains central and critical to knowledge production, and the question ‘why excavate?’ can have little relevance.
Here it is worth making the obvious comment that excavation is the only technique that can produce certain types of data that are crucial for understanding the human past. Only by digging, for example, can we recover C14 samples with sound provenances; reliable archaeobotanical, zoo-archaeological, and osteological materials; perishable finds such as textiles, leather, wood, ivory, or papyrus; the full range of artefactual materials in securely stratified contexts that allow them to be sequenced; detailed understanding of built structures; and so on. One is hard pressed to imagine the future development of any technologies that might dramatically reconfigure the central role of excavation in this regard, so it would be absurd to maintain that it has lost, or is ever likely to lose, its signal importance to the field. Rather, the key questions are whether excavation should continue on such a massive scale in areas where the widespread failure to publish undermines its very purpose, and how and where excavation should fit amongst the now far wider range of available techniques for acquiring data.
I turn to these questions below, but first I must register disagreement with the organizers’ suggestion that we already have an ample stock of certain types of site, rendering further digging pointless. We should not forget that any excavation – indeed, fieldwork of any sort – takes place at a unique point on time's arrow. There exist many examples that illustrate just how productive can be the re-examination of sites first investigated long ago, in another age altogether – Star Carr, Phylakopi, Çatal Höyük, and many, many others – but now approached with entirely novel and different technologies, fresh theoretical assumptions, revised research goals, and new data requirements. In fact, every site is unique. If the postprocessual turn has shown us anything, it is that the past is not ‘out there’, as some static repository of ‘primary data’, to be retrieved by neutral, scientific methods. It is co-created in the very act of investigation itself: excavation as reflexive field practice, or ‘interpretation at the trowel's edge’, to use a currently popular buzz-phrase (e.g. Hodder and Berggren 2003). What we see in the data, of course, depends on what we ask (a central tenet of problem-oriented ‘New Archaeology’), but consideration of archaeology within its social context introduces factors that make purely objective analysis an unattainable fiction. As Witmore (2007, 550) puts it, ‘multiple fields cover the components, contexts and connections implicated in real-time archaeological practice’. This makes me uneasy with the notion that we already have enough Bronze Age barrows (say), or too many Roman villas.
However excavation may be conceived, there is probably near-universal agreement among practitioners that it is integral to the archaeological process – but far less explicit acknowledgement that it is only one part of that process. As a survey archaeologist first and foremost, my own opinion has always been that the most effective archaeological research designs are those that are multi-stage and multi-scalar, with feedback loops. Charles Redman's (1973) well-known formulation envisaged a progression through a series of stages, roughly as follows (with some additions):
1. study of known sites and past research in a wide region, inspection of aerial and space imagery, interviews with local inhabitants about known sites;
2. extensive-mode survey reconnaissance of the region;
3. intensive survey of some carefully chosen sample within this region;
4. detailed studies of the surface and sub-surface expressions of a sample of the sites located, via geophysical survey, phosphate analyses, mapping of visible remains, etc.;
5. selective test excavation of a sample of these sites;
6. large-scale excavation on one or more key sites within the region.
Of course, in the real world such a logical sequence rarely plays out in so orderly a progression. Some stages may take place concurrently, or in reverse order; several may be omitted altogether, for logistical or financial reasons; and others may be entirely precluded by the archaeological laws that apply locally. Above all, explicit research design – which, after all, should guide practice even in the direst of salvage situations – still ‘requires the constant intervention of thinking human beings’ (Redman 1987, 249).
What this idealized sequence highlights, however, is that excavation is only one among a plethora of methods for archaeological data collection, methods which interpenetrate in complex ways. Yet excavation remains, stubbornly, the signature gesture of our discipline. Certainly, both in the popular imagination and as measured by actual employment, archaeology is virtually synonymous with excavation. One may be allowed to wonder, nonetheless, whether archaeologists have too willingly adopted a mindset where excavation is almost always given primacy, irrespective of circumstances. Two brief examples may serve to illustrate the point.
