8

The Roman Empire, AD 14–284

In AD 48 a small delegation arrived at Rome from the distant plains of northern Gaul. The leading nobles of the north Gallic provinces were seeking the right to hold office in Rome itself, in particular the right to apply for membership of the Senate. The Senate was, unsurprisingly, not especially keen on the idea. The issue was decided by a lengthy speech of the emperor Claudius to the Senate in support of the Gauls’ petition.

Do not shudder at the thought of some dangerous novelty being introduced. Reflect, instead, on how many innovations our state has seen; think how many different changes our constitution has undergone, starting right from the very foundation of our city itself. Once, the city was ruled by kings; yet they failed to pass it on to native heirs. Instead, it was other men, foreigners, who took their place. Romulus was succeeded by Numa, a native of the Sabine country – a neighbour, for sure, but a foreigner nonetheless … it was a wholly novel policy, too, when my great-uncle the deified Augustus and my uncle Tiberius Caesar wished to bring into this Senate house the flower of the colonies and municipalities, wherever it was to be found, so long as they were sound and wealthy men.

As we saw in the last two chapters, the question of the incorporation of non-Romans into the Roman state had a long history. Nonetheless, in this speech to the Senate, Claudius was deliberately overturning centuries of received wisdom. The deep conservatism of Roman political thought has been emphasized again and again in the last two chapters. But Claudius now argued that the history of Rome had been characterized by political innovation right from the outset. The main lesson that the past had to offer was the value of political change and novelty. Not only had new men always been freely absorbed into the Roman body politic, but the constitution itself had always been in flux. The Roman historian Tacitus, who included a paraphrase of Claudius’ speech to the Senate in hisAnnals, finished the speech with his own telling flourish: ‘This proposal, too, will grow old, and that which we defend today with precedents, shall one day be counted as a precedent itself.’

Tacitus, looking back to the reign of Claudius (AD 41–54) seventy years after the event, must have appreciated the irony. By the time Tacitus was composing the Annals, late in the reign of the emperor Trajan (98–117), the Republic was gone for good. TheAnnalsbegin with the words ‘The city of Rome from the beginning was held by kings’. Not ‘in the beginning’, but ‘from the beginning’: in Tacitus’ eyes, the principate, rule by a single ‘first man’ or princeps, was monarchy in all but name. The revolutionary politics of Augustus had indeed become hallowed precedent.

The Annals recount the history of Rome under the Julio-Claudian dynasty (Tiberius–Nero, AD 14–68). In the early years of the Julio-Claudians, unlike in Tacitus’ own day, elite resistance to the principate had still been possible. Early in the Annals, Tacitus describes how, in AD 25, the historian Cremutius Cordus was prosecuted for high treason. His crime was to have used his work to praise Brutus and Cassius, the assassins of Julius Caesar, and specifically to have quoted Brutus’ description of Cassius as ‘the last of the Romans’. Defending himself before the Senate and the emperor Tiberius (AD 14–37), Cremutius argued that he had done nothing unusual or seditious. Many earlier historians – Livy, Asinius Pollio, Messalla Corvinus – had praised Brutus and Cassius in fulsome terms. Surely the assassins had now been dead long enough for there to be no danger in giving them their due honour? As Cremutius himself must surely have expected, this argument cut no ice at all with Tiberius. The historian was compelled to commit suicide, and his books were burned. In AD 25, when the last days of the Republic were still within living memory, Caesar’s assassination could not simply be dismissed as ancient history. The figures of the ‘tyrannicides’ carried a particular political and emotional charge. Junia, the sister of Brutus and wife of Cassius, had died only three years earlier, in AD 22; images of her brother and husband were, says Tacitus, ‘conspicuous by their absence’ in the funeral procession. The contested memory of Brutus and Cassius would remain a raw spot on the Roman imperial psyche for at least another generation; as late as AD 65 a lawyer could be sent into exile for keeping an image of Cassius among the portraits of his ancestors.

In AD 41, after the assassination of the unpopular emperor Caligula (AD 37–41), there was a brief moment when the restoration of senatorial government may have seemed a real possibility. The Jewish historian Josephus, our main source for the last days of Caligula, depicts Caligula’s killers as Republican idealists trying to re-establish liberty and the ancestral constitution in the face of tyranny. At any rate, it rapidly became clear that there was no popular support for a return to the Republican system, and the praetorian guard, the emperor’s personal bodyguard, acted quickly to have another member of the imperial family, Claudius, acclaimed as emperor. Tyrant-slaying had had its day.

The new ideology of the principate is at its clearest in a chilling essay by the Stoic philosopher Seneca. In AD 55, only fourteen years after Caligula’s death, Seneca published a short treatise On Clemency, addressed to the new emperor Nero (AD 54–68), Seneca’s own former pupil and protégé. It is, says Seneca, right and necessary that the emperor’s power should be absolute. The emperor is the animate soul of the body politic; without him, the state would swiftly descend into chaos. Although Seneca stops just short of calling Nero a ‘king’ (rex), he is quite clear that Brutus’ fears of autocracy, which led him to assassinate Caesar, were misplaced: the perfect state is, on the contrary, that which is well governed by a just king. The main purpose of the treatise is to urge the new emperor to exercise his limitless power with mildness and restraint, just as Augustus had done. This was the best for which a member of the Roman elite could now hope.

People like Seneca and Tacitus responded to the new world order with realism and caution. Tacitus’ first historical work, the Agricola, published in AD 98, is a biography of his father-in-law, Julius Agricola. Agricola had been one of Rome’s foremost generals under the Flavian dynasty of emperors (Vespasian, AD 69–79; Titus, 79–81; Domitian, 81–96). After a notably successful governorship of the remote province of Britain (AD 78–84), Agricola returned to Rome, only to be hustled into a humiliating premature retirement. There were limits to the glory that a general was allowed to win for himself; no one, however talented, could be allowed to outshine the emperor himself. ‘A great reputation’, says Tacitus, ‘was as dangerous as a bad one.’ By accepting his removal from public life, Agricola showed not only good sense, but even a kind of heroism. ‘Let those whose habit it is to admire forbidden ideals know that there can be great men even under bad emperors; that duty and discretion, if combined with industry and energy, will bring a man to the same heights of honour as others have achieved through perilous courses and ostentatious deaths, with no advantage to the state.’

Much of the Agricola is taken up with an account of Roman rule in Britain, from the initial conquest of the southern part of the island under Claudius (AD 43) to the campaigns of Agricola in northern England and Scotland. Britain was the last major part of western Europe to be incorporated into the Roman empire, and the conquest was drawn-out and bloody. In AD 60 a fierce rebellion broke out in the south-east under the native queen Boudica. The rebels succeeded in sacking the Roman colony at Camulodunum (modern Colchester), along with the towns of Londinium and Verulamium (modern St Albans). The revolt ended in a bloodbath, somewhere in the west Midlands, at which 80,000 Britons were said to have been killed. The defiant speech which Tacitus later puts in the mouth of a – probably fictional – Scottish chieftain, Calgacus, has some truth to it: ‘to plunder, butchery, and rapine, they give the lying name of “government”; they create a desolation and call it peace.’

From Boudica to Boadicea

The figure of Boadicea (as she has usually been called) has been put to various different uses by the British over the centuries. During the war with Spain in the late sixteenth century, Elizabeth I was often compared to the original British warrior queen; it is even possible that Elizabeth imitated one of Boadicea’s speeches (as reported by Tacitus) in her own speech to the English army at Tilbury in 1588, shortly before the battle of the Armada. In William Cowper’s Boadicea: An Ode (1782), Boadicea’s heroic defeat at the hands of the Romans was set up as an icon and precursor of British imperialism.

Probably the best-known modern image of Boadicea is Thomas Thornycroft’s great bronze statue of Boadicea and her Daughters (started in 1856, but not set up until 1902) which today stands in front of the House of Commons in London, next to Westminster Bridge (see Plate 24). Prince Albert took an active interest in the statue’s design, urging Thornycroft to render the queen and her chariot in as regal and poetic a manner as possible (a ‘throne upon wheels’). Boadicea is depicted as a dignified warrior, wielding a spear and standing poised on her chariot behind a rearing pair of horses. The inscription on the front of the pedestal claims Boadicea as a British patriot and national hero: ‘Boadicea, Boudicca, Queen of the Iceni, Who died AD 61, After leading her people against the Roman invader’. On the east side of the pedestal there appears a quotation from Cowper’s Boadicea: ‘Regions Caesar never knew, / Thy posterity shall sway’. Thornycroft’s statue quietly encourages the viewer to see Boadicea as the ultimate ancestor of Queen Victoria, Empress of India.

