This book offers a welcome analysis of urbanization mechanisms and city network(s) in the Roman Balkan and Lower Danube provinces. Donev's (D.'s) research was carried out within the framework of the European Research Council project “An Empire of 2000 cities: urban networks and economic integration in the Roman Empire”Footnote ¹ and brings forth a large-scale analysis of an area not often studied holistically. Alongside the book's intrinsic value of surveying a large area panoramically, it proposes an interdisciplinary methodological framework that advances the study of settlement patterns and their emergence in the Roman Empire.

Chapter 1 defines key concepts and presents the author's methods from economic geography, such as rank-size analyses. The theoretical excursus helps frame the book and points the reader toward its focus, which is not economic performance in itself (as one might anticipate), but the nature of economic relations implied by the variable distribution and size of urban settlements. The research spans the High Empire, from the time of the Principate, and deals specifically with the provinces of Upper Macedonia, Epirus, Dalmatia, the Pannonias, the Moesias, Thrace, and Dacia. The approach is ambitious: the area is large and complex, with many fundamental differences not only interprovincially but also within the provinces themselves. D. also emphasizes a shift from the older studies of individual settlements toward a focus on urbanization of larger areas, situating his book as part of a growing trend.

The data set is well defined, and D.'s choices are explained in detail: rather than using population size thresholds to attribute urban status (and thus risk leaving out settlements that may reflect a critical part of Roman life), the author also includes town-like secondary agglomerations. A potentially larger-than-necessary sample is preferred to one that may prove overly restrictive (6).

The idea of imperial agency is only touched upon, with D. stating that most settlements can be considered part of a minimalist strategy that focused mainly on control of key traffic nodes, natural resources, etc. (9). While this may be generally correct, it perhaps underestimates cases of imperial agency that are not part of large-scale imperial urbanization processes but reflect local imperial priorities (for example, the propaganda behind the founding and naming of Ulpia Traiana Sarmizegetusa, the capital of the Roman province of Dacia).Footnote ²

The author briefly explains his choice of study area, highlighting personal reasons for it (access to literature) and also acknowledging that the region likely never “form[ed] a compact territorial block” (4) in Roman times. This raises an important methodological issue. The chapter would have benefited, in my opinion, from a discussion on the choice of the term “periphery,”Footnote ³ its definitions from both a sociological and a geographical point of view in the context of the Roman Empire, and its meaning for the author. These matters are not dealt with in depth. The Balkan and Danube provinces can be considered geographically peripheral, as most of them are parts of the limes – but in this line of thought, very few European provinces are not. They could also be considered culturally peripheral, for example, when comparing the Greek part of the Empire to the Latin one, or the Pannonias or Dacia to the more “cultured” extreme west (Gallia, maybe?). In fact, could all provinces be considered a periphery of Italy? All these hypotheses hold true to a point, but none is axiomatic; hence a deeper terminological dive from the author would have been extremely welcome.

The genesis of the settlement network in the research area is presented in Chapter 2. In his search for continuity, the author encounters far more “green-field settlements” – that is, settlements built from scratch in a previously empty location rather than evolving from previously existing settlements – than pre-Roman ones. The number of urban settlements founded after the conquest is, in all the discussed provinces, very high; overall, fewer than 30% were pre-Roman (41). Noteworthy is the fact that most of the Roman settlement network remained active up to the High Empire, while the same cannot be said about the pre-Roman one. The sharp discontinuity between the pre-Roman and Roman urban configurations characterizes the whole area, although the reasons for the contrast vary. There are certain parts, such as Upper Macedonia and the Adriatic coast, where the existing Hellenistic settlement network continued to meet most of the Empire's needs, and therefore it persisted. In order to make the data more easily available and to reinforce the conclusions, this chapter includes graphs presenting the pre- and post-conquest situation of each province, tables with all settlements analyzed, and maps highlighting the chronology of regions. This type of visualization, which continues throughout the book, helps the reader better grasp some of the area's particularities

