New thoughts about the use of archaeological stratigraphies! Is this so? The discussion article by Patricia A. McAnany and Ian Hodder aims at the construction of a theoretical framework to expound and discuss the problems of archaeological stratigraphy. Such a theoretical framework is urgently needed, they feel, and has been largely neglected until now. Reading and interpreting an archaeological stratigraphic record, if carried out according to the guidelines they try to establish, may reveal much more information about past social processes that led to the formation of the specific stratigraphy. In the authors' own words, ‘thinking about stratigraphic sequence in social terms is more than an academic exercise’ (quoted from abstract). As the record left behind by ancient communities, archaeological stratigraphies, in their view, take a middle place on a scale from micro-records endowed with meaning (artefacts) to macro-records of contextual meaning preserved in archaeological landscapes. The in-between, the immediate residues of meaningful past human behaviour encapsulated in archaeological sites, remain, in their view, undertheorized.
As someone who has been involved in practical fieldwork for more than 20 years, the authors’ statement that archaeological stratigraphy remains largely undertheorized, and that therefore many chances to gain deeper insights into past lives are passed up, comes as a bit of a shock at first. Have we really overlooked so many aspects during the last decades of fieldwork and dug away so many meaningful records unnoticed? And is there indeed the potential to gain a better understanding by using the ‘toolkit’ proposed by McAnany and Hodder? With the following short contribution, I hope to discuss the authors’ request from a perspective of archaeological practice.
The authors set out to scrutinize one of archaeology's most fundamental tools: archaeological stratigraphy, principally meaning the superposition of layer upon layer of human-generated earth (and other) layers within an archaeological site. Practising archaeologists rely on the interpretation of archaeological stratigraphy in order to distinguish meaningful patterns of spatial and sequential order that, necessarily, reflect conscious and unconscious decisions taken by the creators of any archaeological record, intermingled with natural processes. By doing so, they reach reasonable conclusions regarding the development of the archaeological site over time, possible functional and/or social distinctions obvious from patterns of spatial and artefact distribution, and so on.
McAnany and Hodder challenge this valuation of archaeological stratigraphy as a neutral method for unveiling patterns of past behaviour. They instead advocate the view that the formation of an archaeological stratigraphy – which in their words is always a built environment and therefore cannot be compared directly with geological stratigraphy – must in itself be understood as a record of meaningful social behaviour in the past.
McAnany and Hodder examine today's archaeological practice, which relies on archaeological stratigraphy as one of its foremost tools to reflect, at least in part, the logistic necessity of dividing and delegating labour on colonial-type large-scale excavations or under rescue conditions (and thus of delegating and postponing interpretation). With such a hierarchical decision-making chain in practice, archaeological stratigraphy can by no means be a neutral tool to classify and understand the archaeological record.Footnote 1 This leads, in their view, to an allegedly currently prevailing archaeological practice of treating archaeological strata as ‘containers of artefacts’ rather than as units meaningful in themselves. Thus an abundance of information on past behaviour must be lost to us. Instead, they advocate a perspective that regards archaeological stratigraphy as the residues of past social behaviour that they, consequently, call ‘social stratigraphy’.Footnote 2 In distinguishing and systematizing the various past activities that lead to the formation (or rather, construction) of archaeological stratigraphies, it should become possible to recognize patterns of meaningful and intentional past human behaviour that lend themselves to further interpretation.
The authors base their challenge to the traditional use of archaeological stratigraphy on a survey of current archaeological practice as they understand it. This perspective is not globally valid, but implies an enormous bias since their brief survey of the history of archaeological stratigraphy is restricted to British and American schools (as is explicitly stated). If the authors were engaged in fieldwork in England or North America, such a narrow perspective would be comprehensible. However, both conduct fieldwork away from their respective countries of residence, in settings where international expeditions are customary and where days off are often spent visiting neighbouring expeditions.Footnote 3 Indeed, extending the scope reveals almost immediately that the neglect of archaeological stratigraphy ascribed to practising archaeologists may actually be a chimera. To add a few remarks on the historical survey: the usage of the principles of superposition for the ordering of artefacts from archaeological strata was not only employed by the famous Augustus Pitt-Rivers, but by the German Wilhelm Dörpfeld in the much-taunted Troy excavations of 1882–84 (Dörpfeld 1902) and by the Dane Sophus Muller in his attempt to understand the Danish burial mounds (Muller 1888–95). Along the same line of reasoning, Edward Harris was not the only person to attempt to systematize the analysis of archaeological stratigraphy. Already in the 1970s a series of international workshops tried to establish guidelines for the description of archaeological stratigraphies for west Asian archaeology (for a preliminary summary see Gasche and Tunca 1983). Although there is still a major focus on the possibilities of establishing relative