However, not all British images of Boadicea have been so positive. In Tennyson’s Boädicéa (1859), she is presented as a bloodthirsty fanatic:

Mad and maddening all that heard her in her fierce volubility,

Girt by half the tribes of Britain, near the colony Cámulodúne,

Yell’d and shriek’d between her daughters o’er a wild confederacy.

Tennyson’s picture of a crazed barbarian, ruthlessly butchering the defenceless Roman colonists at Camulodunum, is surely intended to recall the atrocities unleashed on the British colonial settlers in India during the Indian Mutiny of 1857.

The site of Boadicea’s tomb is unknown; the modern urban myth that she lies buried beneath Platform 10 of King’s Cross railway station is no less plausible than any other theory.

Some truth, but not the whole truth. It is of course true that this huge empire was founded on overwhelming military dominance. The western provinces had been created by invasion, looting and, at times, killing on an awful scale. But Roman expansion had more or less come to a halt by the reign of Hadrian (117–38). The Roman empire of the second century AD stretched from the hills of Cumbria to the Nile valley, from the Portuguese coast to the desert plains of Jordan. Military superiority alone is no guarantee of a stable peace; yet that was precisely what Rome achieved. From the late first to the fourth century, with a few notable exceptions, the Roman provinces saw remarkably few internal revolts. Tacitus was right to complain of the difficulty of writing history under the principate; to all appearances, the history of Europe had indeed come to a full stop. The extraordinary success and stability of the Roman empire over the first three centuries AD is a historical problem which cries out for explanation. Much of this chapter will be dedicated to exploring the reasons for that stability.

*

We begin by considering how Roman rule outside Italy actually worked. The Roman empire was not, on the whole, governed by Romans. In any given year, the central government sent out a total of around 160 officials for a subject population of 50 million or more – fewer overseas officials than were sent out from Athens to administer its Aegean empire in the fifth century BC. The most important of these officials were the forty-odd provincial governors, appointed by the Senate or the emperor, each of whom held office for a period of between one and five years. The duties of a governor were not burdensome, being largely confined to provincial jurisdiction and local dispute-resolution. Governors’ day-to-day activity seems to have been, for the most part, crushingly mundane: around AD200, we hear of a governor of Asia (western Turkey) stepping in to deal with a bakers’ strike at Ephesus, and in 254 a governor of the same province was expected to fix the day of the month on which a local farmers’ market could be held. Instead, most of the real administration of the empire was undertaken by the local communities themselves.

The Roman empire was a world of cities. More than three hundred cities are known in the province of Asia alone; across the whole empire, they certainly numbered in the thousands. The cities – or, more precisely, the local civic elites – were responsible for the assessment and collection of taxes, urban and rural police duties, road-building and maintenance and their own food- and water-supply. One of the most significant changes in provincial administration between the late Republican and early imperial periods was the gradual phasing out of the firms of private tax-collectors (publicani), whose activities had caused so much rancour among provincials in the first century BC. By placing the burden of tax-collection in the hands of local bigwigs, Rome neatly delegated a major source of provincial resentment. Cities in the Roman empire, then, were not merely urban conglomerations; they were the indispensable cogs by which the whole imperial machinery turned.

In the eastern half of the Roman empire – particularly in Greece, Asia Minor and the Levant – there was a long tradition of city life, and for the most part, Rome simply preserved the pre-existing urban network. In the west, where far fewer towns had existed at the time of the Roman conquest, the necessary urban centres had to be created from scratch. In the early empire, Roman settlers, usually army veterans, were often parachuted into new model towns (coloniae), either founded on virgin territory or, more usually, replacing small native settlements. Elsewhere, pre-existing communities were encouraged to reorganize themselves on an urban model. The experience of the Aedui, a large Celtic tribe in modern Burgundy, was characteristic. In the immediate aftermath of Caesar’s conquest of Gaul in the 50s BC, furious building work started at the Aeduan hilltop oppidum of Bibracte, on Mont Beuvray. Thatched roofs were replaced by tiles, the streets were laid with paving-stones, space was cleared for a central marketplace and large new Italian-style courtyard houses started to appear. But the basic layout of the town, its meandering streets and dense clusters of tiny dwellings, remained stubbornly un-Roman. Around 15 BC, a full generation after the conquest, the Aedui finally decided that this simply would not do. Taking a deep breath, the inhabitants of Bibracte emigrated en masse to a greenfield site some 20 kilometres away at Augustodunum (modern Autun; see Figure 27). Working from a blank slate, the Aedui planned their new capital so as to look as much as possible like the orderly towns of Roman Italy: laid out on a regular grid plan, orientated around two main streets forming a cross, with a theatre and amphitheatre, massive stone city walls, forum and temples. The scale of the project was immense. Augustodunum was planned from the outset to cover around 200 hectares, with a 6–kilometre circuit wall; it may have taken another two generations before the place ceased to look like a permanent building-site. The really extraordinary thing is that no Roman governor had ordered any of this. It was the Aedui themselves who decided that they wanted a proper Roman city of their own, to the extent that they were prepared to give up on their first attempt altogether and plan an entire new city from the ground up.

Figure 27. Bibracte (Mont Beuvray) and Augustodunum (Autun)

The planned cities of Wren and Hawksmoor

Within a week of the Great Fire of London in September 1666, Christopher Wren presented to Charles II a plan for rebuilding the destroyed parts of the city (see Figure 28). Wren’s planned city was profoundly influenced by what he knew of the urban layout of Roman cities under the empire. The whole area around Fleet Street was to be reorganized on the model of the Roman architect Vitruvius’ description of the ideal city; the City and Stock Exchange were laid out on the model of the Roman Forum (as rather fancifully reconstructed by Palladio in the mid-sixteenth century), and Greek-style stoas overlooked the river Thames at Billingsgate. Major civic buildings (including, of course, Wren’s new St Paul’s Cathedral) were situated prominently at the end of long axial roads. Wren’s plans for ‘Romanizing’ London had strong political overtones. In his first Tract on Architecture, Wren had written that ‘Architecture has its political Use; public Buildings being the Ornament of a Country; it establishes a Nation, draws People and Commerce; makes the people love their native Country, which Passion is the Original of all great Actions in a Common-wealth’. The architects of Augustodunum would surely have agreed.

Figure 28. Christopher Wren’s plans for the redesign of central London, 1666.

Wren’s new ‘Roman’ London was never built, but architects continued to toy with the idea of remodelling England’s great cities on a Roman pattern. In around 1712 Wren’s pupil Nicholas Hawksmoor (1661–1736) drew up a grand plan for the redesign of central Oxford on the model of a Roman city. Hawksmoor was an unusually ambitious and eclectic architect (the door-lintels on many of his London churches were influenced by the architecture of the old Persian capital of Persepolis), but his plans for Oxford are still breathtaking. The new city would be entered from the east through a grand processional gate (at the end of modern Longwall Street); the centre of Oxford would be laid out around a civic forum (forum civitatis) and a University forum (forum universitatis). He even sketched out a rough design for a new University Church, just off the University forum, in the form of a huge peristyle temple, based on the Roman temple of Bacchus at Baalbek in Lebanon. As with Wren’s designs for London, and as at Augustodunum, Hawksmoor’s vision for Oxford was designed to serve a specific political function: to bridge the traditional gulf between town and university through the creation of grand new public spaces, both civic and academic. The only element of Hawksmoor’s design ever to be completed was his fine Clarendon Building, on Oxford’s Broad Street.

The size of the provincial cities varied widely. In AD 6 the census of Quirinius had come up with a total of 117,000 citizen men, women and children for the population of the city and territory of Apamea in northern Syria – one of the bigger cities of the empire, with an urban area of around 250 hectares. Alexandria, the largest city of the eastern provinces, had more than half a million inhabitants, and the population of Rome itself may have topped a million in the early imperial period. But these are exceptional cases. The population of the average Roman city should certainly be counted in the tens, not the hundreds, of thousands. Pompeii, on the bay of Naples, had an urban population of around 12,000, with maybe twice that number living in the surrounding countryside. Many were smaller still; indeed, physically there was little to distinguish a large village from a small city. The inhabitants of the Italian city of Rudiae, the birthplace of the poet Ennius, numbered between 2,000 and 2,500. By contrast, the village of Umm el-Jimal in north-west Jordan – its ancient name is unknown – had well over 2,500 inhabitants throughout much of its history, without ever receiving city-status. The crucial difference between Rudiae and Umm el-Jimal was not their physical size, but their administrative function. Similarly, in the modern United Kingdom, St David’s in Pembrokeshire (population 1,797, plus cathedral) is a chartered city; the town of Reading (population 230,000, with no cathedral) is not.

