The fields of Britannia represents the report of the eponymous research project that addressed issues of landscape continuity from the late Roman to the early medieval periods in England and Wales, in relation to a) field systems and land boundaries, and b) vegetation history as a proxy for land use. To this end, the authors have collected the most reliable excavation results and pollen sequences, obtained mostly from relatively recent developer-funded work. They have attempted to control for geographic variation by considering them in relation to eight regions, defined by physiography, the extent and nature of Romanisation, the apparent degree of ‘Germanic’ cultural influence, and the character and development of the ‘historic landscape’. Each region is sub-divided into pays. In the cultural landscape, continuity is assessed by seeing how far the orientation, or the actual boundaries, of (ideally, late) Roman field systems are perpetuated as historic field boundaries or the edges of medieval furlongs. (It is fortunate that most Romano-British field systems are coaxial enough to facilitate this methodology.) The authors insist that all excavators should consult nineteenth-century maps to see whether their field ditches line up with, or underlie, destroyed historic landscape boundaries; by not doing so, several have not properly understood their sites. Vegetational continuity is assessed by looking at variations in percentages of indicator species for land-use categories, such as improved pasture or woodland. Animal bones are included in the survey but on a less comprehensive basis, and alluviation and snails are also considered. There is also a very good discussion of taphonomic and interpretational issues.

The results make a useful contribution to contemporary debates. Part of the strength of this kind of survey is that it will be possible to build on it in 30 years’ time, using more data, or to interrogate the existing data using different sampling units or modified taphonomic principles. This survey, then, gets away from anecdotal, site-by-site narratives, instead applying a consistent, methodical analysis in order to paint “the big picture” (p. 17). The fullest and most interesting documentation comes from the south-east, central and south-west regions (Chapters 4, 6 and 7 respectively). The headline news is that woodland did not regenerate much in the fifth century (the mostly small rises being explicable by the increased pollen generated by trees in neglected hedges) and that around 63 per cent of reasonably well-dated late Roman field systems are perpetuated in some way by medieval or later boundaries. “The legacy of Roman Britain is all around us” (p. 154), say the authors, yet of course they are not claiming that two-thirds of our historic fieldscape has Roman roots. The ‘hot’ periods of change were the late Roman (more intensive agriculture, high population, towns, markets and so on) and ‘the long eighth century’: the period once described as the Dark Ages was relatively ‘cold’, more comparable with early Roman times. Perhaps the most unexpected point is the suggestion that in the Central Zone (the ‘champion’ English countryside), the authors’ findings call into question the ‘great replanning’ of mid to late Saxon times. Most readers will not need telling that this survey can only be a rough proxy for an agricultural history of the period, or that it makes no direct contribution to the ongoing debates about how southern Britain became Anglicised. Nor does The fields of Britannia much increase our understanding of coaxiality (although fig. 4.14 provides a helpful five-scenario developmental model).

For this reviewer, the most interesting aspects of this book are the lower case headlines, the undiscussed and things glimpsed briefly. ‘Droveways’ are mentioned (how wide does a thoroughfare have to be before it is a droveway?), and apparently there were seasonally used pastures in the Essex marshlands. There is, however, little mention of transhumance (to hill, fen, marsh and distant wood-pasture), which was surely quite widespread by the early post-Roman centuries. I had not realised how confidently pollen analysts now feel able to ‘identify’ hedged landscapes, at least in southern England; as I understand it, sporadic evidence for hedges goes back to the later Bronze Age, and these now have to be factored into our landscape narratives. The authors make the crucial point (p. 317) that hedges could well have survived the silting-up of their ditches, so that field systems might have continued in use considerably longer than their supposed terminal dates; what with that and the cessation of potsherds scattered in manure, the perpetuation of Romano-British field systems under a pastoral regime would leave little dating evidence.

It is interesting to note how far regional differences in woodland cover go back. As fig. 3.3 demonstrates, the relatively open landscapes of East Anglia and Central England were extant by Roman (and early medieval) times, which rather confirms my long-held belief that eastern England is not necessarily a good reference region for British woodland history. The authors also seem to have identified something of a paradox in south-west England, where, despite the relatively late arrival and incomplete linguistic triumph of English speakers, ‘discontinuity’ seems more marked than in regions farther east! Within these authors’ frame of reference, the durée of coaxial landscapes, and/or coaxial field systems—or at least some of their components—is often longue. In this book's passing references to ‘late prehistoric or Romano-British field systems’, however, we get tantalising glimpses of an even longer durée. Coaxiality was no Romano-British invention (although it may have been a re-invention); we surely now need a ‘Fields of the Pretani’ project to address these issues (and, indeed, continuity across the Iron Age/Roman Britain frontier).

The fields of Britannia seems a little constrained by the research project's terms of reference; it would have been no bad thing if the authors had permitted themselves a more discursive final chapter. Nevertheless, this is a landmark study (no pun intended), clearly documented and explained: the outcome of an excellent and valuable piece of research.