chronology, the preliminary guidelines clearly distinguish anthropic, natural and complex causes behind the formation of various layer units. It should perhaps come as no surprise that the most explicit perception of archaeological stratigraphy as a ‘constructed space’ is notable in the field of building archaeology (based largely on Hachmann 1969; Echt 1984). In his introduction to the study of the Uruk excavations, Ricardo Eichmann sees archaeological stratigraphy as ‘Beschreibung stratigraphischer Phänomene nach ihren Ursachen, als Folge menschlicher Handlungen und geologischer und biologischer Prozesse’ (Eichmann 1989, 1). He therein clearly discriminates between anthropogenic and non-anthropogenic causes; also clearly distinguished are depositing (e.g. filling) and cutting (levelling), respectively constructive and destructive, episodes. Cornelius von Pilgrim, in his study of the Middle Kingdom town on Elephantine Island, explicitly criticizes the Harris system for its neglect of processual aspects of stratigraphy formation and its concentration on relative chronology (Pilgrim 1996, 26). To finally return to the anglophone sphere, Marc Verhoeven's study of abandonment patterns in the village of Sabi Abiad goes far beyond Michael Schiffer's artefact-focused formation processes (Verhoeven 1999, esp. 46–60). And among the specific determinates of stratigraphic formation on tell sites discussed by Sharon Steadman were aspects of cutting, retrieval and avoidance that reappear here (Steadman 2000). In brief, many of the thoughts claimed to be new were indeed voiced long ago, possibly in a less explicitly theoretical manner but nevertheless highly systematically, by numerous people engaged in active archaeological fieldwork, and have successfully been applied to the interpretation of archaeological stratigraphic records. The idea suggests itself that the undertheorized archaeological stratigraphy is indeed a straw man.
Nevertheless, the classification presented in the ‘toolkit’ may be of some use in generalizing about what we understand when we dig. It may, at least, serve as a reminder of all the possible questions we should be asking during documentation. According to the system provided in figure 3, techniques can be divided into depositing, cutting and relocating, and a combination of all these. They are created by past operations (here called processes) of raising or lowering, hoarding or retrieving, continuity of use or avoidance, and more. Possible interpretations presented in the third section of figure 3 are manifold; there is, necessarily, no means of establishing a one-to-one relationship between specific operations and their purpose that can only be approached through interpretation. For example, the intentional collection and burying of a group of artefacts can serve the alternative intentions either of remembering and memorial, or of hiding and forgetting. It can equally relate to a failed purpose of hiding for future retrieval. The possible interpretations are hence mutually exclusive. It is in the nature of archaeology that any convincing interpretation of the record therefore depends as much on the careful ‘reading’ of the record and on contextual information and pre-existing knowledge.
This said first, let us now put McAnany and Hodder's toolkit to the test. Let us do so by running a virtual excavation in an archaeological site and trying to understand our own doing through the lens of ‘social stratigraphy’, by using the proposed terminology. Imagine yourself standing in the middle of a small central Anatolian plain assigned to you as a working area. The task is to make a contribution to the better understanding of local (pre)history by carrying out an archaeological investigation, and you have been allotted a comfortable, though not infinite, budget for this purpose. The work ahead of you consists according to traditional practice of three subsequent steps – pre-excavation, excavation and post-excavation.
Pre-excavation: before even setting out to do practical work, you have to make a series of choices – first, where do you decide to work? We take it for granted that, as a responsible scholar, you are aware of the archaeological monuments common in the area so that you can make an informed choice. In the small central Anatolian plain there exist, most probably, visible sites, notably settlement mounds such as are common there; to these may be attached a lower off-site area; equally possible is the existence of flat settlement and activity areas in the vicinity of a visible mound. There may, further, be invisible sites, for example graveyards or flat settlements that can be detected through intensive fieldwalking survey. No matter which site you choose, you will decide on a specific place where you want to work – and the probability is high that this will be a visible place.
Step two in your chain of choices is the specific area of the site that will be investigated. Making an informed choice will require the application of intensive survey, remote-sensing techniques and possibly also the removal of the surface of parts of the site through scraping. Subsequently, you will select a specific area for excavation. Ideally, your choice will be determined by which area promises to yield the best evidence to answer questions posed by your research design. Most probably, factors of feasibility and the hope of uncovering something ‘important’ (relative to your research design) will equally contribute to this decision.
Excavation: it is only now that you will begin to organize invasive work. Up to this point, you have not yet applied any technique of cutting or depositing (except for the surface scraping that will, however, hardly leave a recognizable record in the long run). However, your decisions will have been guided by a whole set of operations related to remembering and possibly memorialization (which is what history is all about) as well as a wish to make something to endure. None of these are visible in the earthly history.