The status of these small communities could change over time. In the mid-second century AD, the village of Pallantion in the central Peloponnese successfully appealed to the emperor Antoninus Pius (138–61) to be elevated to city-status, on the grounds – so they argued – that the legendary Arcadian hero Evander had named the Palatine hill in Rome after his native town of Pallantion. For Pallantion, being a bona fide city was clearly a matter of local pride. The financial benefits may also have been considerable; the emperor was also persuaded to grant them blanket immunity from taxation.

Map 30. The western Roman empire in the second century AD.

As we saw at the end of Chapter 5, the historian Polybius understood the history of the Mediterranean and Near East in terms of a succession of empires: Persian, Spartan, Macedonian and at last Roman. But the Roman empire differed in several significant ways from the imperial states that had preceded it in the eastern Mediterranean and Near East. For better or worse, two centuries of Persian rule left almost no impact on the culture of the millions of subject peoples in central and western Asia. When Alexander the Great entered Babylon in October 331 BC, he found a city which in terms of administration, material culture, language and religion, was more or less as it had been when the city fell to Cyrus in 539 BC. The Roman empire of the late Republican period, like the Persian empire, was an incoherent mosaic of different cultures, unified by little more than the common experience of Roman political domination. Yet in the course of little more than a century every piece of that mosaic was transformed by the experience of Roman rule. The process of transformation took very different forms in the western and eastern provinces. We shall look first at the western provinces (North Africa, Iberia, Europe north of the Alps), before turning to the eastern or ‘Greek’ provinces of the Balkan peninsula, Asia Minor, the Levant and Egypt.

By the end of the first century AD, the material world of the western provinces, their architecture, clothing, food and bric-a-brac, had undergone revolutionary changes. We have already seen how the urban landscape of communities like the Aedui was reshaped beyond recognition under Roman rule. Throughout the western half of the Roman empire, similar changes occurred in almost every sphere of life, from language to religious practice, from drinking habits to personal names. The process is generally known as Romanization, ‘becoming Roman’.

The concept of ‘Romanization’ in the western provinces needs to be used with great care. Some modern scholars wish to jettison the term altogether; this is unnecessary, so long as we are clear about exactly who is doing the ‘Romanizing’. In what is probably the most frequently quoted passage in all of Tacitus’ work, we are told that during his governorship of Britain, Agricola

gave private encouragement and official assistance to the building of temples, forums and villas … he educated the sons of the native leaders in the liberal arts, so that those who had a short time ago spurned the Latin language altogether now strove to speak it with eloquence. Our national dress, the toga, became prestigious and fashionable; and little by little the Britons were led towards the amenities that make vice agreeable, arcades, baths and sumptuous banquets. Among the unsuspecting Britons these things were given the name of civilization, when in fact they merely added to their enslavement …

– a characteristically sharp Tacitean coda. This idea of a deliberate ‘civilizing mission’, whether undertaken from honest or cynical motives, has had a long and unfortunate influence on modern studies of the process by which the provinces ‘became Roman’. It is undeniably the case that temples, forums and villas started to be built in ever-increasing numbers in Britain, that Latin was spoken ever more widely, and that the sons of the native elite were educated in Roman ways: by the reign of Tiberius at the latest, the pastiche Roman town of Augustodunum in central Gaul was equipped with a school, where, says Tacitus, the elite youth of the Gallic provinces could receive a decent liberal – that is to say, Greek and Latin – education.

Agricola and India

Tacitus’ description in the Agricola of the ‘civilizing’ influence of Rome on the native British elite had a profound impact on nineteenth-century British conceptions of empire. The 1830s saw a major debate over the best mode of education for Britain’s Indian subjects, and in particular, whether it was appropriate for Indians to continue to be educated in their native languages, or whether all higher education should be conducted in English. Thomas Babington Macaulay, the later author of the Lays of Ancient Rome and the History of England, argued strongly in favour of an English-language educational system, for reasons that Agricola would have found very familiar. ‘We must do our best to form a class who may be interpreters between us and the millions whom we govern; a class of persons, Indian in blood and colour, but English in taste, in opinions, in morals, and in intellect.’ Macaulay is here directly transposing the Tacitean idea of cultural Romanization into an Indian context. In 1838 Charles Trevelyan optimistically predicted that Indians educated in the English style would ‘become more English than Hindus, just as the Roman provincials became more Romans than Gauls or Italians … The Indians will, I hope, soon stand in the same position towards us in which we once stood towards the Romans.’

By the late Victorian and Edwardian period, that optimism was starting to give way. The assimilation of the peoples of India into the British empire had proved harder than expected. Moreover, archaeologists of Roman Britain had now abandoned the idea of a small, highly Romanized native ruling class acting as local ‘partners in empire’. Instead, Francis Haverfield (1860–1919), the pre-eminent Romano-British archaeologist of the Edwardian era, argued that the Romanization of the native peoples of Britain was deep and far-reaching: throughout the province, ‘the material civilisation [and] the external fabric of its life was Roman, and the native element almost wholly succumbed to the foreign conquering influence’. By those standards, the ‘Anglicization’ of India had been an utter failure.

Lord Dufferin, Viceroy of India 1884–8, took a more pragmatic view of the relevance of the ancient world to modern imperial government. The Greeks and Romans were ‘people who didn’t talk our tongue and who were very strong on sacrifice and ritual, particularly at meals, whose gods were different from ours and who had strict views on the disposal of the dead. Well, you know, all that is worth knowing if you ever have to govern India.’

However, most historians now believe that Agricola – if he did indeed have a policy of active ‘Romanization’ – was wholly exceptional. For the most part, the conquerors displayed no missionary zeal to convert their uncouth western subjects to the joys of Roman culture. On the contrary, Tacitus in his Germania, an ethnographic study of the German peoples (composed shortly after the Agricola), showed himself to be an admirer of many aspects of native German society. As we have seen, the new Roman town at Augustodunum was the result of the Gauls’ own enthusiasm for the trappings of civilized city life, rather than any top-down initiative by the central government. Put simply, the political dominance of Rome made Roman things fashionable. People aspired to be and look Roman, since Roman-ness was associated with power. (We need only compare the overwhelming influence exerted by the material culture – denim, Coca-Cola, basketball – of the only modern superpower, the United States, on the dress, diet and behaviour of the rest of the world.) Moreover, the Roman state was always willing to assimilate people who looked, spoke and behaved like good Romans. Hence there were strong incentives favouring the ‘self-Romanization’ of the western provincial elites.

Figure 29. Fishbourne Roman palace.

The speed with which native elites adopted elements of Roman culture is startling. In the late 70s AD, within a generation of the Roman conquest of Britain, a huge palatial villa was constructed at Fishbourne, near Chichester on the Sussex coast (see Figure 29). This monumental complex, built of stone masonry, consists of four wings surrounding a large formal garden; the rooms were decorated with gaudy wall-paintings, stucco mouldings and fine floor-mosaics. The Fishbourne palace is entirely Mediterranean in layout and decoration; indeed, the labour force may well have been brought over from Italy. Yet the owner of the palace was almost certainly a Romanized Briton, Togidubnus. This man was a member of the pre-conquest aristocracy who had ruled southern Britain as a Roman client king in the years immediately following the conquest. In an inscription from Chichester, his full name and title are given: ‘Tiberius Claudius Togidubnus, great king of the Britons’. The interesting thing here is the presence of the Romanpraenomenand nomen (first and family name), ‘Tiberius Claudius Togidubnus’. Evidently this local chieftain had been granted Roman citizenship as a reward for his compliance at the time of the original conquest. Togidubnus needed no support or encouragement from Agricola to build an Italian-style villa on the Sussex coast: it was all too clear to men like Togidubnus that in order to retain their elite status in the post-conquest world, they needed to become as Roman as possible, as quickly and visibly as possible.

