The search for Roman landscapes

Landscape archaeology has established itself as a generic term for spatial archaeology. It has its roots in Romanticism and has increasingly found its way into archaeological research since the early 1970s, leading to a variety of concepts and approaches. In most cases, however, landscape remains a passive background for finds and archaeological sites viewed in isolation. This paper therefore conceptualises the terms landscape and landscape archaeology, examines their methods with a focus on archaeological prospection and discusses their contribution to Roman archaeology using two geographically, structurally and historically different examples.
It is argued that landscape consists of a physical and a cognitive component. The landscape existing in physical space is mediated by human perception into the imaginary world of its inhabitants and given meaning. Landscape archaeology is therefore a holistic concept that does not view archaeological sites in spatial or temporal isolation, but places them in the context of their natural and social environment. It describes and interprets the material remains of a landscape in order to obtain an idea of the original cognitive concepts of its inhabitants, which formed the starting point for spatial behaviour. As landscapes are changed and passed on from generation to generation, a diachronic perspective is required.
The archaeological investigation of a landscape requires the use of large-scale archaeological prospection methods. These include field survey, aerial archaeology, geophysical prospection and airborne laser scanning. After a discussion of these methods in comparison with each other and with regard to their individual contribution to landscape archaeology, their contribution to the exploration of various Roman landscapes is outlined.
Firstly, the rural landscape of Halbturn on the Danube Limes is discussed. The starting point for the investigations was a completely excavated Roman cemetery with around 300 burials from the second to fifth century A.D. While the excavation results provided insights into burial habits, demography, social structure and the state of health of the population, a systematic and integrated prospection strategy provided its context. This included the remains of residential and farm buildings, at least three other cemeteries, parcelling, roads and paths and presumably the remains of a limitatio. Together with the excavation results, the history of a community that bred horses and practiced agriculture unfolded. The individual components of the landscape proved to be more closely interlinked than expected, and a complex interaction between the graves and the agricultural landscape was revealed.
In a second case study, the Roman harbour town of Osor on the largest Croatian island of Cres is presented. The choice of location appears to have been based on a strategic advantage and control of the narrow passage between the islands of Cres and Lošinj. However, while the artefacts recovered in excavations point to an extensive trade network, the underwater topography provides little evidence of suitable harbours for large trading or military vessels. This contradiction arises from the difficulty of interpreting the historical sources for Osor, but also from the fact that conventional archaeological excavations alone provide a great deal of detailed information, but do not reveal any spatial relationships. Here, detailed terrain models from airborne laser scanning and airborne laser bathymetry have expanded our knowledge of the hinterland. They show a very dynamic landscape with dry stone walls that contain a wealth of information about a complex sequence of human activities, including a possible Roman limitatio. This raises the question of the forms of land ownership or the extent of Roman land surveying.