The Birecik Dam, on the Euphrates River near the Turkish–Syrian border, was completed and filled in 2000, as part of Turkey's Southeast Anatolia Project (Shoup 2006). Displacing thousands of local villagers, the reservoir flooded 100 square kilometres of arid land, rich with archaeological resources, along the Euphrates River valley, in an area almost completely blank on the archaeological map. The only regional-level response to this disastrous situation was a modest six-week survey, whose final report (Algaze, Breuninger and Knudstad 1994) briefly catalogued just 82 sites, within an area that assuredly contained many hundreds more that were flooded before study of any kind. But the chief outcry was about the partial flooding of Zeugma, a provincial Roman capital with remarkably well-preserved architecture, mosaic floors and wall-paintings. As the rising waters of the dam began to inundate the site, billionaire David Packard provided $5 million to assist in the recovery of the mosaics, affording lively media coverage of archaeologists racing against time to retrieve them, with the reservoir's rising waters literally lapping around their ankles. We can conduct a thought exercise here. What if those $5 million had been devoted to a thorough and systematic survey of the entire catchment of the reservoir, with selective test excavations on a range of representative sites? Would that not have been more archaeologically informative overall than the excavation of what, in the end, turned out to be less than 1 per cent of the flooded area of Zeugma (Kennedy 1998; Early et al. 2003)?
As a second example, take the Boeotian city of Thespiai in Greece, a large and important place, as is borne out not only by its continuous occupation for some 5,000 years, but by the very high density of surface material from the 5th and 4th centuries B.C. covering about 140 hectares. Because the site is essentially unexcavated and lies on land that is not today built up, it was possible to conduct an intensive survey of its entire extent, in the context of a comparable survey of its hinterland, that revealed in great detail its changing size and layout over five millennia, as well as allowing examination of those centre–periphery interrelationships which made the classical polis such a distinctive institution (Snodgrass and Bintliff 1988; 1991). This, however, is the telling statistic. It required about 300 person-days to cover the whole city. For the same expenditure of time and money, it was estimated (Snodgrass and Cherry 1988, 13), one could have conducted an excavation down to natural covering only about 75 square metres, or roughly 1/20,000th of the classical site! That choice was not a difficult one for the project's organizers. One can think of plenty of examples, furthermore, where resources expended on geophysical survey at large sites (e.g., in Turkey alone, the lower town at Troy, Kerkenes Dag, or Zincirli Höyük) have yielded very detailed plans of urban layouts that it would have taken many decades of excavation to match.
Excavation, in short, is a technique that does many things well, but it cannot do everything, above all where the spatial dimensions of the research problem (or the salvage contract) greatly exceed the scale of any standard dig. At Thespiai, some limited test excavation subsequent to the urban survey would certainly have been desirable, as a logical step in the sort of multi-stage research programme mentioned above, but that was not an option. Greek archaeological law allows members of foreign schools and institutes to apply for (an extremely limited number of) survey permits, or excavation permits, but not both at once. In other words, this is an unavoidable constraint that operates locally, and greatly affects strategies and practices of archaeological research in Greece, including the balance between excavation and other methods for data collection.
This observation leads on to the final comment I wish to make. The question ‘why excavate?’ is a good one to think with, especially in the context of the current state of archaeology in north-western Europe, where it arguably has the most relevance. But it is also a question about which it is very difficult to generalize or to provide answers that are widely applicable across the discipline. All answers are surely regionally specific and dependent on a whole range of extremely variable factors – the history of archaeology in any given part of the world (and thus the store of previously accumulated knowledge and the relative need for basic research); the disciplinary, governmental, legal and ethical frameworks under which archaeology operates in a country; the balance of salvage versus research needs; the sources and availability of funding; and so on. Let me illustrate the point by providing three brief snapshots of the very disparate situations I have encountered in my own fieldwork in recent years in Greece, Armenia and the Caribbean.
