From Antiquity, an archaeological journal.
The bold paragraph is my emphasis.
Against the grain? A response to Milner et al. (2004)
Introduction
A recent publication in this journal (Milner et al. 2004) called into question the increasing body of human stable isotopic data showing a rapid diet shift away from marine resources associated with the beginning of the Neolithic in parts of north-western Europe, particularly in Britain and Denmark. While we very much welcome informed and positive debate on this issue, we feel we must respond to this specific paper as it is problematic at a number of levels.
Stable carbon and nitrogen isotope analysis of human bone is beginning to challenge what we would argue is the current orthodoxy of a gradual dietary transition between the Mesolithic and Neolithic. Indeed, the stable isotope data support some elements of a previous orthodoxy, which saw the advent of the Neolithic as a 'revolution'. This is not to say that all elements are supported by the isotopic data; the question of the interactions between any incomers and indigenous people, for example, is still very much a live issue. And it is still far from clear exactly how the shift occurred, how rapid it was in human terms (in generations rather than radiocarbon years), and why it occurred. And there is still the possibility of regional and supra-regional variation to be addressed fully. But the implications of the stable isotope data are beginning to be acknowledged and addressed (e.g. Thomas 2003). This is an important independent line of evidence, and has been available since the early 1980s (Tauber 1981a), yet until recently little consideration has been given to the picture of a very rapid and significant shift in diet across the Mesolithic-Neolithic transition. Instead, it is during this very period that the view of the transition as a long, drawn-out process began to emerge and dominate discussion (Thomas 1991).
It is in this context that criticisms made of the isotopic data, particularly by Milner et al. (2004) need to be addressed. Their dismissal of the isotopic evidence for a rapid and significant transition, while to some extent encouraging debate, also prematurely attempts to close it. Milner et al. (2004) present their critique along three main fronts (see also Bailey & Milner 2002). Firstly, they contend that the zooarchaeological and archaeological evidence for diet is at odds with the stable isotope data; secondly, they point to problems of sample size and bias in the human skeletons used for analysis; and thirdly, they argue that there are problems with the interpretation of stable isotope data. We address each of these concerns in turn.
The (zoo)archaeological data
Milner et al. (2004) make much of the zooarchaeological evidence for the continued use of marine resources in the Neolithic, taking examples mainly from Denmark but also from Britain and Ireland. They argue that the presence of the remains of marine foods (especially shellfish) in Neolithic contexts, and the occurrence of apparent seal-hunting stations and of fish traps, somehow counters any argument of a large-scale dietary shift at the start of the Neolithic. Despite the numerous problems and biases with the use of zooarchaeological data, they present this evidence as if it were some sort of 'spoiler'; that finding any evidence, however slight, of any Neolithic person consuming marine foods undermines the isotopic data of a large scale shift. Simply put, the continued occasional use of marine resources in the Neolithic is not at all incompatible with the isotope data, but is largely irrelevant in the overall question of large-scale dietary shifts. The isotopic evidence presents a long-term measure of lifetime diets, and clearly shows a significant change in human diet between the Mesolithic and the Neolithic. Remains of fish and shellfish recovered from archaeological sites are the remains of individual meals, but are not indicative of the overall diet of a human population. As Geoff Bailey himself has elegantly argued (Bailey 1975, 1978), shells are highly visible archaeologically due to their preservational properties, but misleading in terms of determining diet composition, as they are nutritionally poor. Bailey (1978) writes that:
'The ease with which molluscs can be over-rated as a source of food will be swiftly appreciated from the fact that approximately 700 oysters would be needed to supply enough kilocalories for one person for one day, if no other food were eaten, or 1400 cockles, or 400 limpets, to name the species most often found in European middens. I have estimated that approximately 52,267 oysters would be required to supply the calorific equivalent of a single red deer carcase, 156,800 cockles, or 31,360 limpets, figures which may help to place in proper nutritional perspective the vast numbers of shells recorded archaeologically.' (Bailey 1978: 39, emphasis ours) Therefore, the occasional Neolithic shell midden is in itself hardly indicative of a continued marine-based economy in this period. The nature of the exploitation may have been very different, for example, from a central aspect of subsistence in the Mesolithic to one more peripheral in the Neolithic.
In addition, it should be emphasised that, aside from these shell middens and special purpose sites, there are actually very few Neolithic faunal assemblages known from Denmark. Bone survival is poor away from the shell middens, but where mammalian fauna is preserved from the Early Neolithic, it is dominated by domestic fauna (see Fischer 2002 for a recent review). Thus Milner et al.'s (2004) discussion touches upon only one aspect of the Neolithic economy, and likely a very limited one.
In the context of Britain, where much of our own research on this issue has been focused (i.e. Richards & Hedges 1999; Richards et al. 2003a; Schulting & Richards 2002a, b), Milner et al. (2004) do agree that there is substantially less evidence for marine exploitation in the Neolithic. They suggest that this is partly because of inundation of coastal sites by rising sea levels. However, sea levels were quite close to their present position by 4000 cal BC (the generally accepted data for the appearance of Neolithic material culture in the UK), so that this argument holds far less relevance than it does for the Mesolithic period, when it is very much a factor (Schulting & Richards 2002a, b). Milner et al. point to shell middens of Neolithic date along the Firth of Forth in south-east Scotland and along the coast of Co. Sligo, western Ireland, and to evidence for fishing from Neolithic Orkney. The shell middens are subject to the same issues already raised above--their simple presence, while certainly interesting and worthy of further investigation--says little about their quantitative importance in long-term diet. The Forth and Sligo middens are notable for the absence of much in the way of cultural material, or indeed …