Similar patterns can be seen in more modest contexts. No type of object was more ubiquitous in ancient Europe than pottery, clay vessels used for the storage, preparation and consumption of food and drink. Pottery is thus an unusually good index of cultural change. In some cases, changes in pottery types can be used to map changes in consumption. For example, in the late first century BC and first century AD an Italian type of shallow baking dish made of a rather coarse red pottery (‘Pompeian Red Ware’) spread rapidly throughout the western provinces, from Spain to Britain. The widespread adoption of these new baking dishes reflected a far-reaching change in western European cookery. The main source of carbohydrates in pre-Roman north-west Europe was porridge, washed down with the main local beverage, beer. However, no self-respecting Roman would dream of boiling grain for porridge; the civilized person ate bread baked in an oven, not porridge boiled in a pot. Hence the sudden appearance of new Pompeian-style baking dishes across the north-west probably marked a mass switch from porridge to bread as the main source of carbohydrates for the aspiring elites of Spain, Gaul and Britain. Drinking habits, too, changed. In the sixth and fifth centuries BC the elites of the West Hallstatt zone north of the Alps had paraded their elite credentials through the consumption of Mediterranean wine; in the early first century BC, the Greek historian Posidonius had noted that the richer European Celts drank wine imported from Italy or Massalia, while the poorer classes drank either wheat beer mixed with honey, or plain barley beer. Over the course of the first and second centuries AD, beer-consumption declined sharply across western Europe; by the end of the first century AD, more than half of the wine being drunk in the Besançon region was produced in Gallic vineyards.

New foods required new dining-sets. Since the mid-first century BC, the standard Roman mass-produced tableware had been a fine glossy red pottery with elaborate rococo relief decoration, mould-cast rather than thrown, known as Arretine ware, after its main centre of production at Arretium (modern Arezzo, in Tuscany). This style of pottery was wildly popular throughout both the Mediterranean world and north-western Europe. Around the turn of the millennium, the market for Arretine ware in Gaul had reached such proportions that independent workshops were set up in southern Gaul to produce imitation ‘Arretine-style’ pottery to serve local demand. The best known of these local workshops was located at La Graufesenque, near Millau in southern France. The scale of production at La Graufesenque in the latter half of the first century AD is simply staggering: tablewares produced at this single workshop are found not only throughout the Gallic provinces, but as far afield as southern Britain and North Africa. For a generation or more, one of the marks of a civilized Gallo-Roman household was a dinner service of cheap Roman-style Arretine ware from La Graufesenque.

We know a great deal about the mechanics of pottery production at La Graufesenque. When a batch of pots was ready to be fired, a list of the craftsmen responsible for the pots in each batch was scratched onto a plate, which was then fired along with the finished objects and stored away in the factory archives. More than 160 of these lists survive, written in a baffling mixture of Latin and the native Celtic language, showing that the workmen at La Graufesenque were used to switching backwards and forwards between the two languages. Some of the craftsmen on the batch-lists still possessed traditional Celtic names (Cintusmos, Petrecos, Matugenos), though others had already adopted good Roman names (Cornutus, Secundus, Albinus). Each individual finished pot, too, was stamped with the name of the particular potter who had made it – but here something curious starts to happen. Not one of the Celtic names from the batch-lists reappears among the makers’ stamps on the pots themselves. Instead, the makers’ stamps all offer Latin equivalents for the potters’ Celtic names: Cintusmos reappears as Primus, Petrecos as Quartus and Matugenos as Felix. When a consumer bought his dining-set from La Graufesenque, all he would find on his pots was the respectable Latin name ‘Felix’; he would have no way of telling that ‘Felix’ was in fact a bilingual Celt whose real name was Matugenos. The potters at La Graufesenque were, in fact, pretending to be more Romanized than they actually were.

The case of the mock-Roman potters of La Graufesenque is a chastening lesson in the difficulty of judging how far the various local languages of the empire, both east and west, survived the Roman conquest. Officially, the business of the Roman empire was conducted in two languages only: Latin in the western provinces and North Africa (except Egypt), Greek in Egypt and the east. Even in the east, documents of critical importance for the Roman administration – birth certificates, which provided evidence of Roman citizenship, and wills – had to be in Latin, at least until the early third century AD. At first sight, the vast mass of documentary evidence surviving from the period of the Roman empire (inscriptions on stone, papyri, writing tablets, pottery stamps), supports this picture of a linguistically Romanized empire, operating exclusively in Latin and Greek. From the Asia Minor peninsula alone, we have tens of thousands of inscriptions on stone dating to the first three centuries AD. The overwhelming majority are written in Greek, with a much smaller, but still substantial number in Latin. Most of the various native languages of Asia Minor – Lycian, Lydian, Galatian, Carian – are entirely absent. Only in the deep countryside does the Phrygian language appear, on a few dozen bilingual tombstones of the third century AD; we also have a tiny handful of Pisidian texts. To judge from the documentary evidence alone, we would certainly have concluded that the local languages of Asia Minor had died with the Roman conquest, if not before. It is something of a shock, then, to read in the New Testament Acts of the Apostles that when, in the mid-first century AD, the apostles Paul and Barnabas arrived at the small Roman colony of Lystra, the local people hailed them as gods ‘in the Lycaonian language’. Not one single word of the Lycaonian tongue has come down to us. Evidently, at both La Graufesenque and Lystra, there was a sharp divide between the languages of administration and public affairs (Latin and Greek respectively) and the languages that people actually spoke in their daily life (Celtic and Lycaonian).

The Latin language

More people speak the Latin language today than ever before. Latin is the native language of some 700 million people, including almost all the inhabitants of South America and western Europe. It is true that most people now give their local dialect of the Latin language a different name (‘Italiano’, ‘Español’, ‘Français’, ‘Português’, ‘Occitan’); in English, we generally call this family of languages the ‘Romance’ language-group. None of this should blind us to the fact that Spanish, Italian and the rest are simply modern dialects of Latin. It would be quite reasonable to call Latin ‘Ancient Spanish’.

Here, once again, is what Tacitus has to say about the Romanization of Britain: ‘Our national dress, the toga, became prestigious and fashionable; and little by little the Britons were led towards the amenities that make vice agreeable, arcades, baths and sumptuous banquets. Among the unsuspecting Britons these things were given the name of civilization, when in fact they merely added to their enslavement.’

Here is the same passage in Latin:

inde etiam habitus nostri honor et frequens toga; paulatimque discessum ad delenimenta vitiorum, porticus et balineas et conviviorum elegantiam. idque apud imperitos humanitas vocabatur, cum pars servitutis esset.

And here is the same passage again, in a modern Mexican Spanish translation:

Desde entonces, también nuestros hábitos fueron un honor, e frecuente la toga; y paulatinamente se cayó en la seducción de los vicios: los pórticos y los balnearios y la elegancia de los banquetes; y eso era llamado humanidad entre los imperitos, cuando era parte de servidumbre.

habitus nostri, nuestros hábitos (‘our (national) dress/habit’); paulatim, paulatinamente (‘little by little’); pars servitutis, parte de servidumbre (‘part of their servitude’). The spelling and grammar have undergone a few changes over the past 2,000 years, but the language is recognizably the same.

Perhaps the most striking case of cultural forgetfulness in the western provinces of the Roman empire is that of the Punic societies of coastal Iberia and North Africa. Unlike the decentralized, pre-literate societies of Iron Age Europe, the Phoenician colonial world of the southern and western Mediterranean was a highly complex city-state culture, with its own flourishing historical and literary traditions. At the time of the destruction of Carthage in 146 BC, entire libraries of Punic literature were broken up and handed over to the petty princes of North Africa; a Punic-language treatise on agriculture in 28 books, written by a certain Mago, was brought to Italy to be translated into Latin. The Punic language was in fact remarkably resilient under the Roman empire: long inscriptions in Punic are found down to the end of the first century AD, and shorter texts continue to appear well into the third century. Punic was still widely spoken in the late fourth and early fifth centuries AD, not only in the North African countryside, but even among St Augustine’s urban congregation at the town of Hippo. Nonetheless, memories of the pre-Roman past were obliterated just as effectively in Punic Africa as they had been in Gaul. We cannot write a history of the Phoenicians in the west before 146 BC; no one considered the history of the Punic world before the coming of Rome to be worth remembering.

Indeed, some of the inhabitants of the old west-Phoenician world positively celebrated the end of Punic history. A century after the Romans had razed Carthage to the ground, Julius Caesar established a new colony of Roman citizens on the site of the old Punic city, calling this too ‘Carthage’, which eventually developed into one of the largest and most prosperous cities of the western empire. In the second or third century AD, an imperial official by the name of Classicius Secundinus claimed to have found an ancient inscription at Carthage carrying the original victory dedication of Scipio Aemilianus, the Roman general who had destroyed Carthage in 146 BC. Secundinus had the dedication reinscribed and set up once again in its original form. In the third century AD the citizens of Roman Carthage could still muse on the words of Aemilianus, as he gloated over the ruins of Punic Carthage and the bodies of those who had dared to stand against Roman power. This was the only ‘version’ of Carthaginian history available to the inhabitants of Carthage under the Roman empire.