The excavation of your chosen area, no matter how this is organized in specific cases, will be a composition of cutting and depositing activities. From the surface down, you will uncover – expose, lay bare – and possibly lift remains of past activities (or will assign helpers to do so). Necessarily, these remains will also be partly or completely destroyed through cutting and excavating. Ironically, your motivation to do so is indeed the wish to retrieve knowledge about past human lives, not the wish to destroy. In this sense, your cutting/excavating/lifting activity is indeed linked to the construction of a historical memory. The material from the archaeological strata or contexts removed during excavation must be deposited, usually after as much data retrieval as is possible: you will extract samples and artefacts, and only a much reduced part of the originally removed earth or other matter will finally be deposited, away from its original place of primary deposit.
Archaeological excavation hence assembles all aspects of cutting and depositing and results in the creation of a palimpsest, the gaps in which can hopefully be bridged by information from the paper and digital excavation records. Additionally, running an excavation in a specific place conforms to the process listed as continuity of space: clearly, your archaeological activity is the result of historical memories about this specific locality being transmitted or retrieved through research.
Post-excavation: the last field covered by a modern archaeological excavation is the conservation and presentation of the archaeological results. The awareness that archaeological excavation necessarily means destruction has prompted increasing measures of counterbalance through programmes of restoration or reconstruction of the excavated archaeological sites. You will hence most probably engage in measures to stabilize selected parts of the excavated archaeological record and protect them against the rigours of further natural and human destruction. Feasible measures include covering the excavated areas temporarily through light and accessible constructions or permanently through complete refilling, stabilizing the archaeological remains through chemical and mechanical conservation, and possibly also the complete reconstruction of selected parts of the site. To this must be added the analysis of samples and the study and restoration of finds, plus their final display (most probably not in their original place but relocated to specifically assigned facilities such as a museum). All these measures conform to the notion of construction of space (and knowledge). At the same time, they are part of your specific ongoing construction of historical memory. How do you decide what to preserve and what to reconstruct (and therefore what to select for transmission and presentation)? Most probably, accidents of preservation (archaeological stratigraphies tend to be highly selective) and an immanent tendency to present rather the exceptional but the ordinary play their part in this decision. But there may also be constraints originating from the social or political environment,Footnote 4 or from politics of fundraising, and other inherent necessities. The relevance that you assign to certain specific parts of the archaeological record above others will be reflected in the (re)constructed archaeological remains. Following all this, the legal protection of the site may finally lead to a pattern of avoidance such that no modern construction will be allowed atop it.
To take this mind game a little further, let us see what evidence for your archaeological activities might finally be found and interpreted by future archaeologists working according to the principles of ‘social stratigraphy’, provided that there is no direct historical tradition informing them. Would they be able to interpret the record accordingly? We have already seen above that the pre-excavation part of the archaeological process leaves no earthly trace at all. The excavation proper combines processes of cutting and related depositing-cum-relocation that comply with patterns of destroying, outbalanced only through careful recording. Any future archaeologist would probably be able to recognize traces of cutting, erasing and possibly retrieving. However, there is hardly any means of relating this earthly record to the virtual record created alongside it. Future excavators will therefore probably hardly be able to recognize anything beyond the destructive aspect of your work. In identifying your excavation area, they may, however, diagnose a pattern of continuity of space used (although probably not uninterrupted), and intertwined with subsequent avoidance through site protection. Traces attesting the retrieval of finds and samples will, however, most probably be lost. Finally, depositing and construction processes such as, respectively, restoration and reconstruction may hint at your motivation to memorialize selected aspects of an ancient site for future times.
This virtual excavation experience demonstrates that the suggested system is functional as far as the processes described are concerned. Although not at all new, it has the merit of classifying aspects related to the processes responsible for its creation, although the interpretation of it remains unlinked and dependent on the accidents of archaeological transmission. Almost certainly, the careful use of other systems may lead to comparable results. If we decide to cling to this one, a future extension of the toolkit is possible and desirable. The notion that archaeological features often consist rather of the interfaces between deposits than of the deposits themselves deserves further elaboration (Pilgrim 1996, 16). Of equal importance is the recognition of the episodic nature of many archaeological features, an aspect that often may go unnoticed. The toolkit, and possibly future extended versions of it, may help to create a fuller awareness of the fortuitousness of our archaeological sources.
No matter which system is employed, past behaviour encoded in the processes of stratigraphy-making remains open to interpretation. Given the flaws immanent in any archaeological record, we will not overcome the notion that the record in itself is highly selective. Whatever construction of historical memory we aim at can only be a faint shimmer of past Lebenswelten. It has, at the same time, the advantage of a long-distance perspective that may enhance the visibility of patterns of past behaviour. In conclusion, thinking about archaeological stratigraphy allows us possibly to understand better the formation and construction of individual archaeological stratigraphies. In the lucky case of one large or of several related stratigraphies, comparative studies of the individual sequences may allow us to gain some insight into the dynamics behind their formation. Beyond the resolution level of the individual site, however, any gain in historical knowledge remains negligible.
