Although many of the best-known archaeological sites are ancient cities and urbanism was a central feature of the Roman Empire, the majority of its occupants lived and worked in the countryside, or at the very least, owned property there (Kehoe, this volume). The spaces between cities variously housed farmland, country retreats, industrial complexes, sanctuaries, cemeteries, military zones, hunting terrain, and wilderness. Villas, forts, roads, bridges, aqueducts, and other visible or monumental structures in this landscape have been studied for some time through excavation and structural surveys (Percival 1976; Sitwell 1981; O’Connor 1993; Hodge 2002). Humbler sites and the landscape itself have not received as much attention, but farming and other kinds of rural production leave structural and environmental traces (MacKinnon 2004; for economic issues, see Mattingly, this volume).
Field survey, the study of surface artifacts and other remains of occupation over a broad area, has been one of the most important developments of the latter half of the twentieth century in Roman archaeology. Unlike excavation of a single site with its particular vicissitudes, field survey offers the opportunity to look at a broad region over a long time-span and evaluate (among other things) changes in types or locations of sites, agricultural usage, military occupation, or trade patterns over time (for the archaeology of production and trade, see Mattingly, this volume). The territory of an ancient city or a geographical feature such as a river valley may serve as the unit of investigation, often approached through a sampling strategy. Methods employed in survey have become increasingly systematic and multidisciplinary (Fran-covich et al. 2000). Survey has the advantage of being non-destructive and noninvasive. Also, it is usually less expensive than excavation.
The UNESCO Libyan Valleys survey investigated the pre-desert of Libya, where substantial farms and olive presses attested considerable productivity in ancient times in this highly arid zone that receives 35-75 millimeters of rain per year (Barker et al. 1996; D. J. Mattingly 1994). That the ancient climate was the same as today’s is shown by the preserved seeds and pollens of naturally occurring species (weeds), which were all suited to this climatic regime, whereas the cultivated species were all ones that required considerably higher rainfall. Mapping sites, structural remains, and sherd scatters, archaeologists demonstrated how complex systems of retaining walls in the seasonal creek-beds (wadis) channeled runoff from the brief but intense rains into agricultural areas, where it sustained substantial production of olive oil, exported to the coastal region and beyond. The local population, so identified by (among other things) Libyan names on inscriptions, developed this system of floodwater farming under the pressure of Roman taxation. The gradual abandonment of this system of agriculture came about as part of a regional economic decline rather than through environmental change or warfare.
While the need for surplus to pay taxes stimulated agricultural intensification in Libya (and elsewhere), annexation into the Roman Empire seems to have had the opposite effect on Greece, the Roman province of Achaea (Alcock 1993). Here numerous different field surveys show both a drop in overall numbers of sites and the loss of the smallest sites in the countryside during the late Hellenistic and early Roman periods. Where it is possible to judge, it further appears that there was instability in occupation of sites, with long-occupied sites going out of use, and altogether new ones being developed. Contemporary literary sources, scant as they are on such a non-elite topic, create an image of instability followed by rural depopulation in Roman times, and modern scholars have sometimes taken this as an indication of extensive population decline in Greece in the Roman period. However, along with the disappearance of the little sites that archaeologists usually associate with independent, small-scale farmers comes an increase overall in the size of occupied sites and the appearance of some very wealthy sites (usually referred to as villas) in the landscape. These factors taken together indicate substantial change in the system of land ownership, with fewer, larger estates replacing a wider scatter of smaller farms. Rather than disappearing entirely, it is possible that some small farmers resided in nucleated settlements instead of on their property. In a few instances, surveys show that early imperial sites are clustered around urban centers, while more remote areas cease to have sites. Moreover, the continued prosperity and even expansion of Greek towns indicates a steady and substantial urban population base. Thus, archaeological data demonstrate that the apparent emptiness of the Greek landscape reflects not a straightforward and drastic population decline, but considerable change in patterns of land tenure, probably also coupled with some demographic attrition.
World History
