Book Reviews
This well written volume explores continuities between Romano-British and early medieval landscapes using published excavations of pollen, animal bone, charred cereals and elements of Romano-British field systems. The collated data are presented in five external appendices accessible through the OUP website. Based on the proposition that ‘landscape character varies in both time and space’ (p 44), the book is structured around nine regions subdivided into pays. Their identification is based on geographic determinants — surface geology, relief, modern agricultural land classification — combined with cultural attributes: extent of Romanisation (urban hierarchies, the distribution of villas, temples, mosaics and pottery), degree of 5th-century cultural shift (eg new forms of settlement epitomised in Grübenhauser), changes in material culture, extent of woodland (from place names and Domesday Book), ‘scale of Anglo-Saxon immigrations’ and the ‘creation of villages and open fields in the later first millennium ad’ (p 46). The volume’s strength lies in its solid reprise of the wide-ranging evidence and established arguments for agricultural continuity across the Romano-British and early medieval centuries, the inadequacy of ‘Anglo-Saxon’ as a cultural descriptor and the utility of setting the period within the longue durée. It concludes that ‘Roman landscape character was different’ across central southern England (p 340), and that subsequent evolution of field systems occurred within ‘clearly distinct local and regional differences in agricultural practice across Britain in the first millennium ad’ (p 315). Readers should explore the variation of data between counties through the Appendices, since the evidence from some regions may be stronger than that from others. Table 3.4 (pp 76–7), for instance, summarises the numbers and proportions of major domesticated animals from 146 early Roman, late Roman and 5thto 8th-century rural sites (listed in Appendix III) across 15 counties from Cornwall to Norfolk; the text (p 78) refers to 319 sets of evidence. The number of reports from each county represented in the Table is variable: eg Appendix III cites around 28 sites in Northamptonshire, about 24 in Buckinghamshire, three in Cornwall and none in Devon. The conclusion that, on the fen islands, the proportion of late Roman ‘cattle increased very slightly from 38 per cent to 40 percent’ in the early medieval period (p 80) comes from a single site. Similarly, pie charts showing changes in the proportions of Roman (ad 43–400) and early medieval (ad 400–1066) woodland, arable and grassland pollen — usefully disaggregated by region and period in histograms (pp 66–7) — compare results from 15 Roman and 23 early medieval sites in the South West, for example, with only three Roman and four early medieval sites in East Anglia which is roughly equivalent in area. There are inaccuracies too. Sawtry, for example, is wrongly included with fen islands in Appendix III; elsewhere C Taylor is cited as proposing an ‘even more modest estimate of “no more than 10,000 Saxon settlers”’ in total (p 114) when what he actually wrote was significantly different: ‘Professor Charles Thomas has recently estimated that during the fifth century no more than 10,000 Saxon settlers came to this country’ [my emphases]. Despite these criticisms, however, the volume contributes solid support for the growing swell of research which suggests that most changes to the early medieval landscape represented evolution and innovation from a consistently Romano-British base.