Introduction

The Vidourle River runs south from the Cevennes hills for around 95km before debouching into a lagoon to the east of Montpellier on the Mediterranean coast of southern France. It traverses a range of different geographical units—hills, foothills, basins, coastal plain, lagoon, garrigue and scrub—draining a catchment area of 800km² (Figures 1 & 2).

Figure 1.

Location of the Vidourle valley (© M. Scrinzi).

Figure 2.

Research situation in the study area, cities and different landscapes (© S. Sanz & M. Scrinzi).

Various archaeological projects carried out over the past fifty years throughout the valley have identified 832 settlements dated between the seventh century BC and the tenth century AD. This significant dataset is the basis for analysing population dynamics throughout the valley across seventeen centuries (Reference ScrinziScrinzi 2014). In what ways did settlement evolve between the Early Iron Age and AD 1000? Are there differences in the human occupation and exploitation of the upper and lower Vidourle valley? And does the river influence the choice of settlement location? All of these questions fall into a more general theme concerning the spatial and temporal analysis of settlement, in accordance with the Archaeomedes programme that forms our methodological model (Reference Van Der Leeuw, Favory and FichesVan der Leeuw et al. 2003).

Results

The relationship between coast and hinterland, geographically different but linked by the river, is at the core of our analysis. For each major chronological phase, our research highlights similarities in settlement forms across the entire valley; from the concentration of populations within protohistoric oppida and medieval castra and villae, through to the networks of cities and dispersed settlements of classical antiquity, all areas of the watershed share the same settlement distribution and dynamic. Beneath this apparent uniformity, however, there are significant differences in the population density of the coastal plain and inland areas, the latter being much lower. Why? A first response to this question concerns the Mediterranean Sea and its possibilities for commercial and cultural exchange; another response focuses on the natural resources and favourable soils that influenced the development of agricultural and crafting activities. Even if full understanding of settlement in the high valley zone is hindered by vegetation that makes these areas difficult to survey, the levels of settlement remain, in all cases, well below those documented at lower elevations. Indeed, the scarcity of settlement dispersal in the cultivated areas of the high valley indicates limited investment in the hinterland; here, population appears to have remained largely concentrated.

Our settlement database allows us to establish the number of sites occupied in each century, providing a reasonably accurate assessment of settlement trends (Figure 3). The results fit well with knowledge of the wider settlement context of southern France identified in various comparative studies (Reference TrémentTrément 1999; Reference Van Der Leeuw, Favory and FichesVan der Leeuw et al. 2003; Reference VialVial 2011). Each major chronological phase demonstrates a unique picture: population mainly grouped within oppida and ports during protohistoric periods; the development of agricultural facilities scattered across the countryside during Antiquity, with urban centres typically located on strategic heights; and settlement on the coastline, and in the hills and plains of the middle valley during the high Middle Ages (Figures 4, 5 & 6). Despite these long-term differences, the river and its tributaries continuously structured human occupation of the valley. As a strategic passageway from the coast to the interior, a supply of fresh water and a fertile alluvial plain, the Vidourle and its tributaries have animated and guided the locational choices of its past inhabitants. These demonstrate a desire to be close to the river for convenience and in order to control it; in parallel, the hills and elevated areas provided security against flooding, particularly for urban areas, and the possibility for wider territorial control. As an axis of penetration into the hinterland, the Vidourle was also a trade route between the interior and the sea (Reference ScrinziScrinzi 2014: 331–95).

Figure 3.

Number of newly created sites and site abandonments per century between the seventh century BC and the tenth century AD (© M. Scrinzi).

Figure 4.

Distribution of settlement in the Vidourle valley in the sixth century BC (© S. Sanz & M. Scrinzi).

Figure 5.

Distribution of settlement in the Vidourle valley in the first century AD (© S. Sanz & M. Scrinzi).

Figure 6.

Distribution of settlement in the Vidourle valley in the tenth century AD (© S. Sanz & M. Scrinzi).

Analysis of long-term settlement and land use in the Vidourle valley provides a model for the eastern Languedoc. How does it compare with the wider Mediterranean context? Is it typical of the settlement systems encountered in the river valleys of Spain, Italy and North Africa? Or is each Mediterranean valley different? The results from investigations in the Vidourle valley stand as a starting point for a wider research programme extending into neighbouring regions and countries, with the aim of establishing a methodological and analytical framework for the study of settlement in the river valleys of the western Mediterranean.