The impact of Roman rule on the religious life of the western provinces is particularly difficult to judge. Before the Roman conquest, the Celtic tribes of northern Gaul seem each to have worshipped their own, highly local gods: the god Mullo, for instance, was worshipped only in a triangle north of the river Loire, between Rennes, Nantes and Le Mans. Most of these local gods survived the Roman conquest through being identified with one of a very limited range of Roman deities (particularly Mars and Mercury). For example, the principal god of the tribal group of the Remi, in the Champagne region of north-east France, was a figure called Camulus. By the first century AD Camulus was being interpreted as the local manifestation of the Roman god Mars. In his new guise as ‘Mars Camulus’, the god continued to be the main patron deity of the Remi and their new urban centre at Reims (Durocortorum).

It is at first sight tempting to regard the survival of native deities like Camulus as a sign of deep-seated resistance to the process of Romanization: although the Remi were now obediently shaving their beards, baking bread, drinking wine and learning Latin, they remained Camulus-worshipping Celts at heart. The trouble with this view is that it fails to take into account the profound degree to which the Remi identified their interests with those of the ruling power. One of the four great triumphal arches leading to the monumental city centre at Reims, the Mars Gate, survives in relatively good shape. The central arcade of this arch carries an image of Mars Camulus; the east arcade has a depiction of the she-wolf suckling the twins Romulus and Remus (see Figure 30); and the west arcade shows Leda and the swan. The imagery is at first sight rather puzzling. Both of the side arcades seem to be depicting episodes from the foundation myths, not of the Celtic Remi, but of the city of Rome – since the offspring of Leda and the swan was Helen, who was responsible for the Trojan War and thus, indirectly, Aeneas’ foundation of Rome. What interest did the Remi have in the myths of the founding of Rome? It has been plausibly suggested that these ‘Roman’ images on the Mars Gate at Reims represent a claim by the Remi to be the ultimate descendants of Romulus’ brother Remus. This would of course make the Remi privileged kinsmen, even ‘brothers’, of the Roman people, a game which we have seen played repeatedly in the course of this book. This is a strikingly creative response to the fact of Roman rule: while preserving the key elements of their traditional religious practices (the cult of Camulus), the Remi also developed a version of their own mythical origins which linked them to the broader power of the Roman empire.

Figure 30. Sculptural decoration on the ceiling of the east arcade of the Mars Gate at Reims: Romulus and Remus being suckled by a she-wolf, surrounded by a frieze of shields, helmets, armour and weapons.

Clearly, the mere survival of elements of indigenous, pre-Roman culture cannot automatically be taken as evidence for positive cultural resistance to Rome. What, for instance, are we to make of the following Latin inscription, discovered at Southwark in south London in 2002?

Num(inibus) Aug(ustorum)	To the divine will of the emperors
Deo Marti Camulo	And to the god Mars Camulus:
Tiberinius	Tiberinius
Celerianus	Celerianus
c(ivis) Bell(ouacus)	citizen of Beauvais
moritix Londiniensium.	seafarer of the Londoners.

The inscription, which dates to the late second century AD, takes the form of a dedication to the ‘divine will of the emperors’ and the god Mars Camulus. The dedication was set up by a man carrying a good Roman name: Tiberinius Celerianus, a native of Beauvais (ancient Caesaromagus Bellovacorum) in northern Gaul. Celerianus describes himself as ‘seafarer of the Londoners’ (moritix Londiniensium), and we should probably understand him to be the agent of a shipping company which transported goods between London and northern Gaul. It is very striking that Celerianus chose to define himself with the curious term moritix. Moritix is not a Latin word at all, but an ancient Celtic term meaning ‘seafarer’. There is, of course, a perfectly good Latin word meaning exactly the same thing (nauta). Why did Celerianus chose to use the old Celtic word? Was he trying, consciously or unconsciously, to emphasize his local Celtic identity? The real cultural affiliations of a man like Celerianus are desperately difficult to recover: a native of northern Gaul, with a Roman name; worshipper of a superficially Romanized Celtic deity of his native region, but also of the reigning Roman emperors; capable of setting up a dedicatory inscription in impeccable Latin, but opting for a local Celtic term to describe his profession of seafarer.

Map 31. The eastern Roman empire in the second century AD.

In the western provinces, as we have seen, there was virtually no institutional memory of the pre-Roman past. The history of native societies before the conquest was largely forgotten, to be replaced by a new, more acceptably ‘Roman’ past; the local mythology of the Remi in north-east Gaul seems to have been imported wholesale from Rome on the basis of a chance similarity between the names ‘Remi’ and ‘Remus’. Local languages went into a precipitous decline; even local drinking and dining practices were obliterated by the spread of ‘Romanizing’ pottery and crops, above all the vine. When we turn to the eastern half of the Roman empire, however, the picture looks very different indeed. The spread of Roman cultural artefacts (villas, foodstuffs, Roman personal names, the Latin language) was comparatively restricted. Not only did the Greek language survive, but Greek memories of the pre-Roman past remained as vivid and culturally potent as ever. ‘Romanization’ is clearly not an appropriate way of thinking about the development of the eastern provinces in the first three centuries AD. Nonetheless, the impact of Roman rule on the Greek-speaking half of the empire ought not to be underestimated, for all that it is less immediately visible than in the western provinces.

By the first century AD, the Greek language had evolved a long way from the language spoken and written in the Classical Greek world. Several cases and moods, notably the optative mood, had all but disappeared. The pronunciation of Greek, too, had changed radically. The letter beta was now pronounced as v rather than b, and the word kai (‘and’) was normally pronounced as ke (as in modern Greek). However, over the course of the first and second centuries AD Greek writers and intellectuals increasingly came to reject this ‘common’ Greek language (koinē) of their own day as vulgar and unsuited for literary production. Instead, they reverted to the ‘pure’ Athenian dialect of the fifth and fourth centuries BC. Huge dictionaries, grammars and handbooks of the Classical Athenian language were compiled to help writers and orators to avoid inappropriate modernisms. The medical writer Galen (129–216) complained at the ‘pestilential pseudo-erudition’ of writers who use the obsolete Athenian word rhaphanos for cabbage, ‘as though we were in conversation with Athenians of 600 years ago’. An older contemporary of Galen, Lucius Flavius Arrian of Nicomedia, composed a seven-book history of Alexander the Great entirely in the style of Xenophon, an author more remote from Arrian’s own day than Shakespeare is from ours. In one case, a speech attributed in our medieval manuscripts to the second-century AD writer Herodes Atticus, the imitation of classical Athenian prose is so uncannily accurate that some modern scholars have argued that the speech is a genuine Athenian product of the fifth century BC – the ultimate compliment to Herodes’ skill.

This culture of ‘classicism’ eventually came to infect all areas of Greek cultural life. Members of the Greek civic elite – even, and perhaps especially, those who had been granted the privilege of Roman citizenship – named their children after famous Athenians of the Classical period: the emperor Trajan’s court doctor, Titus Statilius Crito, carried the name of a friend of the fifth-century BC philosopher Socrates, and his grandson, Titus Statilius Solon, was named after an Athenian lawgiver of the early sixth century BC. Sparta, by now a quiet olive-growing country town, attracted ever-increasing crowds of cultural tourists. The main draw was the opportunity to observe at first hand the legendary Spartan educational system, the agogē, supposedly established by Lycurgus in the ninth or eighth century BC. In the late second century AD we hear of a Spartan civic official with the unique title of ‘expounder of the Lycurgan customs’; according to one plausible explanation, his job was to act as a professional, full-time tourist guide. Greek and Roman tourists came to watch teenage Spartans practising military drills, wrestling and boxing, playing a rough ball-game called sphairistikē (apparently similar to rugby), and participating in a bloody annual ‘contest of endurance’ at the festival of Artemis Orthia, in which ephebes (older teenagers) had to try to reach an altar protected by whip-bearers – fatalities were apparently not uncommon. In fact, many of these violent games were inventions of the later first century AD, developed to satisfy non-Spartan expectations of what the Lycurgan education ought to have been like.

Just as with the survival of native deities in Roman Gaul, it is a delicate question whether this obsessive Greek focus on the distant past reflects a deep-seated cultural resistance to Rome. It is certainly the case that the cities of the Greek world preserved a far stronger sense of their own local identity than the communities of the western provinces. In the mid-third century AD Goths from the Black Sea region launched a series of devastating attacks on the Roman provinces in Asia Minor and the Balkans. Athens itself was sacked in 267/8, the first time that the city had been assaulted since the Mithradatic wars of the early first century BC. Nonetheless, under the leadership of an ageing historian and intellectual by the name of Herennius Dexippos, the Athenians assembled a scratch force of 2,000 men in a remote rural district of Attica, which succeeded in ambushing and destroying the greater part of the Gothic raiding band. It is tempting to explain the remarkable Athenian popular resistance of 267/8 through the deep emotional and cultural ties which bound them to the heroic Athenian past. The survival of a strong Athenian civic identity, still focused around the great days of Marathon and Plataea, may well have contributed to the Athenians’ ability to mount an effective civic defence against the Gothic invasion. A telling contrast could be drawn with the cities of the Roman west, very few of which managed any such resistance to the (ever more frequent) barbarian invasions of the third and fourth centuries.

