This remarkable, highly readable book is the first reliable overview of the archaeology, history and geography of ancient North Africa ever published in English. It is timely: scholarly interest in the region is growing fast, and it makes an ideal complement to the essays recently published in Bruce Hitchner's Companion to North Africa in Antiquity (Reference Hitchner2022). It takes a new approach to the topic, covering a vast geographical range reaching from Cyrenaica to the Canaries and south to the Sahara across more than a thousand years, from the beginnings of what Mattingly labels here the North African Iron Age (NAIA) through to late antiquity. And it is a radical call for change in the field, a reflective and reflexive approach to ancient North Africa that confronts the failings of past scholarship, but not at the expense of the stories of North Africans themselves: the primary focus is not on ancient colonial powers based in the north but on the activities and experiences of autochthonous peoples. Indeed, if there's anything wrong with the book, it's that the subtitle undersells it: this isn't only about ‘Africa in the Roman Empire’ – itself a great improvement on the traditional ‘Roman Africa’ – but Africa beyond the Roman Empire, too, and before it, in a period that Mattingly shows was foundational for later imperial success.
Part 1 introduces the project. Chapter 1 lays the groundwork, introducing the sources and their problems, as well as the geography, environment and inhabitants of the region. Mattingly modifies Brent Shaw's (Reference Shaw2003) model of ancient North Africa as a series of environmental ‘islands’ by adding more areas of dense occupation, including oases, and questioning the notion that these islands were cut off. Chapter 2 then sets out the stall, surveying the defects of traditional approaches to North Africa, where modern imperial powers presented themselves as successors to the Romans, and celebrated their positive impact on the region as a whole. The notion of ‘nomads’ roaming beyond a Roman frontier helped to emphasise the primitive nature of the indigenous population, and the dangers they posed to imperial civilisation, while the ‘Romanisation’ model was an attractive paradigm for empire itself, ‘both palliative and compensation for subjugation’ (5). Mattingly prefers ‘discrepant identity’, part of a more general and most useful revision of the scholarly vocabulary relating to the region: not only does NAIA replace ‘protohistory’, with its suggestion that history itself only starts with the introduction of writing by outsiders, but we also lose ‘tribes’, ‘natives’, ‘slaves’ and ‘frontiers’, where Mattingly prefers ‘security zones’.
Part 2 (chapters 3–5) covers Africa before Rome, dealing in turn with the settler populations (Phoenician, Latin/Italic and, to a lesser extent, Greek-speaking); the local ‘Libyans’ or ‘Afri’ of the north, including the Mauretanian and Numidian kingdoms (where the account can now be supplemented by Ardeleanu Reference Ardeleanu2021, which appeared too late for full incorporation here); and to their south ‘Gaetulians’ and ‘Garamantes’, building on Mattingly's own revolutionary body of work on these populations to paint a vivid picture of oasis-dwellers farming the desert from the early Iron Age.
More generally, this section debunks the myth of an underdeveloped pre-colonial Africa. The ideas that urbanisation, agriculture and metallurgy were imported to the region from the northern and eastern Mediterranean are ‘keystones of the orthodox interpretation of paternalistic incomers jump-starting social developments in Africa’ (126). It doesn't help that there has traditionally been little archaeological work on the pre-Roman period, and what there was tended to focus on Carthage and its own ‘Punic’ world. But in the last 15 years that picture has changed dramatically. The best new evidence comes from Althiburos in the Tunisian Tell, where excavations have revealed a substantial sedentary farming settlement dating to at least the ninth century BC, with a ‘fully urban character’ by at least the fifth. The question of the transmission of metallurgy is still hazy – the eighth-century evidence for ironworking at Althiburos probably just postdates evidence for settlement at Carthage – but Mattingly authoritatively summarises the picture that has emerged from projects across North Africa in recent years that urbanism and agriculture were already widespread well into the desert before Mediterranean settlers first arrived.
Parts 3–5 (chapters 6–11) then assess the evidence for military, urban and rural communities under Rome, underlining the indigenous contribution to what might otherwise seem a miracle of empire. Militarily, the African frontier can appear unusually calm. Only one legion held it, after all, by contrast with the 15 or more permanently stationed along the northern border of the Roman empire. But Mattingly disrupts this picture, pointing out the broad scope of military operations within and beyond the provinces, as well as the weight of the evidence for resistance and revolt over the first two centuries of Roman presence: ‘a logical response to Roman aggression or imposition of imperial structures’ (243). The Mauretanian provinces in particular were highly militarised, and Roman authority gradually extended south over oasis settlements beyond their notional border. It is no coincidence that military communities kept their cultural distance from the rest of the African population: the vast majority of dedications to the Dii Mauri, for instance, come from military contexts, while a similar proportion of dedications to specific African gods are by local civilians.
