Internet Archaeol. 9. Millett et al. Ave Valley Survey

1. Introduction

1.1 Background and research design

This report presents the results of a field survey undertaken between 1994 and 1997. The survey set out to examine a sample area of the landscape in northern Portugal with the primary objective of investigating the impact of Roman annexation on the indigenous landscape of a region which had a sophisticated pre-Roman Iron Age culture. The original aim was thus to undertake a survey which would produce results for direct comparison with those produced by the Ager Tarraconsenis survey which had recently been completed (Carreté et al. 1995). The research design for the project is summarised in the following extracts from the document prepared in 1993 as the basis for funding applications:

The aim of the survey is to investigate the Iron Age, Roman and early Medieval periods in the Ave valley in the Minho region of northern Portugal through the application of surface survey techniques. Initially work will utilise the field walking methods used in the Ager Tarraconensis survey with the aim of collecting comparable data. At a later stage it is intended to undertake more detailed studies of a representative selection of sites through the use of geophysical and geochemical techniques.

It is also intended to develop techniques for the application of remote sensing techniques to the mapping and interpretation of the survey data in relation to the topography, geomorphology and soils.

There has been considerable work on the well-known castro sites in this part of the Iberian peninsula. Good stratigraphic sequences have been produced through excavation, and contacts with the Mediterranean world have been demonstrated through the study of the distribution of Greek and Punic pottery on indigenous sites prior to the eventual Roman annexation by 19 BC. A recent synthesis of these data has been provided in Dr Queiroga's Oxford D.Phil. thesis. This makes it clear that there are a series of research questions which can be addressed through survey.

From a broader perspective there is now an accumulation of survey evidence from Mediterranean Spain and the Guadalquivir valley. In contrast, these techniques have not been widely used for the gathering of rural settlement evidence from the Atlantic regions.

The Ave valley basin provides an excellent sample area for survey. It lies in the far western part of the Roman province of Hispania Tarraconensis, close to the border with Lusitania and on about the same latitude as Tarraco. The river basin also contains the conventus capital for this region, Bracara Augusta (Braga) so it will be possible for the survey to provide evidence to make direct contrasts with the results from the Ager Tarraconensis survey.

The known sites in the region are presently almost exclusively found on hill tops. There is a rich agricultural landscape, with the valley floors and coastal plain heavily cultivated for the production of arable crops and vines. The river basin drains westwards into the Atlantic where there is a broad coastal plain. In its middle reaches the river flows through a reasonably broad valley with steep and wooded interfluvial ridges. Further inland the valleys narrow and there is extensive terracing and more widespread woodland.

The survey will be designed to investigate a series of 1km wide northsouth transects across the river valley, spaced to sample the landscape at intervals from the coast to the interior. This will enable us to examine the following specific questions:

a) What is the distribution of archaeological material across the valley within the survey area? How does this vary through time?
b) How does this distribution of material relate to the known sequences of the castros? In particular is there any evidence for the High Empire when many castros appear to be deserted?
c) Is material in the valleys the result of cultivation (manuring) and/or the dispersed agricultural settlements?
d) Does the distribution of material and the composition of assemblages vary with distance from the coast up the valley?
e) Does the distribution of material and the composition of assemblages vary with distance from Braga?
f) Are there geomorphological changes within the river valley which can be correlated with changing land use in antiquity?

In the course of the development of the survey, various of the lines of enquiry described in the original research design were modified as a result of our changing understanding of the region and its potential for field-walking survey. Details of the modified methodology are outlined in sections 4.1 & 4.2). Of particular importance, however, was the discovery that in contrast to the Tarragona area, ceramics were much less common as surface finds in this region. This, combined with difficulties in establishing the chronology of the pottery collected, meant that questions concerned with the transition from the Roman to medieval periods could not be addressed adequately. Nevertheless, the work has produced significant new insights into the Iron Age and Roman landscapes of the region.

Interim reports on the project were published annually between 1995 and 1998 (Millett and Queiroga 1995; 1996; Millett et al. 1997; 1998). These summarised the development of the project as understood at the time but the results they describe are superseded by the present publication.

This report is designed to provide the definitive publication of the fieldwork undertaken on the Ave Valley survey between 1994 and 1997 although it will be complemented by a conventional print publication which places the results in a broader context (Millett forthcoming). This article begins with a geographical introduction and an assessment of previous knowledge of the area. It then presents the methods of study, the finds reports and distribution maps. A discussion of settlement patterns is based on analyses of these results.

As a final introduction to this study, it is also appropriate to note the method of publication, as this is one of the first archaeological surveys which has attempted to use the medium of electronic publishing to fulfil its aims. As well as exploring the potential for publishing spatial information through this new medium, we have also made a first step in attempting to link the publication with an integral database. Although this has been a learning experience, we hope that by publishing in this way we have been able to make more information fully available in a more versatile manner than would have been possible through conventional publication media. We trust that this will enable others interested in the region both to see the information on which we have based our conclusions and to interrogate the data.

1.2 Authorship

The article as presented has been put together by Martin Millett and Kris Strutt and contains material prepared by Steven H. Willis, Jeremy Taylor and Franciso Queiroga. This information has been integrated into a single text except in the case of the pottery analysis which is wholly the work of Steven Willis.