In southern Greece, where I have been active in survey archaeology on and off since 1974, one is constantly mindful that the work we do stands on the shoulders of seven or eight previous generations of archaeologists, and against the backdrop of many thousands of past excavations. Why excavate, indeed – especially given that foreign archaeologists are generally not allowed to involve themselves in the very salvage digs where their funding and expertise could be of greatest assistance (Cherry 2003)? Here the clearest need, as many of us came to realize during the 1970s and 1980s, was to provide a regional context for all those excavations by conducting high-resolution surveys to explore the rural landscapes of Greece, a palimpsest of the visible traces of the last 10,000 years of its history, which are increasingly being damaged by development, intensified agricultural practices, and a variety of other threats. Similar circumstances apply in neighbouring countries, and the responses have been comparable; thus there has been astronomical growth in the quantity of regional survey-based projects around the Mediterranean in recent decades (Alcock and Cherry 2004). In a number of countries in this part of the world, it would be fair to say (in the words of the panel organizers) that excavation is no longer quite ‘the same defining practice of our discipline that it used to be’ (editorial, this issue, p. 1).
In Armenia, where I codirected a project from 2004 to 2008 (Zardaryan et al. 2007), things could not be more different. A far less restrictive permit regime than in Greece allows for virtually any sort of fieldwork to take place. Because so little archaeology (except for burial excavations) has taken place in the south of the country where this project was located, there is a pressing need to establish artefactual sequences, site typologies, and relative and absolute chronologies – the kind of basic work, in fact, that took place in Greece many decades ago. Here, therefore, extensive and intensive regional surveys and excavations at several key locations have fruitfully gone hand in hand, the former to locate and map hundreds of hitherto unknown sites, the latter to investigate stratified deposits and provide firmly dated assemblages. In this situation, excavation is essential to provide material that can help anchor down materials encountered by survey that would otherwise remain chronologically vague, while survey provides the necessary information to have a better understanding of the sites we have excavated in their wider regional contexts.
Lastly, my current work on the island of Montserrat in the eastern Caribbean presents a different situation again (Cherry et al., forthcoming). Two-thirds of the island is now a no-go zone, having been devastated by a volcano that has been erupting continuously, and with devastating impact, since 1995; the population that has remained has been forced to relocate to the north, and active resettlement and infrastructural development there is already doing serious damage to archaeological resources. Yet there has only ever been a handful of excavations on Montserrat; there exists no adequate inventory of known sites, nor an archaeological service to make one, nor even a resident archaeologist on the island. Here is a rescue situation, if ever there was one. Our project's strategy thus prioritizes urgent extensive-mode survey, alongside the analysis of space and aerial imagery, to locate and map archaeological resources, as well as a risk assessment of the sites we encounter; test excavation of at least some of the sites that are in most immediate peril is certainly a desirable option, but mainly as a follow-up procedure.
I was invited to participate in this interesting panel discussion, evidently, because its organizers wished to hear the voice of someone such as myself whose career has not been dominated by excavation. I have certainly served my time in the trenches, including several years in the late 1960s employed in year-round rescue archaeology in the United Kingdom, as well as long spells as a trench supervisor on large-scale urban excavations in the classical lands. The fact that most of the projects I have co-organized over the past several decades have focused on regional survey indicates no antipathy to excavation per se, but merely reflects the fact that the types of research question I have been most interested in pursuing can best be addressed – perhaps can only be addressed – by acquiring archaeological data on a regional scale. Several of these projects have taken place in parallel with substantial excavations, under the primary direction of other colleagues, and that has been mutually very beneficial. I am still not digging, much, myself. This does not mean that I fail to appreciate the dilemmas faced by archaeologists working in those parts of the world where (until recently, at least) a relative abundance of resources for excavation, but a paucity of support for post-excavation analysis and publication, has led to the crisis of confidence reflected in the opening statement by this panel's organizers. In the final analysis, I think it is not helpful to frame the question in terms of whether excavation is, or can remain, the defining practice of archaeology as a discipline, as compared to the ever-growing number of other methods for acquiring archaeological data. So many locally specific factors intervene in any given situation, in different parts of the world, with variable research goals in play, that the only appropriate answer to the question ‘why excavate?’ must be: it all depends.