Discrepancies between two regions of the same province were sometimes major. In Dalmatia, for example, where the Adriatic coast and the inland territory presented two completely different environments, the Empire faced challenges in urbanizing the inland. The differences persisted, and in fact a slight decrease in mainland urbanized sites can be observed when compared to the Hellenistic period.Footnote ⁴ The reality behind this phenomenon, not only in Dalmatia but also in the Pannonian provinces, is to be identified, at least partially, in the fact that the Romans brought an end to the tribal fortification system and its small networks. The High Empire's settlement network never incorporated the La Tène oppida (26).Footnote ⁵ Unlike in certain areas of Epirus and Macedonia, however, the army was not directly responsible for the demise of existing settlements; instead, they “simply failed to integrate into the new socio-economic reality caused by the Roman conquest” (27). The fate of oppida and strongholds, significantly more sensitive to political change than commercial emporia, had a profound effect on the urban network. In Dacia, the scene of the most brutal conquest in the area, the change in settlement patterns is the same as in the rest of the Danube provinces, but intensified. The pace of urbanization in the province was “mind-boggling” (38), based almost exclusively on Roman-founded settlements. Regardless of local specificities, urbanization apparently had a rather uniform pattern across the whole area. This idea is reinforced by the analysis in the following parts of the book and supports the author's choice of this area as a coherent case study.

The next chapter (Chapter 3) focuses on the myriad new foundations. The analyses are not limited to the autonomous towns, as this would have led to losing a considerable segment of the integral network. Besides this obvious category, the author also includes military settlements that in time evolved into urban-like agglomerations, as well as settlements that evolved around special-purpose facilities (for example, road stations, ports, hot springs, mines, quarries). Moreover, D. considers that larger rural settlements are centers of micro-regions. This presumption is in most cases epigraphy-based, with written sources setting these places apart from average villages. Unfortunately, they are generally poorly researched archaeologically, which makes the data appear biased and the conclusions disputable, as the author justly remarks. D. states that his primary purpose is not to draw the “elusive line” between urban and rural, but to study the distribution of urban functions within different formal/administrative and quantitative categories (52–53). As one can see, the data selection process is quite laborious. It requires a large quantity of information, and some data is unavailable or difficult to pinpoint in scarce publications. The results are usable for various types of research at the level of case studies; the slight shift in methodological approach also seems useful, offering nuance to the perception of ancient realities. The maps throughout the chapter highlight settlement clusters and the gaps between them, enhancing the picture of an urban network that incorporates old and new settlements, and military and civilian ones.

The next three chapters form the core of the book, dealing with the general size-distribution of the settlements and the agricultural and administrative territories of the urban sites, assessing them from a quantitative perspective and employing geographical methodologies. Chapter 4 opens with an inquiry into what exactly is reflected in settlement size: population numbers and degree of urbanization, of course, but at the scale of a province, also perhaps internal settlement hierarchy. This last feature is the chapter's true focus, and the author seeks explanations for why certain size categories cluster in certain areas.

Methodologically, for each town, upper and lower estimates of the size of the territorium are taken into consideration, resulting in maximum-minimum curves for the rank-size distribution (92). The explanations of the graphs are comprehensive – and necessary, as they reflect the interdisciplinary aspect of the work, as well as clarifying notions that are not familiar to all historians. In economic geography, city-size frequencies that follow a log-normal distribution are associated with advanced economies. The rank-size graphs indicate that Roman Macedonia, for example, did not achieve a very high level of economic integration; this is reflected in the historical record, as the trade ban imposed after Pydna had medium- and long-term implications for the area (95). Thus, sometimes we get confirmation of what we already know, along with testing the applied methodologies. At other times, the new tools open up fresh perspectives. Such is the case of Dalmatia, for which the concave shape of the rank-size graphs has been attributed, generally, to the excessive growth of average towns. There must also be another cause in this case, the author argues, namely the failure of the top settlement (Salona)Footnote ⁶ to reach the size projected by the rank-size rule. The explanation must be sought in Salona being a port, where incoming wealth was only partially retained. In addition, the city was part of a different economic network, connected to the exterior, the Adriatic shore. These external connections and relationships shaped it more profoundly than the provincial mainland network did (97). In the limes provinces, another phenomenon is visible: large towns did not generally act as emporia but instead owed their development to primarily non-economic factors, such as a military presence, or their role in fulfilling the administrative needs of the provinces. At the same time, the second quantitative tier was much better developed than elsewhere, probably again due to the military. Interesting in this analysis is the case of Thrace.Footnote ⁷ It is demilitarized, but the graph pattern is very similar to the other provinces; the divides in this province, though having different socioeconomic implications than in the limes provinces, align to the same rank-size pattern, once again suggesting a certain degree of uniformity throughout the area.

