Not all archaeological remains cause discussions concerning stratigraphy. In Sweden, for example, intense stratigraphy discussions have taken place among archaeologists working mainly with urban sites (see e.g. Larsson 2004), and many of the illustrative examples in the text by McAnany and Hodder are rather well-preserved remains with complex stratigraphic sequences. This is, of course, due to the fact that different remains are stratified to different extents and are thus valued differently regarding this issue. Poorly preserved, sketchy remains scattered in the ground may lack complex stratigraphic relations and are regarded as less relevant for this discussion. However, all archaeological remains have some stratigraphical relation and – as McAnany and Hodder mention – interpretation of stratigraphic sequences is a part of archaeological identity. A greater interest in how stratigraphic sequences are formed in social terms should be relevant for all archaeologists. I believe that archaeologists working with complex stratigraphic sequences, and those who work with less stratified remains, have something to gain from this discussion, but in different ways. McAnany and Hodder state that stratigraphy may be both overdescribed and undertheorized. The problem of overdescription concerns complex stratigraphies, while I think less stratified remains are suffering from a lack of discussion concerning stratigraphy all together.
During an attempt to implement a systematized reflexive field method in a contract archaeology project in Malmö, Sweden, much inspired by the reflexive methodology of Çatalhöyük, we encountered some related problems. As part of a method to encourage and document the process of interpretation, we introduced digital diaries for the field staff for use during fieldwork. The diary did not function the way it was intended to, at least not in all cases. There were several problems and various reasons for this that I will not mention here because of the limited scope of this comment. (These problems are described elsewhere; see Berggren 2009). But one explanation that was discussed stands out as relevant to the topic discussed here. It concerns the nature of the remains and how this affects the method of documentation and interpretation. The project consisted of nine different excavation locations. Two locations had medieval remains that were better preserved and more stratified than the scattered prehistoric remains of the other sites. The use of the diaries varied among the sites. Some diary entries were written in a descriptive manner, rather than reflexive, and contained a repetition of the feature or unit sheet instead of reflections on the interpretive process. These descriptive diary entries were written at most of the sites, but there were more descriptive entries written at the two excavations with medieval remains. In an article, the site supervisors at one of the medieval sites offer an explanation for this imbalance. Using a single context method for the complex stratigraphies at this site, the archaeologists regularly summarized groups of contexts. It was thought that these summaries contained the information that was supposed to be written in the diary and could replace it. However, the summaries were very detailed and technical, and in most cases lacked the level of reflexivity we wanted in the diary. The site supervisors suggest that this lack of reflexivity was caused by the complex character of the medieval remains that demanded a more detailed investigation at an initial stage. In comparison the prehistoric remains at the other sites were more spread out and allowed a more general perspective which they thought could generate other questions at an initial stage of investigation (Ingwald, Koch and Lövgren 2002). Now, I do not believe that complex remains and complex methods prevent archaeologists from being reflexive. But I do think that Ingwald and his colleagues have identified a part of the problem that McAnany and Hodder are trying to solve. Complex stratigraphic sequences are often meticulously described, resulting in very elaborate and complex matrices and so on, which perhaps are sometimes regarded as final results instead of vehicles to reach interpretations. It is very time-consuming to deal with complex stratigraphies, and at the initial stage the technicality and the details took over at the site of Ingwald and his colleagues, even though they later reached interesting interpretations of their material (Ingwald and Löfgren 2009). I do not think archaeologists regard stratigraphic sequences as socially neutral, but the social formation processes may be overshadowed by the amount of recorded detail. Could other, socially charged terms for describing the stratigraphic sequences at the site of Ingwald and his colleagues have promoted a more reflexive attitude at the initial stage of interpretation? This is a hypothetical question that may not be answered here. But adding terms that are clearly not descriptive, but rather interpretive, to the initial documentation could perhaps augment a reflexive approach to the material.
The other type of remains I would like to mention consists of the more scattered, prehistoric remains typical of the heavily cultivated landscape around the town of Malmö. Most remains that are preserved are dug into the ground underneath the topsoil, consisting of pits of various kinds, postholes and so on. The features are chronologically mixed and only rarely cut each other. So the stratigraphic relations between features are rare, but there are always some stratigraphic relations within the feature, at least one between cut and fill. In many cases there are several fills which add to the stratigraphic sequence.
