This chapter reviews the spread of irrigation technology across the Sahara in antiquity, and its effects on settlement agriculture and the movement of people. Recent work has stressed the close connections between the introduction of foggara technology and the rise of Garamantian civilisation, which featured intensive agriculture and incipient urbanism. However, many oases achieved substantial size through the use of well technologies, artesian springs or a combination of technologies. Another key question relates to the effects of the eventual decline and failure of these irrigation systems in terms of population movement and fragmentation of states such as the Garamantes. After presenting new AMS dating evidence for Garamantian foggaras, the chapter advances the discussion by examining the wider picture of foggara distribution within a survey of the evidence of irrigation technologies across the Sahara and whether and to what extent the distribution of foggaras beyond the core Garamantian heartlands might be seen as an indication of Garamantian control or influence. It explores what foggaras, wells and new crop introductions might suggest about agricultural intensification and organisation. This has implications for assessing agricultural intensification in the ancient Sahara. Finally, it considers causes and possible effects of irrigation failure and in some cases collapse.

The southern oases of the Sahara can be split into two broad groups. First, there are a series located in and around the Saharan mountain ranges. From east to west these are: Ennedi, Tibesti, Aïr, Tassili n’Ajjer, Ahaggar and Adrar des Ilforas. These massifs receive higher levels of rainfall than other parts of the Sahara, resulting in seasonal rivers that feed permanent bodies of water (small lakes or pools/gueltas) or that support a high water table beneath the wadi beds. There are still populations of crocodiles living in Ennedi and nineteenth- and early-twentieth-century reports of them also in Ahaggar and Tibesti. However, these areas have traditionally been primarily exploited by pastoralists and they have little suitable land for oasis agriculture and highly variable rainfall. The dynamic wadi systems may also be a factor in the poor preservation of archaeological remains with less sturdy or older constructions either swept away or buried under sediment.

This chapter focuses on the oases of the northern Sahara, both those close to and in some cases incorporated within the frontiers of the Roman provinces of Africa. With the exception of some outstanding contributions from our co-author Pol Trousset, there has been little consideration of the potential scale or on the ground reality of oasis development in the Roman frontier region. This is only partly explicable in terms of the lack of detailed archaeological work at these sites – as we shall demonstrate there is quite a lot of fragmentary evidence to support the case for widespread oasis development in pre-Islamic times. In large measure the lack of recognition of the importance of oases here relates to the long-prevailing myth that Rome was confronted in this frontier zone by nomadic (or at best transhumant) peoples. It is hoped that what follows will provoke a full re-evaluation of Rome’s African frontiers and what they were designed to deal with.

This volume explores a series of linked themes that have wide relevance in world archaeology: sedentarisation, urbanisation and state formation. In this opening chapter we review some of the key background to recent debate on these themes and identify some of the Saharan particularities which complicate the application of models developed elsewhere.

A dominant discourse on the Sahara throughout history has been the idea of a ‘nomad menace’, coupled with a persistent emphasis on the Sahara as largely uninhabited and uninhabitable. It is true that pastoralism has at all times been a key mode of life and mobile populations have underpinned the development of networks variously used for trade and raiding. Yet the lifestyle and inter-relations of mobile peoples of the historic Sahara, such as the Tuareg, have always been contingent to a greater or lesser extent on the existence of sedentary communities, both within the Sahara and at its fringes.

This chapter will review the evidence of early oasis development in Western Egypt and Eastern Libya, broadly following the course of the ‘route of the oases’, running west from the Nile to Siwa, then onwards to Awjila and al-Jufra in Libya, where it met the major north-south route from the Mediterranean to Garamantian Fazzan and beyond to Chad. The evidence presented for pre-Islamic oasis development is particularly strong in this part of the Sahara; indeed the origins of agriculture at some of the Egyptian oases went back to the third millennium BC and the route as a whole seems to have been well-developed by the fifth century BC.

We suggest that the ultimate origins of oasis agriculture in the Western Desert are to be sought in the Nile Valley and the Fayum, with a package of plants and irrigation techniques first developed there, then adopted in the oasis depressions of the Western Desert – notably Kharga, Dakhla, Farfara, Bahariya and Siwa (Fig. 3.1).

This chapter is intended as an analytical counterpoint to the descriptive presentation of the evidence of pre-Islamic activity at Saharan oases and as an engagement with some of the more theoretical introduction to the volume. Here we make the case for a dynamic Saharan history with frequent episodes of sedentarisation and urbanisation in the past. The related issue of state formation will be considered as the conclusion of the book, with a discussion of how we might define states and state formation in the Saharan milieu and the contribution that oasis centres have made to early desert states, like the Garamantes. We do not seek to deny the importance of pastoral lifeways, or the social complexity of such mobile societies, but we argue that the pastoral and mobile dimension of the Trans-Saharan world has been over-emphasised to the neglect of its sedentary communities. Several factors play into the choices made by communities as to how to live: social and political circumstance, environmental affordances, regional dynamics and trade. The fact that pastoral groups have tended to be more dominant in later history does not exclude the possibility that in some earlier ages sedentary and oasis populations have had greater social and political weight. Trans-Saharan history we suggest has been characterised by a continually changing power balance between pastoralists and sedentarists (as is also evident at times in the recent past).

In this concluding chapter we focus on debates about state formation and to what extent it is possible to recognise historical instances of this key process among Saharan societies. Here we make the case for a dynamic Saharan history with frequent episodes of urbanisation and state formation in the past. However, we also emphasise the ephemeral nature of these polities, which challenges more evolutionary models of the rise of towns and states. Towns and states have existed among a range of forms of organisation utilised by Saharan societies, with the balance of power swinging back and forth between sedentary and more mobile lifestyles.

Our exploration of these factors commences with general questions about the definitions and types of towns and states that are detectable in the Trans-Saharan world, coupled with a discussion of the sorts of models that can explain their rise and fall.

In this chapter we present a case study of early oasis development relating to the area of south-west Libya known as Fazzan and the ancient people known as the Garamantes. The Garamantes were first firmly identified with this area by Duveyrier in the mid-nineteenth century. As will be explained in more detail below, the archaeological rediscovery of the Garamantes properly commenced with a pioneering Italian mission in 1933. There were some large-scale excavations in the 1960s by Mohammed Ayoub and importantly this included work at the Garamantian capital of Old Jarma (ancient Garama), but the poor quality of the work limited the value of the published outputs. More reliable results were achieved by the team led by Charles Daniels between 1959 and 1977, excavating and surveying a number of sites in the area close to Jarma.

This chapter focuses on the northern oases of the Western Sahara Fig. 6.1). As with the Northern Sahara, the North-western Sahara has also too often been dismissed as wholly nomadic before the development of Arab trading networks. There have been only limited attempts to view the region as a whole, as opposed to restricted studies of the different oasis groups. In this chapter we start with the western end of the Saharan Atlas and move west towards the Atlantic through the Wadi Gir, Wadi Saoura, Tafilalat, Wadi Draa and Wadi Noun. We conclude with a diversion to some outlying oases to the south, Gourara and Tuwat at the southern end of the Wadi Gir and Tabalbala c.300 km to the west of Tuwat. This broad zone also needs to be considered in relation to the south-western Saharan sites that are presented in Chapter 7.