The world which the Greeks of the second century AD harked back to was, above all, the pre-Roman – and, indeed, pre-Macedonian – Greek world of the fifth and fourth centuries BC. It is hardly surprising that the problematic Hellenistic age, an era of uneasy collaboration and, at times, violent resistance to Roman imperialism, was quietly erased from Greek history. From the perspective of the Roman ruling power, the Greeks’ obsessive nostalgia for a safely distant past was something to be encouraged. Athens, in particular, was ostentatiously favoured by Roman emperors as a cultural centre of the east Roman world, and by the mid-second century AD, the city had attained a level of wealth and prosperity not seen since the days of the Athenian empire in the fifth centuryBC. In the early 130s the emperor Hadrian made Athens the centre of a new religious confederation of bona fide Greek cities, the Panhellenion (‘All-Greeks’). The Panhellenion was a largely ceremonial body, and its actual functions are very obscure; the interesting thing from our perspective is the definition of ‘Greekness’ which Hadrian used in vetting potential members. The Spartans, of course, were admitted without hesitation. So too were cities which claimed to be descendants or colonists of the old Greek cities of the mainland: Magnesia on the Maeander, a small town near the west coast of Asia Minor, was waved through on the basis that they were ‘colonists of the Magnesians in Thessaly, the first of the Hellenes to cross over into Asia and settle there’. In 135 the town of Ptolemais-Barca in Libya petitioned the emperor Hadrian for membership of the Panhellenic club. The Barcaeans had high hopes of success, since everyone agreed that Barca was an ancient colony of Cyrene, a city which already sent two annual delegates to the Panhellenion. However, the emperor pointed out that Barca had been renamed ‘Ptolemais’ by one of the Macedonian kings of Egypt (probably Ptolemy II Philadelphos, reigned 282–246 BC). It was quite unreasonable for the Barcaeans to claim the same status as their neighbours at Cyrene, ‘whose ancestry is pure Greek, and specifically Dorian’; by way of compromise, the Barcaeans were permitted to send one delegate per year.

Some curious birds found their way into Hadrian’s Panhellenic nest. The obscure city of Eumeneia, lying high in the mountains of central Asia Minor, was not at first sight an obvious candidate for membership of the Panhellenion. The history of Eumeneia was short and not very eventful. Like Ptolemais-Barca, the town took its name from a Hellenistic monarch, the Attalid king Eumenes II of Pergamum, who had founded Eumeneia in the 160s BC as one of a string of military forts along the border with Galatian territory. But by the second century AD the Eumeneians had mysteriously acquired a past as respectable as any city of mainland Greece. They were, so they claimed, the descendants of colonists from Argos in the Peloponnese, who had migrated to Asia Minor late in the heroic age. The proof of their Argive ancestry lay in the town’s name. During the exile of the children of Heracles (Heracleidae) from their ancestral home in the Peloponnese, one of Heracles’ sons, Hyllos, had travelled to Eumeneia, where he had been ‘happy to stay’ (in Greek: eu menein). The Eumeneians were duly admitted to the Panhellenion on the basis of their Argive ancestry, and henceforth proudly declared themselves to be the ‘Eumeneian Achaeans’ on their inscriptions and coinage – no doubt to the extreme irritation of the citizens of Ptolemais-Barca.

We may well conclude from this that whoever vetted applications for the Panhellenion ought to have been sacked. But the episode is nonetheless revealing. There were good practical reasons why Hadrian should have wanted to support the Eumeneians’ flimsy claims to a place in the history of the heroic age: Eumeneia was home to the main Roman military garrison in the province of Asia, and as such held a key position in the Roman administration of the province. In such cases, where Rome’s favoured Greek allies lacked an appropriately distinguished history, a history was simply invented. The city of Aphrodisias, also located in western Asia Minor, had been founded at around the same time as Eumeneia. Thanks to her conspicuous loyalty to the Roman cause from the first century BC onwards, Aphrodisias held the privileged status of a ‘free city’, and was eventually made the capital (metropolis, literally ‘mother-city’) of the late Roman province of Caria. As one of Rome’s staunchest allies in Asia Minor, the city required a correspondingly impressive pedigree. So, in the late first century AD a new Roman civic basilica at Aphrodisias was decorated with relief sculptures depicting the newly discovered mythological founders of the city; in pride of place stood an image of the Greek hero Bellerophon with his winged horse Pegasus, consulting the oracle of Apollo at Delphi, and presumably receiving divine sanction for the foundation of Aphrodisias. The ‘discovery’ of a pre-Trojan War origin for Aphrodisias allowed the Roman emperors to carry on favouring her above her neighbours with a clear conscience. The creation of a respectable early history for Eumeneia and Aphrodisias is closely reminiscent of the invention of Messenian history at the time of the foundation of the polis of Messene in the mid-fourth century BC (above, pp. 134–6).

It is clear that the eastern provinces of the Roman empire did not become ‘Romanized’ to anything like the same extent as the western provinces. In the west, becoming Roman involved a wholesale obliteration of local languages, local history and pre-Roman forms of settlement. By contrast, the dominant cultural force in the eastern provinces was classicism, conformity to a homogeneous and unthreatening model of Greekness. This culture of classicism, safely focused on the distant past, represents a different kind of response to the reality of Roman dominance.

‘Romanization’ in the west and ‘classicism’ in the east were only in very limited ways conscious articles of policy of the Roman government. Far more important, in both cases, were the needs and aspirations of the provincials themselves. Certainly there was resistance to these processes too, not least among the selfsame provincial elites. But the interesting question is not so much why the subject peoples of the empire resisted Roman rule, as why, by and large, there was so little open resistance. Native rebellions, like that of Boudica in AD 60, were rare, and usually fizzled out within a generation of the initial Roman conquest.

The question becomes particularly pressing when we consider the one subject people who did try repeatedly to throw off the Roman yoke: the Jews. In AD 66 the Jewish inhabitants of Jerusalem rose up in open revolt and massacred the city’s Roman garrison. The whole of Judaea soon followed suit, and the entire region saw mass ethnic cleansing of the Gentile population. A Roman force under the governor of Syria, which came within an ace of recapturing Jerusalem from the rebels, was all but wiped out by Jewish partisans in the autumn of 66. The Roman response was brutal. After a five-month siege, led by the future emperor Titus, Jerusalem fell in September AD 70. The entire population of the city was killed or enslaved, and the great Temple at Jerusalem, the centre of the Jewish faith, was burned to the ground. The second revolt in Judaea, in 132, seems to have been sparked by Hadrian’s attempt to found a new Roman colony, Aelia Capitolina, on the site of Jerusalem – just as Antiochus IV’s attempt to turn Jerusalem into aGreekcolony had kicked off the great Jewish revolt against Seleucid rule in the second century BC (above, p. 157). For three and a half years, the Jews, under the charismatic leadership of Shimon bar Kokhba, struggled to recapture Jerusalem. By the time the rebels were finally crushed, in the autumn of 135, Judaea was a wasteland; the Roman historian Cassius Dio claims that more than half a million Jews had been put to the sword.

Why was resistance to Roman rule in Judaea so fierce? Josephus, our main source for the first Jewish revolt, argues that the uprising was the work of a few bad apples on both sides. Roman governors in Judaea had behaved in an unusually vicious and insensitive manner; on the Jewish side, the breach with Rome was caused by a few villainous brigands and religious fanatics. However, Josephus has an axe to grind. Himself a prominent member of the native ruling class of Judaea, he had fought on the Jewish side in the first year of the revolt, before being captured by Roman forces. He promptly switched sides, was granted Roman citizenship, and spent the rest of his life in Rome. In part, his history is an attempt to justify his own conduct in the revolt and afterwards. As an apologist for the Judaean upper classes – who were, in fact, profoundly implicated in the revolt – it suits him to present the revolt as the work of a few extremists; as an interpreter of the ‘Jewish situation’ for a Roman audience, he diplomatically places the blame for the revolt on individual wicked Romans, rather than the nature of Roman rule itself.

