Bryn Mawr Classical Review 2013.01.52
Gijs Willem Tol, A Fragmented History: a Methodological and Artefactual Approach to the Study of Ancient Settlement in the Territories of Satricum and Antium. Groningen archaeological studies, 18. Eelde; Groningen: Barkhuis; Groningen University Library, 2012. Pp. xvi, 405. ISBN 9789491431036. $87.00 (pb).
Reviewed by Christopher Smith, University of St Andrews/British School at Rome (firstname.lastname@example.org)
This comprehensive and important contribution to the study of southern Latium emerges from the long-term work of the Groningen Institute of Archaeology around Antium and Satricum (Borgo le Ferriere). It explicitly addresses problems with survey archaeology and therefore reflects the intense methodological approach which has characterised both Dutch and British work on field survey over the past twenty years.1
Tol reconsiders the Astura and Nettuno surveys, both already published, in part by Tol himself, who is therefore extremely well-placed to take this extra look at the data.2 The work begins with a methodological introduction which considers the way in which survey is conducted and how its results are extrapolated for the study of socio- economic factors. Tol notes that survey has problems with site recovery, site typology and site chronology, well- known difficulties for the finer grained analysis one might wish to conduct. This relates particularly to socio- economic issues.
Chapter 1 considers the historical and archaeological evidence already available. Antium and Satricum have been looked at in detail; Satricum has been the object of decades of excellent Dutch research. Mid-Republican evidence for Antium is sufficient to support the idea of a Roman colony on the site, as the sources claim; villas were constructed along the coast in the later Republic, and there was a building boom in the first two centuries AD with strong imperial support. Satricum is better known for its archaic phase, with houses, a temple and evidence of a necropolis; the Republican and imperial period are less well represented. Tol then considers the attestations for the oppidum of Caenon and an emporion and road station at Astura, the archaeological evidence for coastal villas and rural settlement and the tomb known as the Torre de Monumento. Roads must have existed but are underrepresented in the archaeological evidence.
An appendix to this chapter publishes what survives of a third-century BC votive deposit from near Anzio; the remains in the antiquarium at Nettuno consisted only of the black glaze ware fragments. The material appears to have been recovered from caves under the Villa Sarsina in the 19th century.
Chapter 2 returns to methodology. The problems inherent in dating by pottery in surface scatters remain; the date ranges are uneven, and the presentation in histograms gives false impressions. Site function is often attributed vaguely and on the basis of loose interpretation of literary evidence; site development is often obscured both by the material and by its presentation. In order to counteract some of these issues, return visits to survey areas were coupled with two chronological devices: first, the reconstruction of the chronology of each site using the full date range of any given ceramic find; and second, the division of periods into equal blocks of 150 years.
Revisiting sites increased the number of sites and improved the amount of data but did not substantially change the general settlement trends. Chronologies were confirmed or extended by the new material.
A second methodological case study came from studying the largely unpublished material in the storerooms of the antiquarium at Nettuno, mostly collected in the 1970s and 1980s. The generally high quality of finds permitted the evidence for the status and function of sites to be improved, and also increased information for the Republican and especially the later periods. It is unsurprising that the material was on the whole more decorative and aesthetically interesting.
A third approach to the data was through four new intensive surveys of sites in the Nettuno area previously identified in the larger survey. Two or three walkers worked on four by four metre grids, collecting all available material, which was then classified in the field and weighed. Some metal detecting was also used. The data were then used to produce distributions by number and by weight of fragments, and also by date. One site continues to be considered probably a farmstead; another originally identified as an outbuilding may be a roadside shrine, with painted walls and exotic marble in the floor; a third has become from study a villa rustica more complex than previously thought. More significant changes came with regard to lengthened chronologies of use.
Fourthly, an area on the river Astura was excavated. It was predominantly active between the fourth and sixth centuries AD, with local and imported pottery, coins, lamps, glass, bone, metal and animal remains, and there is some evidence for pottery production of a ceramic painted with red bands. Its earlier identification by Piccareta as the road station mentioned in the Tabula Peutingeriana now seems certain, and confirms a general pattern of decreased importance of urban settlement in favour of road stations from Campania north to Etruria in the late imperial period.3 A revival in the 12th and 13th centuries is also attested.
The work concludes with a diachronic discussion arguing for a vigorous archaic period, especially at Satricum; decline in the fifth century (the period of Volscian incursions); rural infill in the middle Republic continuing into the late Republic, with increasing evidence of international trade; a flourishing period at Antium in the early and middle empire; decline in the later empire, with the rise of the Astura vicus; and some early medieval evidence. The work has led to a greater variety of sites, and chronological decisions taken have stretched the dates of some sites, but site development remains elusive.