When it comes to cities, Mattingly argues that ‘Roman’ Africa built on the example of Numidian, Mauretanian and Saharan proto-urbanism as well as the Greek and Libyphoenician cities Romans found there (this should come as no surprise, he points out: throughout the Roman empire areas of distinctive urbanism built on pre-existing traditions, while regions such as Britain and Gaul largely continued to do without). Our own vision of ‘Roman’ North Africa is disproportionately shaped by a small number of sites, often those closely associated with Roman power. By contrast, Iron Age settlements buried beneath Roman ones have traditionally been undercounted or even ignored altogether, though clusters of megalithic burials provide an important clue to their existence. But the truth was always obvious: the vast majority of the hundreds of towns known from the Roman period have local toponyms. Surprisingly few, by contrast, boast the traditional markers of ‘Roman’ urbanism. Only around 10 percent of towns in Roman Africa, for instance, had theatres or amphitheatres.
In rural areas the Roman period did see the spread of agriculture into marginal landscapes, along with increased crop specialisation focused on grain, olives and vines. Large estates belonging to Roman senators and then emperors appeared, especially in areas with less developed agrarian traditions, and on Numidian royal lands where the new owners could exploit existing dependent labour. At the same time, both Punic and Libyan persist as epigraphic languages well into the Roman period, as do Libyan and Punic names, and local cults and burial practices. Mattingly suggests that this ‘language choice reflected a non-Latinate identity’ (491), but he provides evidence for more complex processes of decision-making as well. Function mattered: Latin, Punic and Libyan were all used for funerary inscriptions at Ghirza, for instance, but only Libyan (which appears to have been the everyday language of the settlement) was employed in the temple – perhaps on the grounds that you should address gods in their own language. The politics of occupation may have mattered as well: using Punic on funerary monuments in Tripolitania allowed locals to distance themselves from the local military community, with the emphasis more on ‘them’ than ‘us’.
Beyond the ‘provincialisation’ of Rome in the story, there is an equally important theme running through the book of ‘profound regionalism’, as Mattingly emphasises the variety of landscapes across North Africa, and the diversity of their human populations in all eras. More specifically, rural, urban and military communities under Rome were different social and cultural groups that differed within themselves as well. This approach cuts neatly through some long-standing debates. When it comes to the Roman army's effect on rural areas, for instance, it seems that Lisa Fentress and Brent Shaw were both right: Fentress (Reference Fentress1983) about the disruption it caused in towns with military associations, especially on the Aurès plains; Shaw (Reference Shaw1983) on the high level of continuity found in pre-existing communities, particularly those established on higher ground. It clarifies some problems in the pre-Roman period, too. Mattingly convincingly locates the ‘Libyphoenicians’, who lived on the coast and intermarried with the Carthaginians (Diodorus 20.55.4), to the east of Carthage, associating coastal cities further west more with Numidians and Mauri. He shows that most of the ‘Libyphoenician’ cities appear only after 500 BC, mapping onto increasing Carthaginian hegemony in North Africa, but suggests at the same time that they have multiple origins, ‘from predetermined colonial towns to settlements that evolved and grew out of trading contacts’ (84). He also notes that their material culture suggests broad engagement with Mediterranean societies beyond the Phoenician-speaking world, a conclusion now buttressed by paleogenetic work on human remains from Kerkouane, a city on the coast of Cap Bon that flourished from the sixth to third centuries BC. Only five out of 12 individuals studied had clearly local heritage, while the genetic profiles of the other seven were more similar to those found in ancient Sicily and central Italy (Moots et al. Reference Moots2023).
These new data also support Mattingly's bigger picture of mixed human populations in North Africa, ‘blurring the hard lines’ of ancient and modern categories (37). And there may be opportunities to blur things even further. There is, for example, the question of the Gaetulians, who play an important role in the book as early farmers and urbanisers, and then as a significant source of political and cultural resistance to Roman empire. But Gaetulians appear in our sources rather late. The first surviving work to mention them is Sallust's Jugurthan War, where the author recounts an origins myth that makes them out to be one of the two original populations of North Africa, along with the Libyans. They are rough and uncivilised hunter-gatherers, who were ‘governed neither by institutions nor laws nor by anyone's authority’ (18.1). In Sallust's own time (the 40s BC), they still lay beyond state power, south of Jugurtha's own Numidian kingdom, and this impression is later echoed by other Greco-Roman authors, who place Gaetuli across a wide range of African desert steppe landscapes from the Atlantic to the Syrtes.
To what extent then is this ‘people’ the product not of indigenous practice or experience but the colonial imagination? There is no doubt that an important ecological boundary falls around the edge of the Numidian kingdom, where Sallust's Gaetulians start. This is the northern limit of date cultivation in desert oases, the southern limit of olive, vine and cereal farming. Cultural division is less clear, at least in the pre-Roman era, and lifestyles are not homogeneous: while many of the Romans’ (and Mattingly's) Gaetulians occupy desert oases, like the Garamantes, others live in mountain hillforts to the north that recall Numidian settlements.