The problem with this view is that the revolts were not merely protests against Roman maltreatment, but aimed at the establishment of an independent, self-governing Jewish state centred on Jerusalem. On both occasions, the rebel Jewish state minted silver and bronze coinage with aggressively nationalist inscriptions in the Hebrew language: ‘JERUSALEM IS HOLY’, ‘FREEDOM OF ZION’, ‘FOR THE REDEMPTION OF ZION’. One group of coins, struck in the early days of the second revolt (132), carries an image of the lost Temple – destroyed sixty-two years earlier – and the proud inscription ‘YEAR ONE OF THE REDEMPTION OF ISRAEL’ (see Plate 31f).

Jewish resistance clearly had deeper religious roots than Josephus is prepared to admit. Before the first revolt, Jews throughout the empire had paid a voluntary contribution of two drachmas per year for the upkeep of the Temple at Jerusalem. After AD 70 this voluntary payment was replaced by a special Jewish tax, once again levied on all the Jews of the empire, to pay for the cult of Jupiter Capitolinus on the Capitol at Rome. The interesting point here is that Rome chose to respond to the revolt in Judaea with collective punishment of the Jewish race as a whole. The humiliating ‘Jewish tax’ was imposed not only on the Judaean rebels, but also on the large diaspora communities in Egypt, Asia Minor and Rome itself, which had taken no part in the revolt. All too clearly, Rome saw the uprising as a Jewish problem, not a Judaean problem. The Romans had a point: the Jews of the diaspora were no better integrated into the imperial system than the Jews of Judaea. In 116–17, Egypt, Cyrenaica and Cyprus were rocked by a sudden and violent mutiny of the Jewish minority community, this time directed against their provincial non-Jewish neighbours rather than the Roman government. The Roman response was, once again, draconian; henceforth, no Jew was allowed to set foot on the island of Cyprus, and even Jews whom bad weather forced to land at a Cypriot harbour were to be put to death.

The fundamental problem was that the kind of religious assimilation to Roman rule which happened elsewhere in the empire was impossible for the Jewish people. The Celts of northern Gaul were happy to identify their local deities, gods like Camulus and Mullo, with the Roman Mars or Jupiter. But the Jewish God was simply incompatible with any kind of polytheistic system. There was one true and omnipotent God; He was not the same as the Roman Jupiter. Nor was the line which separated Jew from Gentile open to negotiation. Observation of even the most basic elements of the Mosaic law – circumcision, dietary restrictions, keeping the Sabbath – sharply restricted the extent to which Jews could associate with Gentiles. Finally, subjection to Rome was not just humiliating; it was wrong, a fundamental breach of God’s promise to protect and deliver his chosen people from oppression.

Nonetheless, as we have already seen, the Roman empire was essentially a world of religious pluralism. The provinces of the empire were home to an astonishing number of local cults, most of which were tolerated by the Roman ruling power. Perhaps the most striking example of Roman religious toleration is that of the Mithras cult of the second and third centuries AD. The origins of the cult of Mithras are clouded in mystery. All we can say for certain is that very suddenly, at the end of the first and the beginning of the second century AD, Mithraic sacred buildings (‘mithraea’) and dedications to Mithras appear simultaneously in very widely dispersed parts of the empire (Judaea, the Rhine and Danube frontier, Rome, inland Asia Minor). The cult became widespread in the second and third centuries AD, particularly among the frontier garrisons of the Roman army; three mithraea have been found at forts on Hadrian’s Wall, at Housesteads, Rudchester and Carrawburgh. The Mithras cult was a mystery religion, open only to male initiates, each of whom belonged to one of seven grades, ‘raven’, ‘male bride’, ‘soldier’, ‘lion’, ‘Persian’, ‘sun-runner’, ‘father’, each associated with one of the seven planets (the five which are visible to the naked eye, Mercury, Venus, Mars, Jupiter and Saturn, plus the sun and moon). A Mithraic initiate passed through the seven grades in order, and as he moved from planet to planet, each lying further away from the earth, his soul too travelled further away from the material world, towards the final aim of salvation after death. A bronze plaque from Virunum (near modern Klagenfurt in Austria), lists all thirty-four members of one particular Mithraic congregation as it stood in 183. Over the following eighteen years the congregation was swelled by between one and eight new members per year, but only six of the ninety-seven men who belonged to this particular Mithraic group in the period 183–201 seem to have attained the highest and most prestigious grade of ‘father’.

It would be quite wrong to regard the Mithras cult as an ‘underground’ sect. Most of the adherents of the cult were soldiers, slaves and ex-slaves, but a few initiates came from the upper ranks of the army, including some of equestrian and senatorial status; some of the best-preserved Mithraic sacred buildings are those from the city of Rome itself. Nonetheless, the Mithras cult was radically different from other Graeco-Roman religions. Mithras himself, as represented in cult icons, was a visibly non-Roman figure, invariably being depicted wearing a ‘Phrygian’ cap, associated with the far eastern reaches of the empire and the Persian world beyond its borders. Mithraic cult practice, too, was wholly unlike Graeco-Roman ritual. Women were totally excluded; Mithraic worship took place in secluded, indoor locations (private houses, military camps, even caves); most striking of all, the central icon of the cult, located at one end of the Mithraic sanctuary, depicted the god Mithras slaying a bull. The idea of the god himself performing animal sacrifice is a deliberate and pointed reversal of the ordinary rules of Graeco-Roman cult practice, according to which animal sacrifice was performed by the worshipper in honour of the god. Finally, Mithras was, uniquely, a god whose home lay outside the bounds of the Roman empire, in a mythical cave somewhere in Persia. While Greek and Roman gods were almost always worshipped in their capacity as protectors or inhabitants of a specific place (the Capitoline Jupiter, Athena Polias on the Athenian Acropolis, the Ephesian Artemis), Mithras was a geographically floating deity, tied neither to a particular civic locale nor to the imperial centre.

The Mithras cult seems to have been especially popular among serving soldiers and slaves or ex-slaves, precisely those groups in the Roman empire who were most cut off from the civic world of Graeco-Roman religion. It is tempting to see the Mithras cult, with its rootlessness, its deliberate rejection of the cult practices of civic religion and its emphasis on internal hierarchies and self-advancement, as a response to the experience of military service on the imperial Roman frontier, the limes. For the men of the frontier garrisons, thousands of kilometres from the temples of their native cults and thrown together with soldiers drawn from all corners of the empire, the Mithras cult may have been a way of making sense of their common experience.

The imperial frontier was one of the most distinctive and novel features of Roman Europe. The limes took different forms in different regions. In the east and south, in Mesopotamia, Arabia and North Africa, the effective limits of the empire coincided with natural boundaries, at the point where cultivable territory shades into desert. In continental Europe, throughout the first and second centuries AD, the northern limit of the empire was conventionally marked by the course of the Rhine and Danube rivers: Tacitus speaks of an empire ‘fenced in by the Ocean and by great rivers’. In the second century AD parts of the frontier were still more clearly defined with the construction of physical barriers of earth, wood or stone; the best known example is of course Hadrian’s Wall in northern Britain, but large parts of the limes on the Danube and Rhine were similarly fortified.

It is tempting to think of the Roman empire of the first three centuries AD as a static, non-expansionist state, enclosed by linear, heavily defended frontiers. The first emperor, Augustus, is said to have advised his successor, Tiberius, ‘to keep the empire within its boundaries’. However, the actual behaviour of Roman emperors hardly supports this view: in AD 15, within a year of Augustus’ death, Tiberius himself authorized an advance further into Germany, and few emperors did not at least attempt to add new territory to the empire. It may be helpful to imagine Europe in the first three centuries AD as being divided into three distinct regions. The first and innermost region, the ‘core empire’, was made up of the provinces directly administered by Roman officials, whose inhabitants paid taxes to Rome and were subject to Roman law and jurisdiction. Beyond the core provinces lay a second zone of territory, an ‘inner periphery’ of subject territories (the Latin word is gentes, ‘tribes’), subject to Roman rule but not under direct Roman administration, controlled by client kings and native allies. The Roman emperor expected to be able to influence the succession of power from one client king to another; it was a rare privilege for King Herod of Judaea to possess the right of choosing his successor without reference to Augustus. This inner periphery separated the core provinces from the third region, the ‘outer tribes’ or gentes externae, who were not subject to Roman rule at all.

