on 3 January 2012
For anyone with an interest in Irish and British prehistory and, specifically how the chronologies are assembled through radiocarbon dating, the publication of Gathering Time: Dating the Early Neolithic Enclosures of southern Britain and Ireland has been long anticipated and much, much desired. It is hard to overstate the importance of this book and how it has already rewritten our understanding of Neolithic enclosures, but it also stands as a template for other intensive studies to follow and emulate. The central importance of this study is not simply that it uses a lot of new radiocarbon dates for various sites, but it is how this data is treated and processed on such a large scale that is already leading to new and exciting insights into prehistory. As many readers of this blog, both professional archaeologists and enthusiasts, will be aware, the advance of absolute chronologies in archaeology has, in large part, been due to the development of radiocarbon dating. Prior to the seminal work carried out by Willard Libby and his team (James Arnold and Ernie Anderson), archaeological sites and were only datable through relative chronological means, such as seriation etc. In 1960 Libby, Arnold and Anderson won the Nobel Prize in chemistry for their work on radiocarbon dating. The basis of the method was that the measurement of the amount of the radioactive isotope carbon 14 (14C) surviving in a sample could be utilised to determine when, say, a piece of wood had been cut or grain harvested. These early dates relied on the assumption that the amount of 14C in the atmosphere had remained constant throughout history and, as the discipline was in its infancy, the associated standard deviations were also quite large. Over the years parallel advances in calibration of dates against tree ring curves, more sophisticated methods and machinery, along with increased care and refinement in the selection of materials and samples has led to better results. Today radiocarbon determinations have better accuracy and precision than ever before. Nonetheless, even with careful sample selection and the use of high-quality AMS dating, there is still the possibility that, when calibrated, the date will range over several decades to centuries. Since the 1990s a number of researchers have explored and developed a statistical system known as Bayesian modelling. The approach derives from the ideas of Thomas Bayes, an 18th century Presbyterian minister and mathematician. Simply put, this method allows the calculation of how the degree of belief in a given proposition changes due to additional evidence. In archaeological terms, the application of Bayesian modelling allows the refinement of radiocarbon dates through the addition of contextual information. Such information may include multiple dates for individual deposits, stratigraphic relationships, or even closely datable artefacts such as coins or pottery. To take an example from my own experience: at Gransha, Co Londonderry, I excavated a small pit group. A radiocarbon date from charcoal recovered from one of the feature the indicated that it had been deposited in the Early Neolithic period (4930±70 BP), but the date range was some 405 calibrated years (3943-3538 cal BC). As part of the INSTAR Cultivating Societies project at QUB additional radiocarbon dates were commissioned and then modelled by Rick Schulting and Paula Reimer (Chapple 2008, Appendix 7). The end result was that the potential lifespan of the site was reduced from 405 years to 0-50 years - a vast improvement on the earlier result from a single radiocarbon date. [Introductions to Bayesian modelling may be found here and here].
What Gathering Time set out to do was exactly like the example above, but on an enormous scale. Not only was the aim to produce robust chronologies for individual sites, but to then place them in wider chronologies and within their geographic and typological settings. The book presents 871 radiocarbon dates from nearly 40 causewayed enclosures. To assess how causewayed enclosures functioned as part of the wider Neolithic landscape and society models were also prepared for a range of monument types, including long cairns and long barrows. This brings the total analysed radiocarbon dates to a startling 2350. As such it is the largest Bayesian modelling project ever undertaken. The central findings of the project are that the main period of causewayed enclosure construction lasted from the late 38th century cal BC to the mid-to-late 36th century cal BC. Although a number of sites had an active life of several centuries, many were used for relatively shorter periods - some for only a matter of decades. When this data is incorporated into wider models, encompassing the entirety of the evidence, it is shown that the causewayed enclosures only appeared three centuries after the first Neolithic practices were established in southern Britain. The process of `Neolithisation' is shown to have begun in south-eastern England and spread regionally over two centuries.
Google: "Robert M Chapple blog" to find the rest of this review