1. Introduction

Although visibility analyses have been implemented extensively within archaeological GIS (e.g. Ruggles et al. 1993; Llobera 1996; 2006; Wheatley 1995; van Leusen 1999; Wheatley and Gillings 2000; Lake and Woodman 2003; and papers in the 2007 edition of World Archaeology 39(1), there are grounds to suggest that they are still at an experimental stage. Current research is based upon binary viewsheds, where visibility is shown through a grid of cells with the value 1 (= visible) or 0 (= non visible). However, visibility analysis is a much more complex issue than that, with a wide number of factors affecting its accuracy and interpretation. A number of methods to overcome the limitations of traditional binary viewsheds have been put forward (Wheatley and Gillings 2000; Lake and Woodman 2003). Nevertheless, most practical applications tend to concentrate on only one of the several methodological issues that affect viewsheds' accuracy (e.g. multiple viewer points). This, I will argue, can lead to flawed archaeological interpretations.

Attention will be shifted here towards replacing binary viewsheds by a combination of visibility analyses, taking into account a variety of aspects that are crucial to obtain accurate viewsheds and significant interpretations.

Figure 1: Map of Europe showing Catalonia; the study area is marked by a square (after Esri Data and Maps 2006)

This methodology is used to investigate visibility from Iberian hillforts in a landscape north of Barcelona (Catalonia), around the town of Badalona (Figure 1), during the 3rd century BC, i.e. just before the landscape was affected by processes of Romanisation.

The Iberian people who inhabited the area were part of the Laietani tribe, as mentioned by texts from the Roman period (Strabon Geography III, 4, 8, Pliny Natural History III, 21 and Ptolemy II, 6, 18 and II, 6, 72). According to these texts, the Laietani inhabited the territory between the Llobregat River (south of Barcelona) and the town of Blandae (Blanes). However, cultural differences between the Iberian sites make the limits of the Laietania region a subject of debate (Prevosti 1995, 21-22). Although the degree of political and cultural unity within Laietania is also an unclear issue, it is often understood to have existed as a political entity ruled by Burriac hillfort (García et al. 2000, 30-31; Ruiz and Sanmartí 2003, 42), which is located outside of the study zone, near Mataró. The region analysed lies in the central part of Laietania as described by the Roman sources. There are grounds to suggest that one of the Iberian hillforts within it, Turó d'en Boscà, would have acted as a focal point for the others, probably developing a role of political and administrative centre, secondary to Burriac.

The precise limits of the study zone (Figure 2) broadly follow natural geographical features, framing a coastal strip between the sea and the Serralada Litoral mountain range, which is formed by hills of no more than about 500 metres altitude.

Figure 2: Map of the study area

When the Romans disembarked in Emporion in 218 BC, during the Second Punic War against Carthage, the Iberians of the region would have been living mainly in hillforts, urbanised since about the 4th century BC. These settlements, on the hilltops of the Serralada Litoral, are distributed among an inner and a coastal line of mountains. The only other documented sites consist of scatters of pottery and a single pit cluster (Figure 2). Although they could correspond to rural settlements, the pottery could simply indicate transient occupation of the area, whereas the pit cluster would seem to be related to a nearby hillfort. In addition, points referred to as possible Iberian sites have also been plotted. Since they contain pottery with a broad chronological range, they could have been either Iberian and/or Roman. Accordingly, there are no grounds to establish if the plain was occupied by rural settlements (Ruestes 2006).

A web of routes would have been crucial for the articulation of the territory. Two main pathways parallel to the sea would have been used: one following the ridge of the inner mountains, and another, broadly dated to the Iberian period (Olesti 1995, 243), passing through the southern slope of the littoral mountains. Moreover, natural corridors at right-angles to the sea that follow dry streams would have enabled communication between an inland valley suitable for cereal cultivation (Zamora et al. 2001, 221) and the coast.

The method presented here is designed to address visual control of the landscape from the hillforts and intervisibility relationships between them. These two questions had previously been tackled (Ruestes 2006) through the generation of simple binary viewsheds from a single viewer point. On the basis of the results gained by this research, the hypothesis of a highly structured society was proposed where each hillfort might have visually controlled a given zone of the landscape, linked to the other hillforts through a visibility network. However, if this hypothesis was to be tested, a much more robust visibility analysis was needed and a number of issues had to be addressed.

Due to the inherent inaccuracy of the DEM (Digital Elevation Model), visibility analysis from a given location can lead to divergent results when performed using different DEMs of the same area.
Viewsheds from a single fixed viewer point may result in only partial visible areas from a given settlement being identified, since looking would involve movement and a variety of viewpoints.
Theoretically being able to 'see' an area does not necessarily imply being able to control events taking place in it, not least because visual clarity reduces with distance.

This article proposes a multi-technique visibility analysis intended to address these shortcomings and thus to reach more significant archaeological conclusions.

PREVIOUS NEXT CONTENTS HOME