Sorting the butchered from the boiled

Abstract

Is it possible to identify cooked, rather than burnt, bone? Mild heating (≤100 °C,1 h) – typical of cooking – does not lead to detectable changes in any biochemical parameter of bone yet measured. If it is only possible to detect charred bone, how is it possible to detect cooking in the archaeological record? In a previous paper (Koon et al., 2003, J. Arch. Sci.), we used a Transmission Electron Microscopy (TEM) based approach to investigate changes in the organization of the bone protein, collagen, as it is heated, using bone from heating experiments and short term burials. The work revealed that mineralized collagen, despite requiring aggressive treatment to gelatinise the protein (e.g. 90 °C, 240+ h), readily accumulates minor damage. We believe that the presence of mineral matrix stabilises the collagen enabling the damage to accumulate, but preventing it from causing immediate gelatinisation. Once the mineral is removed, the damage can be observed using appropriate visualization methods.

In this paper the visualization technique was tested in a blind study of bovine bone from the Anglo-Scandinavian site of Coppergate, York. The purpose of the study was to determine if the method could discriminate between bones thought likely, on the basis of zoo-archaeological and spatial evidence, to have been cooked (high meat yield bones from a domestic context) and those which were butchered but unlikely to have been cooked (low yield bones from a butchery site). The results of the TEM analysis identified two clear groups of bones, one set more damaged than the other. This finding was consistent with archaeozoological interpretation, with the exception of one bone from the domestic context, which was not identified as having been cooked.

Introduction

A key part of the daily life among past populations was food procurement and preparation. As a consequence the processes of hunting and butchery have been studied in some detail, both ethnographically and from the study of faunal remains. Cooking has received far less attention, yet it is the act of heating food that sets humans apart from other animals. Some have gone so far as to suggest that cooking transforms an element from being a product of nature to one of culture (Lévi-Strauss, 1969). Indeed anthropological studies have shown that eating habits may be used as a means of non-verbal communication, with which to reinforce relationships between individuals and groups within a society (Douglas, 1972, Lévi-Strauss, 1978). Gender roles, for instance, can be manifested within food processing and consumption (Hastorf, 1991, Montón Subías, 2002). The mode by which a meal was cooked, and the type of foodstuffs that are eaten, can reflect social and economic status within a society (Goody, 1982, Grant, 2002). The division of a carcass within a group can be used as a means to reinforce social order at meal times, with higher quality cuts restricted to those of higher social standing (McCormick, 2002, Stokes, 2000).

The ritualistic meanings behind cooking are not just restricted to the living; heated bone, be it animal or human, also has a role to play in understanding funerary practices. Throughout history animal bones are found associated with human inhumations (Dürrwächter et al., 2006). Deliberate interment of raw or cooked animal remains in a burial are interpreted as ‘grave gifts’, perhaps given for religious reasons, to feed the soul (Lauwerier, 2002) or to express status, either that of the deceased or of the deceased's relatives (Grant, 2002). Evidence for the preparation of these ‘gifts’ would aid interpretation of their significance. Beliefs can also be expressed in the ways that relatives offer their deceased to the gods. Whether the deceased is offered raw, cooked or burnt may reflect what the relatives believe to be their god's wishes (Oestigaard, 2000). Whilst burnt bone is often easy to recognize, bones which have been heated to lower temperatures (boiled, roasted) are often categorized and catalogued as unburnt and thus interpreted as inhumations (Oestigaard, 2000).

From the nutritional standpoint, cooking would have meant an improvement in the edibility of meat both in terms of the quality and extended use-life (Wrangham and Conklin-Brittain, 2003). In addition, boiling the residual meat and extracting lipids from bones would have provided a new and important source of nutrients (Lupo and Schmitt, 1997). Cooking and meat eating have been linked to key changes in the dentition, and enlargement of the brain (the expensive tissue hypothesis; Aiello and Wheeler, 1995). Therefore evidence of cooking and the processing of meat are fundamental to understanding both the cultural and biological development of past populations.