Remains of these kinds are rarely discussed in debates about stratigraphy. The features themselves are often regarded as the main entities of interpretation and much effort is placed on how they belong together, or not, much like the palimpsest case described by McAnany and Hodder. The features are often seen as stratigraphically uncomplicated, even though they are excavated by layer when possible, and finds are often collected by layer as well. The focus of debates on complex stratigraphy in urban archaeology and other well-preserved sites has left the ‘feature archaeology’ out of the discussion. But these pits need to be problematized in this regard as well. Because even though the layers are separated and sampled for macro-fossils and so on, the layers are mainly regarded as carriers of finds, a part of the object-oriented tradition mentioned by McAnany and Hodder. The layers are not in themselves regarded as finds. Perhaps they should be.
Layers in a pit are often just termed ‘fills’, and they are often not discussed or interpreted further. ‘Fill’ is thus used as a rather neutral concept. An exception may be the interpretations of burials, where the infilling, as well as the fills, may be discussed as parts of a burial ritual (see e.g. Gansum 2004), but fills in other pits are rarely interpreted with any detail. If we were to regard the fill explicitly as a find in itself, and describe it in social terms at an initial stage of investigation, I believe we would gain a different attitude towards these layers of soil and perhaps even a better understanding of their function, both practical and social.
So far, I agree with McAnany and Hodder that there are some problems of interpretation of stratigraphic sequences. Interpretations of both complex stratigraphies and less-stratified features may gain from additional attention to their social formation processes. But is the suggested terminology the solution, and, if so, how should it be used? My discussion here is hypothetical and theoretical, as is the suggestion by McAnany and Hodder. They suggest that such terms as remembering and forgetting and other socially charged words should be encoded and used in databases. Could this work? Has anyone tried? I am very curious. I think the suggestion by McAnany and Hodder would have benefited from an example where this had actually been implemented, a real case, instead of the more general examples used in their text. I think the success of such a documentation and registration would depend on the level at which these terms were introduced. I have some difficulty seeing every single unit having to be labelled with socially charged terms. As McAnany and Hodder state, the interpretation of these processes is difficult, which perhaps in part is the cause of stratigraphic relations being undertheorized. Above all, the significance of these processes always depends on the context. That is why a greater overview is needed for interpretations like these. There has to be enough information to form a context, and then the social interpretation may be introduced. Radically different social processes – such as referring to something in a positive fashion by affiliating a construction with an older structure, and overtaking something in a negative way by building over older structures – may result in the same material results. We have to know enough of the people we are trying to understand to decide what social principles were at play in each case. This may only be achieved at a more general level than single units or even groups in a matrix.
Another concern is connected to the difference between memory processes, at what the authors, using Connerton's words, call a habituated and a commemorative level. These processes may be at play separately or together. Many practices that build on history or memory in a society may be habituated, as pointed out by McAnany and Hodder, but the terms suggested in figure 3 give the impression of being at the commemorative level, creating specific social memories. If we use them while recording stratigraphic sequences, do we not risk creating an image of the prehistoric people constantly going round explicitly and intentionally creating memories? We have to find terms that fit the routine and habituated processes as well.
Perhaps a test case would be to study our own society and our own profession using these terms. It could perhaps be useful both to evaluate the functionality of a social stratigraphy, and to look at ourselves through a lens created by concepts we use describing prehistoric societies. In archaeology we find such techniques as depositing, cutting and relocating, such processes as lowering, hiding, hoarding, retrieving, recutting, erasing, returning, remaking and avoiding that may be interpreted as remembering, history-building, memorializing, forgetting, renewing, dominating, displaying, cleaning and destroying. This could tell us something about ourselves as it would illuminate the values behind our decisions to ask certain research questions and not ask others; that is, what would we like to remember and what would we like to forget?
To conclude, do we need a social stratigraphy? Well, I do not think that archaeologists in general regard method as neutral. But some may feel that documentation methods are lacking something, for example the time aspect as pointed out by Lucas. Others mention social agency as something that does not show in our documentation, and that an explicit theory has to be added for this phenomenon to be investigated (Larsson, personal communication). So I believe that McAnany and Hodder have identified a real problem, and that it could be useful to encode the social aspect of our interpretations already in the databases. At what level and with which terms needs to be discussed further, and perhaps tried on a real case.