Regarding socio-economic matters, on the basis of the new material Tol makes the case for the contribution of locally produced wares to exchange systems. Fluctuations in pottery consumption might be seen as proxies for economic change – and the consequence of the data is to flatten out abrupt change, and reduce evidence for decline. Tol is coy about where the data lead regarding high versus low counting in terms of demography, but the data do indicate that many of the sites now appearing were not used for habitation. Town-country relations are considered; some material seems unlikely to have been exchanged through urban markets, but imported pottery tells a different story.
Tol’s conclusion is that the intensity of this study has revealed valuable results, and is therefore the way forward. He writes, ‘I am convinced that large-scale, consistent and intensive sampling provides us with the best opportunity to study and identify similarities and differences in economic exchange and consumption’ (p. 386); and this conviction has been rewarded with a major new grant to study ‘minor central places.’
This is an excellent piece of work; mature and assured which fits with the long series of outstanding Dutch work, led by Peter Attema. Students of methodology and those interested in central Italy will benefit from the careful presentation of the results (it is characterised by substantial illustrations of local coarseware) and the intriguing results.4 It is instructive to read it alongside the other doctoral thesis produced at the same time by the project, by T. C. A. de Haas, who focuses on Roman colonization and uses a off-site survey data.5
Tol makes a determined argument for the methodological value of his survey approach, and this will need more discussion. At the moment, the tendency towards restudy is strong. In part this is because of a clear recognition of the limits of broader surveys, and those arguments are well made. But is there a sense in which restudy, and intensive survey, simply compound the problems? The core problem has always been that pottery is a proxy for human activity. Recovery rates are poor, but deposition rates are utterly unpredictable. The uncertainty between whether we are seeing half a dozen pots being dropped over fifty years or continuous human presence over one hundred and fifty years on a small site can be overcome to some extent in individual cases by contextual data and good sense. Can it be managed better by intensive survey? Methodologically, it might be argued that Tol has replaced artificially sharp changes with artificially smoothed transitions. What we are still looking at is an archaeological record which relates to, but does not mirror, human behaviour. As a consequence, in any individual site, there may have been any number of variations in continuity versus discontinuity, but the archaeology is not fine grained enough to tell.
T. himself points out that site development has been resistant to his methods. This is no criticism of his work; site development is revealed even in the best preserved sites like Ostia and Pompeii only by the most painstaking work. However, there is an open question about how well-preserved a building has to be to entertain revisiting. We need to be very open to the possibility of a landscape of rather rickety buildings which were occasionally visited in the context of a world of regular rapid crossings and individual transactions, rather than one of formal building and complete abandonment.
Tol must be right that movement and exchange at the smallest scale must be both responsive to and productive of economic activity at a larger level; there must have been some multiplication of individual acts to broader trends. The further work of the Dutch teams will be very welcome in trying to get from site to trend. However, scaling up is difficult from partial evidence. There is an inevitable danger that we will have exactly what we had before; incompatible data for highly localised phenomena telling us about the vital variability of the ancient world. Boxing things into similar-sized chronological containers will at best conceal that; it will not create comparability.6 Where Tol makes a very telling point, however, is that local pottery production and exchange may have been underplayed in modern scholarship. This is where one might well start to look for much more systematic petrographic analysis of local wares.
In short, this is a valuable contribution to the debate over survey and its future. It could be argued that it puts too much faith in the capacity of ever-increasing levels of detail to reveal information at an appropriate level of return to effort, and there are of course other methodologies which might be deployed; geophysics, aerial photography and LIDAR are just some. However, the results of this specific reinterpretation of the data in southern Latium continue to add to our knowledge of a fascinating area.
1. For a recent volume of essays on survey methodologies see F. Vermeulen, G-J. Burgers, S. Keay and C. Corsi (eds.) Urban Landscape Survey in Italy and the Mediterranean (Oxford, 2012).
2. P. A. J. Attema, H. Feiken, T. C. A. de Haas, G. W. Tol, ‘The Astura and Nettuno surveys of the Pontine Region Project (2003-2005). First Report,’ Palaeohistoria 49/50, 415-516, and P. A. J. Attema, T.C.A. de Haas, G. W. Tol, ‘The Astura and Nettuno surveys of the Pontine Region Project (2003-2005). Second Report,’ Palaeohistoria 51/52, 169-327.
3. F. Piccarretta, Astura (Florence, 1977), in the Forma Italiae series.
4. For a summary of the overall project’s work in the Pontine region, see P. Attema, T. de Haas, G. Tol, Between Satrium and Antium: Settlement Dynamics in a Coastal Landscape in Latium Vetus (BABESCH Supplement 18) (Leuven, 2011).
5. T. C. A. de Haas, Fields, Farms and Colonists: intensive Field Survey and Early Roman Colonization in the Pontine Region, Central Italy (Groningen, 2011).
6. For a highly pertinent attempt to scale up survey data to produce higher level analysis, see A. Launaro, Peasants and Slaves: The Rural Population of Roman Italy (200 BC to AD 100) (Cambridge, 2011), who compares twenty-seven archaeological surveys – but not the Pontine Regional Project, which nevertheless now seems to cohere with his other data.