The case for the Gaetulians depends in part on a recent analysis of modern Tamazight (Berber) languages by Christopher Ehret that distinguishes between two different versions of proto-Amazigh spoken in ancient North Africa, with one splitting off from an earlier branch further east around 1000 BC and the other emerging or (more likely) arriving in the region about 500 years later (Ehret Reference Ehret2019). Lisa Fentress has made an attractive case for mapping usage of these two hypothesised languages in the later first millennium BC to north and south of the ancient Numidian ‘border’, with the Gaetulians and Garamantes using the younger tongue (Fentress Reference Fentress, Gatto, Mattingly, Ray and Sterry2019). As Mattingly notes, however, it is still speculative: Fentress herself describes her interpretation of the linguistic data as a ‘(pre)historical novel’, while Ehret's method and results have provoked considerable skepticism among specialists in Berber linguistics. At the same time, an extensive corpus of inscriptions in the ‘Libyan’ script that these languages shared appears across the whole region in the later first millennium, from the central Sahara to the Mediterranean coast. And this is not the only way in which the Gaetulian cultural koine bleeds into others to north and south: depictions of chariots are restricted to the flat landscapes of Saharan rock art, but imagery of mounted warriors armed with swords and circular shields reach north from the desert through the mountains to the sea.
Where clear cultural differences between Gaetulians and Libyans do emerge it is after Roman conquest and in the religious sphere. In this era a large number of Libyan communities practised the tophet rituals originally introduced by Phoenician-speaking migrants, and normalised over the early Roman era as ‘Saturn’ cult. Further south, however, Mattingly shows that the tophet cult is restricted to Roman military communities; below the olive line it is completely absent, and we find instead an emphasis on ancestor worship. Mattingly ingeniously suggests that the contrast may reflect the different demands of very different landscapes on their inhabitants. On this reading the dry farmers of the north appealed to the gods who controlled the rain on which they depended, while the irrigated landscape of the south depended instead on water systems built by earlier generations, and now operating in the shadow of their tombs. He observes that sculpture on Gaetulian mausolea and Saturn stelae shares distinctive themes of agricultural fertility and productivity, suggesting that farming concerns were taken to the god in the north, to the ancestors in the south. However that may be, a similar distinction is later found in the ‘reluctant and thin adoption of Christianity in Gaetulian zones beyond the garrison settlement’ (519).
Whether these Gaetulians themselves had any investment in collective identity is less clear. Self-declared Gaetuli are relatively rare even in the Roman period, and they appear only in Roman cultural contexts. Inscriptions reveal Roman citizens at and around Lepcis with the cognomina Gaetulicus and Gaetulia, but in a world where Roman naming conventions are the norm, using a Roman term for people beyond the pale makes cultural sense in Roman rather than Gaetulian terms. This is clearer in the case of a veteran named C. Iulius Getulus who erects a bilingual Latin/Libyan inscription at Thullium in north-east Algeria. Given his military past, the name Getulus again makes Roman sense: there were auxiliary alae of Afri and Gaetuli in the Roman army. But in his own Libyan language he is called Keti of the Misiciri. Romans would have understood the Misiciri to be a subset of their Gaetulians, but that isn't how the Misiciri described themselves to local audiences. Even Apuleius’ famous description of himself as half-Numidian, half-Gaetulian is explained in a Roman courtroom in coastal Tripolitania not as an expression of identity or heritage, but as a simple reference to the location of his hometown of Madauros on the border between what was in the Roman era Numidian and ‘Gaetulian’ (more specifically Musulamian) territory – and therefore, from the point of view of his accusers, suspiciously far away (Apology 24). Apuleius himself rejects their implied association of both these terms with barbarism, but not this basic interpretation of his own words.
If the Gaetuli are a colonial invention, however, they are probably not a Roman one. Sallust's original story of their deep past is not his own, but comes – he says – from a myth recorded in ‘Phoenician books’ that were supposed to have been written by a Numidian king called Hiempsal (Jugurthan War 17.7). The invention of Gaetulians by Numidian imperial rhetoric during the North African Iron Age would be entirely in line with Mattingly's suggestion that ‘the last centuries BC and early centuries AD were marked by massive instability of ethnic groups and other forms of social organisation’ (117). It may also be that by adopting the concept of ‘Gaetulians’ ourselves we continue to divide the world by colonial categories: those within and beyond first Numidian, then Roman, rule.
The book ends with ‘some final themes’: a chapter on the economy, with particularly interesting reflections on trans-Saharan trade, and one on ‘African Agency’, summarising the argument that ‘Africa became one of the most densely urbanised and richest provinces in the empire because it was a densely occupied zone of sedentary agriculture long before Rome’ (561). It is beautifully produced, with many colour illustrations, splendid maps drawn by Martin Sterry and one curiosity: all quotations are translated into English, but the original text is only given for modern, not ancient, languages.
