In this volume, Allison offers a valuable study of Roman archaeology that admirably demonstrates how far the discipline has developed in the past 30 years and also provides a model for future practice. A.'s experience of spatial analysis of artefact assemblages as a means of providing insight into social practice has been applied to Roman frontier studies, which is data-rich in terms of the number of sites and the amount of artefacts available from them. Despite the temptations that such a wealth of data provide, A. presents a very clearly defined study, geographically and chronologically focused on a legionary fortress and other fort sites of the first and second centuries a.d. on the Rhine and Upper Danube frontiers.

The monograph is structured in a logical fashion, broadly consisting of four parts separated into thirteen chapters and seven appendices: introduction; methodology; results; discussion and conclusions. Ch. 1 provides an introduction to the study, while chs 2 and 3 provide discrete discussions of Roman military communities and artefacts as they relate to social practice, respectively. Ch. 4 explains site selection and data processing, while ch. 5 expounds on how artefacts were related to specific activities. Chs 6–10 are each dedicated to a site: Vetera I, Rottweil, Oberstimm, Hesselbach and Ellingen, respectively. Ch. 11 provides inter-site analysis, with ch. 12 focusing thematically on status and gender. Ch. 13 concludes the book. The chapters are amply supported by the appendices, which expand upon methodology and results. Significant further resources accompanying the book are the interactive maps and downloads available via ADS (Archaeological Data Service: http://archaeologydataservice.ac.uk/archives/view/milspace_leic_2008/). The volume has been produced to a high standard, though initially the text seems remarkably under-supported by illustrations. The majority of illustrations can be found in the appendices, and interactive maps for each site in the study are archived with ADS and referred to in the text. These latter punctuate in vibrant colour what is only offered in greyscale by CUP.

While the structure of the volume and index facilitate a reader's ability to simply dip into the text for a quick grab of information, every reader — especially students — needs to be fully aware of the methodology supporting the study. A. is explicit, both in terms of site selection and in the use of artefacts as a means of exploring social activities. The sites were chosen because they were only briefly occupied, reducing issues of residuality and redeposition. Ch. 5 also provides the essential rationale for A.'s attributions of ‘activity categories’ and ‘gender categories’ to specific types and classes of artefact. Significant here is that A. provides a basic confidence rating to distinguish between definite/probable and possible attribution and use of an object, and this is carried throughout the text and illustrations.

There is much to applaud in this study, including the amount of time and commitment required for the compiling of the data. Inevitably, critics will point to the lack of detailed discussion of taphonomic issues relating to artefact recovery and context; yet A. provides an overview of excavation history and stratigraphy for each site, and contentious or insecure contexts are acknowledged appropriately during interpretation. Another criticism might be the prominent focus on gender (and age) as identity constructs, paying little to no attention to other identity categories, such as ethnicity, or military-specific affiliations (like legionary or auxiliary). However, while A. is conscientious in not over-playing status associations linked to particular spaces (such as officers’ quarters), the data were also examined relative to status as well as gender.

In summary, A. is to be congratulated for a stimulating study and promising method. Her results add to the growing body of data that indicate Roman military settlements were more complicated and nuanced than presumed by previous generations of scholars. This is no mean feat, though a greater challenge remains: can we surpass the hurdles presented by regional and chronological variation in Roman artefacts and typologies, not to mention diverse international standards of excavation and publication, to build upon A.'s foundation?