These two recent works on the Roman military in northern Britain highlight both the positive and negative aspects of the BAR model of publishing. On the one hand, it is great to see interesting dissertations and research projects made available to a wider audience. On the other, a firmer editorial and review process often seems to be needed, and one of the volumes considered here exemplifies this issue quite clearly. The stronger of the two books is Evers' study of the evidence for characterising the ancient economy that can be drawn from the Vindolanda tablets. Though covering somewhat well-worn territory, this useful summary of key evidence also contributes a nuanced approach to the economy of the frontier zone that chimes well with current debates and new archaeological results. Giles' volume is based on a promising proposal to investigate site formation and behaviour in a range of military installations, but unfortunately must be regarded as a missed opportunity, hampered by methodological and theoretical flaws that could have been readily corrected by more thorough review.

E.'s volume aims to evaluate the long-standing debate on the nature of the Roman economy in light of details drawn from the Vindolanda corpus, and a small number of other comparable documentary sources. E.'s hypothesis is that the traditional dichotomies which characterise the imperial economy as ‘ancient’ or ‘modern’ are too simplistic, and that in reality there was a mixture of mechanisms operating at the same time, and this seems eminently sensible, if somewhat unsurprising. The first couple of chapters outline this debate and previous work on the Romano-British economy. These are reasonable syntheses, albeit lacking up-to-date references to either work on the military community in the frontier zone or the application of theories of globalisation to the Roman Empire, both of which are highly relevant themes. Chs 3–5 contain the core analyses of specific tablets, in relation to three main modes of economic activity: market exchange, redistribution and reciprocity. In each chapter these concepts are introduced, and then three main case-studies are chosen from the Vindolanda corpus, supplemented by miscellaneous material from this site and others such as Carlisle. Ch. 6 pulls together all of this information to argue for a tripartite model of the economy, operating also on three nested spatial scales of local, intra-provincial and inter-provincial interactions. The parallel existence of market exchange to procure provisions, state redistribution through a range of channels, and the use of goods in reciprocal fulfilment of religious and social obligations is well argued. The implications of this complex web of activity for broader debates about the nature of the Empire are also pursued, with some interesting points made about the nature of the frontier economy which tally with recent archaeological work on rural settlement on both sides of Hadrian's Wall. Though this concluding discussion is brief, perhaps reflecting the difficulty of generalising too far from the tightly-focused context of the Vindolanda tablets, it completes a useful volume.

If E.'s study addresses primarily the networks that existed around forts in the northern frontier zone, Giles' research very much concerns what went on inside them. The project described is an attempt to conduct a distributional analysis of finds within a selection of northern forts, deploying also theoretical insights drawn from New/processual archaeology, particularly aspects of the work of Michael Schiffer. At first glance this seems a curiously retrograde step, given that most of the openly theoretical work in Roman archaeology has drawn from post-processual sources of inspiration. However, it is true that given the timing of the development of explicitly ‘theoretical Roman archaeology’ (emerging at the turn of the 1990s), quite a lot of the insights that the New Archaeology did produce were skipped over, or only incorporated at an implicit level in methodological innovations. Certainly, the study of site formation process is a key aspect of later New Archaeological, or processual, work that remains a glaring omission in Roman archaeology as a whole. The approach taken by G. can thus be clearly justified and indeed holds considerable promise.

However, a range of problems in the execution of the research limit its contribution. Concern is first aroused in the introductory chapters. The review of the historical background to the Roman frontier (ch. 2) is entirely conventional, displaying none of the New Archaeologists' interest in taking a critical approach to received historical wisdom. Ch. 3, on ‘Interpretation and Theory’, is more a history of research in the Wall region, with only patchy coverage of theoretical developments, particularly in recent years. The following chapter on ‘Research Methodology’ is similarly inconsistent, with gaps in content (crucial issues like residuality and quantification do not receive satisfactory coverage) and bibliography. Indeed, for a volume published in 2012, the bibliography is problematic; although a handful of more up-to-date references have been added, there are numerous omissions among works on the history of Roman military archaeology (e.g. by James), artefactual and architectural studies of military sites (e.g. by Allason-Jones, Allison, van Driel-Murray, Gardner, Hodgson and Bidwell), and previous attempts to apply Schiffer's approach to Roman material (e.g. Peña). Nor is there full enough discussion of the ongoing development of that approach in the theoretical literature. While a study of this length might be forgiven some such shortcomings, in this case the gaps really hamper the author's engagement with the material.

This material is derived from hugely variable excavations at some 17 sites, with much of that variation being compensated for by aggregating the finds distribution data into ‘composite forts’, in order to discuss chronological and functional patterning in the core results chapters of the volume. Some interesting insights are produced from the analyses in these chapters, particularly in relation to the functional patterns, but core issues like residuality and context formation remain peripheral when they should be central, and thus we really get no closer to a comprehensive view of formation processes on a Roman military site. An illustration of the problem is the ascription (49) of a higher-than-usual percentage of ‘military’ finds in the Hadrianic period to warfare at this time, rather than to the particular depositional circumstances of primary construction of many forts. Ironically, such recourse to simplistic assumptions to explain aggregate patterns was precisely the kind of practice which drew criticism in the early days of the New Archaeology, leading on to the detailed work of Schiffer and Binford on middle-range processes, and ultimately to the contextual archaeology of Hodder and others. The volume concludes with a discussion of evidence for gender and status deriving from the analyses, which does make some useful points but is again hampered by lack of reference to key work on, for example, the gendering of artefacts in Roman military contexts.

Whereas E.'s solid study of the ancient economy has emerged into published form pretty satisfactorily, G.'s more unusual and ambitious project really needed more robust editorial intervention to update the bibliography and consequently the discussion of a number of key issues. It does take us forward, though, at least in so far as it highlights a major theme of both theoretical and methodological interest that Roman archaeology has yet to adequately address.