Features such as; fragmentation patterns (Gifford-Gonzalez, 1989, Oliver, 1993, Outram, 2002, Outram, 2004), butchery marks (Gifford-Gonzalez, 1989, O'Connor, 1989, Outram et al., 2005) and charring (Albarella and Serjeantson, 2002, Binford, 1981, Gifford-Gonzalez, 1989), are used to indicate food processing. Evidence of burning will show that a bone has been heated, but can also provide information about the condition of the bone at the time it was cooked. If charring affects only the articular ends this indicates that the bone was cooked with some or all of the flesh attached, for instance when a joint is roasted (Albarella and Serjeantson, 2002, Gifford-Gonzalez, 1989). Butchery marks can show how the carcass was dismembered and this may provide an indication of how the meat would have been cooked. The distribution of cut-marks and fractures on the upper limb bones can (for instance) provide an indication of whether whole joints of meat were roasted on the bone or whether pieces of meat were first removed from the carcass and then cooked (O'Connor, 1989, Outram et al., 2005). However cut-marks can sometimes be difficult to recognize due to sub-aerial weathering and diagenetic processes. Fractures can be produced during butchery and dismemberment and again these can also provide an indication of cooking practice. For instance long bones processed for boiling in cooking pots often show irregular transverse fractures resulting from their breakage prior to boiling, because they have to fit into cooking pots (Gifford-Gonzalez, 1989, Turner and Turner, 1999).

Charring is the only one of these features however, that directly implies cooking, yet is unlikely to be present if bones were boiled or covered by flesh, and may be misleading evidence if fatty bone has been utilised as fuel (Costamagno et al., 2005). What the other features may provide is evidence of human intervention. This zoological evidence, along with the patterns of butchery and the spatial distribution of bones on a site, are commonly used to classify the faunal remains into one of the food processing stages. If it were possible to differentiate between cooked and uncooked bone by looking at heat-induced damage to the bone, these classifications could be tested. Furthermore, the total thermal history of the samples, as expressed by the overall degree of alteration, may even permit a more informed interpretation of the cooking methods themselves; for example in terms of thermal history, roasting < boiling/stewing < stock production < glue.

The problem with cooking is that, unlike burning, temperatures remain sufficiently low to transform the food whilst retaining some moisture. Previous studies have focused upon temperature-induced mineral changes, but significant alteration does not occur below 500 °C (Shipman et al., 1984, Piga et al., 2008). Attempts to identify organic changes have so far proved to be unsuccessful, principally because such changes appear to mirror diagenetic effects and require temperatures in excess of 200 °C to achieve significant alteration (Holden et al., 1995, Munro et al., 2007, Nicholson, 1993, Shipman et al., 1984).

Previous attempts using mineral and protein changes to boiled bone (Nielsen-Marsh et al., 2000, Roberts et al., 2002) have proved too crude to detect these subtle changes. The reason probably lies in the fact that the dominant organic component of bone is mineralized collagen. Our work reveals that this has an unusual diagenetic pattern, of all-or-nothing decay, caused by the extraordinary stabilization offered by bone mineral (Covington et al., 2008). Consequently heat-induced changes to the collagen molecule sufficiently large to be detectable by physical or wet chemical methods cause so much damage that they convert it to gelatine, which is then lost from the bone.

The collagen molecules are packed into long rope-like structures called fibrils, using electron microscopy it is possible to detect more subtle heat-induced changes to collagen at the level of the fibril. A TEM analysis of the integrity of collagen fibrils was first used to look at factors affecting the toughness of meat (Snowden and Weidemann, 1978). Later Richter (1986) applied the same approach to examine the effect of boiling on fish bone collagen. More recently the technique has been refined and used to identify roasted sheep bone in heating experiments and from experimental burials (Koon et al., 2003). The results showed that mild heat treatment can leave detectable alteration as localised unpacking within the collagen fibrils.

The aim of this study was to determine if the same approach could be applied to archaeological material. It is easy to cook bone experimentally but proving a method on archaeological remains is much more problematic. Self evidently lacking a method to identify them, there are no definitive cases of archaeological cooked bone. Furthermore it has proven difficult to distinguish those changes caused by cooking and diagenesis (Koon et al., 2003, Roberts et al., 2002, Shipman et al., 1984). In this study a blind test was undertaken on two sets of archaeological bones (those with a high probability of being cooked and those with high probability of being uncooked) in order to test the validity of this new method.