Chapter 5, on the agricultural territories of the settlements, opens with the research questions and assumptions that motivate the author's analytical approach. D. starts from the hypothesis that Roman towns functioned like agro-towns and settles on an optimal 5-km radius for agriculture (123). As in the previous chapters, he introduces each province individually, beginning with a description of the physical geography. We thereby see the differences between areas take shape. Thus, we have small agricultural settlements with well-chosen micro-locations within an area of generally limited agricultural potential, such as the agro-towns in Northern Epirus and Macedonia, or the many villas on the Dalmatian coast. We must note the short distance between cities and the relative scarceness of secondary agglomerations. A change of geography, however, dictates a change of dwelling-agglomeration patterns as well, and in the Pannonias we have settlements with a 15-km market radius (156).

One of the ideas that recurs throughout the chapter is that, notwithstanding the importance of the non-agrarian sector in these towns, agricultural considerations and connectivity ranked high in the siting of all civilian settlements. Major settlements developed out of other considerations – political, administrative, or commercial – but medium-size settlements dominated the agricultural hinterland. The highly urbanized belt of these provinces only partially coincided with the agricultural heartland of the territory, thus exposing the dichotomy that drove the evolution of the settlement networks. While some geographical features were paramount in the development of the urban network (mountain zones), others (fertile valleys) do not fall into the pattern we might anticipate (173).

The last criterion for settlement analysis is administrative status (Chapter 6). In order to make his undertaking intelligible, the author provides separate lists of autonomous towns and state-run districts. The enterprise employs Thiessen polygons (183).Footnote ⁸ This method is borrowed from geography, the polygons being used to allocate space to the nearest point feature. They define an area around a point, where every location inside the polygon is nearer to this point than to all the others. The resulting maps present an idealized potential administrative territory of the cities. Of course, in our case, they ought to be correlated with epigraphy and not taken as definitive; while D. does this when the written sources are rich, he remains hesitant in the case of smaller settlements and ambivalent sources. Despite the chapter being at times hard to follow for the historian untrained in geo-spatial technologies, the results are synthetized in a comprehensible manner, so that the important points become obvious.

The size of the administrative territories varies, but some general rules become apparent. In Dalmatia, the extensive urban coverage is due to obscure settlements on the periphery of the network. Medium-size territories are frequent in the Balkans, while the militarized Pannonia Superior is dominated by small hinterlands. The main transport route is on the limes, while the geometrical (and agricultural) center of these provinces is left out; implicitly, the Pannonias are oriented toward the north. The largest towns do not have the largest hinterlands (210), but oversized settlements with disproportionately large agricultural territory (such as the case of Dacia's capital) appear sometimes as the answer to a very mundane need of some important cities, namely a more consistent tax pool. An interesting connection is made with Late Antiquity, highlighting the fact that dioceses developed in certain urban centers, even if they were not the area's largest cities during the Principate. The area showcases an ambiguous relationship between size and regional influence (256), which needs to be interpreted in each case separately, taking local political, military, and commercial factors into account.

A welcome addition is the final catalog (322–79), comprising records of all “towns and settlements, built-up area, agricultural and administrative territory” in the author's data set: approximately 350 entries, each with a short historical presentation and bibliography. Along with the synthetic tables in every chapter, they are one of the book's main assets, contributing to its utility and ease of use.

The conclusions stress both differences and common trends in patterns of settlement development. Interestingly, one of the revelations about the area is that the high percentage of urban coverage does not necessarily imply a high degree of urbanization (224). The author finally identifies the study region as a periphery, arguing in this final chapter (Chapter 7) that it shares enough common features with other such regions to be included among them, even if the reasons for the development of those features were local. A more detailed comparison with other areas of the Empire would have been welcome, although the necessary data sets and analyses may not have been available.

All in all, the book is a useful addition to the historiographical landscape not only of the Danube and Balkan provinces but of the Roman Empire overall,Footnote ⁹ dealing with a fragmented and heterogeneous territory in a holistic manner. The thorough analysis, the interdisciplinary methods, and the synthetic graphs, tables and maps all contribute to the book's value for future research.