Bryn Mawr Classical Review

Reviewed by Anouk Vermeulen, University of St Andrews (av22@st-andrews.ac.uk)

Version at BMCR home site

The work under review here is based on a thesis with the same title from the Justus Liebig University in Giessen, Germany. Although many aspects of a PhD thesis can still be seen in the monograph, it is also an important publication that contributes to filling a gap in research on Roman Spain, and in particular its rural landscapes. Its objective is to compare surveys and develop a comprehensive, universally applicable methodology, tested in three case studies: the ager Tarraconensis, the Vera basin (hinterland of Roman Baria), and the nearby Upper Almanzora Valley. The approach is strictly statistical and quantifiable, which is admirable but also narrows the interpretative part somewhat. In fact, my main issue with the publication is the lack of more qualitative interpretation, a general lack of extrapolation of the specific data to a larger provincial or even Mediterranean historical context, and a more explicit indication of what the data and spatial analysis means in a larger setting.

The work is divided in seven chapters and seven appendices, which include colour maps, catalogues and statistical tests. The first three chapters form an extended introduction, of which chapter 2 on methodology is most important. This brief and very dense chapter details how exactly the comparison between different surveys can be effectuated, including discussion of site/off-site problems and categorisation of villae and other rural settlement types. In my opinion, this chapter should have been at least twice as long, as none of the aforementioned issues are dealt with satisfyingly. Rather, a very brief overview of the discussion, with somewhat random and incomplete bibliography, is given, before the issues are swept aside in the name of pragmatism. Although an argument can be made for doing so, I would have appreciated a stronger theoretical underpinning and more thorough analysis of the debate, and especially some indications on how this study contributes to advancing the debate. The third chapter is an introduction on the two case studies discussed in Chapter 4, discussing geographical context as well as a brief historical background.

The core of the work is reflected in chapters 4 and 5, where the case studies are analyzed. The hinterlands of Tarraco and of Baria are directly compared in chapter 4, while the Upper Almanzora Valley is presented separately in Chapter 5. Chapter 4 follows a clear structure throughout, discussing each aspect first for Baria, then Tarraco, and then comparing the two. Under the heading 'Diachronic Aspects' follows an analysis of the changes in settlement patterns broken down by century, ranging from 'phase 0,' before 200 BC, to the 7th century AD. Each century is accompanied by a map of decent size with minimal colouring, which enhances clarity, while plenty of tables and diagrams accompany the comparing sections. The maps of Tarraco include the centuriation system, which is unnecessary in my view, since it is not part of the analysis and clutters up the otherwise simple and clear maps. 'Synchronic Aspects' heads a section on size and monumentality of the sites, again divided by century. The maps in this section show sized and colour-coded circles for each settlement, and are mostly too small to be really useful for anything other than illustration. The standard GIS features of slope and aspect then show up in the geographical section of the analysis, as well as soil type, infrastructure, and economic potential (for agriculture, but also vicinity to certain metals or quarries). The chapter concludes with an overview by century of all the changes, again separating both case studies. Although the analysis is thorough and generally informative, there is a missed chance for a deeper and more qualitative interpretation, which could have included details like an awareness of differences in water supply throughout the year. In the ager Tarraconensis, for example, this means that during summer only the Francolí has some water, and even this is unreliable, which has a significant effect on availability of drinkable surface water. Currently, the overview entails a large amount of repetition without pushing the data further and at the very least attempting to offer an explanation.

The fifth chapter on the Upper Almanzora Valley is separated from the two main case studies because it has a much smaller sample size. Considering that the sample sizes in the other two case studies break down to very small, hardly representative numbers as well, I personally think this division is unnecessary, but neither is it very disturbing. Representative numbers and statistically relevant sample sizes are very difficult (and in most cases impossible) to obtain for ancient landscapes, which does not necessarily mean quantitative analysis should be dismissed, and I fully agree with the author's decision to include the smaller sample from the Upper Almanzora.

Conclusions are summarized at the end of each chapter and brought together in a final chapter (Chapter 6), which makes some steps towards integrated interpretation. There is some repetition and I would have liked to see a more qualitative constructive interpretation, but nevertheless the conclusions are solid and consistent with the general scope and objective of the work. The three regions are shown to develop relatively similarly in the pre-Roman and early Roman periods, with large rises in number of settlements mostly in the first century BC, although the ager Tarraconensis is notably richer, with settlements both having higher status and being larger in size. The coastal regions of Tarraco and Baria show a decline in numbers in the second and third c. AD, with initial continuity in Almanzora. While decline in the ager Tarraconensis abates, with it retaining a strong urban and suburban presence, there is slow but consistent decline in Almanzora, with barely any sites left after the fifth c. AD. In the Vera valley the urban centre of Baria disappears in the course of the fifth century, which has a decentralising effect on the settlement pattern. On the other hand, Tarraco remains a strong urban presence, and sites in its immediate hinterland remain continuously occupied. In both areas, however, a decline in site numbers is visible, parallel to the decline in Almanzora. In terms of ecology and environmental factors, the old adage of contrasting situations in plains and mountains is confirmed: the ager Tarraconensis has a plain-related economy, with important viticulture and associated pottery production (something that is surprisingly not extensively mentioned by Schneider), as well as important trade links, while the Vera Basin and the Almanzora valley have a mountain-related economy, with smaller and lower status sites, more elevated sites, and a close connection to the production of metals and mining activities. The settlements in the ager Tarraconensis are also more focused on the urban centre itself, with all high status sites found within 15km of Tarraco.

The main strengths of the work lie in its complete openness of methodology and of data. That makes the study very valuable in relation to recent developments in open access and data-sharing, and I applaud the author for his choices in this. He could have gone even further and published his databases and maps electronically, either on a website or on a disk, to make it even easier for other scholars to continue work on these areas. I for one would have loved to play around with the databases and maps in GIS, to have closer illustration of the data behind the work. I am very much aware, though, that especially for early career researchers, there is a risk involved in alternative publication means and open access publishing, and I suspect this has informed decisions for a more traditional publication format. Permissions from third parties for open access publication may also have been an issue, and might continue to be for a long time. However, change has to start somewhere, and this kind of data-heavy work is perfect for the exploration of online open access publishing. The current volume has been very thought-provoking in that sense as well as for its contents.