When we talk about the ‘frontiers’ of the Roman empire, then, we should really be thinking of the whole of the second zone: a broad, heavily militarized swath of territory, in places hundreds of kilometres in depth (in Britain, stretching from York to Newcastle upon Tyne). Actual lines of fortifications could stand either at the outer edge of this zone, as in the case of Hadrian’s Wall, or at its inner edge, as in the case of the Danube frontier in Hungary. The fortifications built by the emperor Commodus (180–92) along the banks of the Danube south of Aquincum (modern Budapest) are a particularly interesting example. These towers and forts were certainly not built in order to protect the frontier from barbarian attacks: they were in fact constructed immediately after a peace-treaty with the barbarian kingdom on the far side of the Danube, the Iazyges. As part of the peace-treaty, the Iazyges were permitted to enter the Roman province to visit market towns on appointed days. Consequently, as we are informed by a series of building-inscriptions, Commodus ‘fortified the whole river-bank with brand new towers and guard-posts, situated in opportune places for preventing the secret crossings of smugglers’. The Danube fortifications were needed precisely in order to regulate traffic between the province and the friendly barbarian kingdom outside it to the north. It is striking and appropriate that the only surviving map of the whole Roman empire, the Peutinger Table (a late medieval copy of a map of the fourth or fifth century AD), marks no frontiers at all; instead, roads continue smoothly beyond the nominal ‘borders’ deep into non-Roman territory.

The impact of the army on the societies of the frontier zones was immense. In second-century AD Britain, some 60,000 Roman soldiers were stationed among a total population of perhaps 4 million inhabitants, a ratio of one soldier to every sixty-five civilians. Britain was, of course, a heavily garrisoned frontier province, but even a peaceful ‘core’ province like Egypt, with a population comparable in size to that of Britain (4 to 5 million), may have had as many as 10,000 soldiers permanently stationed on its soil. In 238 the inhabitants of a village called Skaptopara in Thrace (modern Bulgaria) sent a desperate petition to the emperor Gordian III (238–44). Their village, they explained, was well situated: it possessed thermal springs with excellent therapeutic qualities, and once a year, a famous market-fair was held 3 kilometres from the village. Senior Roman officials, including the provincial governor, often came to stay at the village in order to enjoy the use of the thermal baths. (Delicate Italians often found the climate of the Balkans uncongenial: the poet Ovid, exiled to Tomis on the Black Sea in AD 8, suffered from indigestion, insomnia, fever and miscellaneous aches and pains.) However, Skaptopara also happened to lie just off the road connecting two military camps. Soldiers travelling between the two camps regularly descended on the village to use the hot springs; the villagers were required to house and provision them without compensation. As a consequence of this constant military traffic, the village, which had previously been prosperous, had been reduced to utter destitution. Attempts by the provincial governor to crack down on these abuses of power had made not the least difference, so that the villagers were now reduced to asking the emperor himself to step in. The emperor’s unhelpful response was simply to refer them back to the governor.

The army was clearly a highly visible and unwelcome presence in the daily life of Skaptopara. Nor was this an exceptional case. In the militarized zone immediately to the south of Hadrian’s Wall, in Co. Durham and North Yorkshire, no cities and very few rural villas have been located, and native sites have produced very few Roman coins and artefacts; most of the native population (the Brigantes) continued to live in traditional roundhouses, in increasingly impoverished conditions. To all appearances, the massive military presence in northern Britain had a parasitical effect on the people living in the frontier zone; native economic and cultural development simply ground to a halt.

A story in Apuleius’ second-century AD Latin novel, The Golden Ass, makes the point about the relation between soldier and civilian as well as any. Somewhere in the northern Greek countryside, a legionary centurion comes across a market gardener riding a donkey. The soldier accosts the gardener, and demands, in Latin, to know where he is taking his donkey. The gardener, not knowing any Latin, rides on in silence; the soldier takes this as an insult and lands a vicious blow on the man’s head with his staff. After striking the gardener to the ground, he asks the same question again, this time in Greek: where are you taking that donkey? The soldier is bilingual; he must have assumed that the gardener was a Greek-speaker. Nonetheless, he had chosen to address him in Latin, solely in order to have the opportunity to beat up a peasant for insulting behaviour. It is no surprise to read that the soldier goes on to demand that the gardener hand over his donkey for military use. That such arbitrary requisitions of locals’ labour and livestock were a daily occurrence in first-century AD Judaea is clear from Jesus’ advice in the Sermon on the Mount: ‘And whosoever shall compel thee to go a mile, go with him twain’ (Matthew 5: 41).

We have seen that the frontier ‘zone’ shaded off almost imperceptibly into the lands of the gentes externae. Among these neighbouring peoples, we can map a kind of ‘ripple effect’ of Roman cultural influence on their lifestyles and material culture. To the north-east, beyond the Rhine, Roman-style armour and weapons become steadily more popular over the course of the first three centuries AD. In the second century AD the first examples of the Runic alphabet, based directly on the Latin script, start to appear in Danish burials on weapons, tools and prestige objects. Most importantly, we start to see the emergence of local aristocratic dynasties, characterized by their ostentatious use of imported Roman luxury objects. At Himlingøje in eastern Denmark, the tombs of at least thirteen members of the same family have been uncovered, dating from the mid-second to the late third century AD. The grave goods from the Himlingøje burials are a mixture of high-quality Roman imports (bronze buckets and ladles, glass cups and drinking horns) and local prestige objects (gold finger- and arm-rings). It was through its Romanizing pretensions that the Himlingøje family chose to distinguish themselves from their lower-status neighbours.

Map 32. Rome’s eastern neighbours.

The most powerful of the gentes externae lay beyond the river Euphrates frontier to the east. Roman relations with the great civilizations of inner Asia were far more intense and complex than with the barbarian peoples of northern Europe, for the simple reason that most of the essential luxuries of the Roman overseas trade – Arabian incense, African ivory, Chinese silk and Indian pepper – came from beyond the eastern frontiers of the empire. Most important of all was the Roman seaborne trade between Egypt and southern India, channelled through two great Roman trading stations on the Red Sea coast of Egypt, Myos Hormos and Berenikē. Recent excavations at Berenikē have brought this trade to life for us for the first time. Enormous quantities of Indian black peppercorns have been excavated there, including 7.5 kilograms of peppercorns still packed into a huge Indian storage jar which for some reason was never shipped onwards to the Mediterranean. Rome’s Indian trade was dazzlingly profitable. Pliny the Elder, in his Natural History, claimed that the Indians received some 50 million Roman sesterces per year, in return for goods which were sold in the Mediterranean for a hundred times the price.

Pliny’s figures have been routinely dismissed by modern historians as an absurd exaggeration, but a recently published Egyptian papyrus has provided a spectacular vindication of Pliny’s statement. This document, dating to the mid-second century AD, records the transfer of a shipment of goods overland from the Red Sea coast of Egypt to the Nile valley. We learn that these goods had been shipped to the Red Sea from Muziris, in the south of India, on a single Roman merchantman called the Hermapollon. The journey of the Hermapollon can be reconstructed with some confidence: a thirty-day journey down the Red Sea to the gulf of Aden; twenty days across the Indian Ocean, with the monsoon winds at their backs; three to four months at Muziris, waiting for the wind to turn; and another fifty days back to Berenikē, making a round trip of almost eight months. The cargo carried on this ship on a single journey back from India to the Red Sea, including ivory, luxury fabric and spikenard (a perfumed oil) from the Ganges valley, is valued in the papyrus at around 7 million sesterces. To put that figure in context: 9 million sesterces would (on one estimate) have been enough to equip a city the size of Pompeii with all its necessary public buildings and amenities. The Hermapollonwas a true floating treasure-chest.

At the eastern end of this particular channel of trade, staggering numbers of Roman coins have been found in Sri Lanka and southern India, along with transport jars, and Roman bronze- and glass-ware. The Peutinger Table (a fourth- or fifth-century AD map of the Roman empire, mentioned above) marks a ‘temple of Augustus’ in the vicinity of Muziris, which may suggest the existence of a permanent Roman trading post in southern India. Rome’s connections in South Asia may have extended even further to the east: two Roman medallions, one of Antoninus Pius (138–61), the other of Marcus Aurelius (161–80), have been excavated at the trading post of Oc Eo, in the Mekong delta of South Vietnam, and there is some evidence that the builders of log-boats in Vietnam in the early first millennium AD were aware of Roman shipwrights’ techniques.

By contrast to its seaborne trade, Rome’s overland trade with its eastern neighbours is very poorly attested indeed. There is no doubt, for example, that Rome imported huge quantities of Chinese silk, the ultimate luxury fabric: the mistresses of the Augustan love poets shimmer around in translucent silk dresses, and fragments of silk clothing have been excavated as far north as Holborough in Kent. Whether this silk came to the Mediterranean by an overland ‘Silk Road’, however, is far more controversial. Evidence for direct Roman contact with China is very hard to come by. It was long believed that Roman glass was exported in large quantities to China. More recent research has shown that most of this glass was produced in China itself; sadly, one particularly spectacular example, a cast glass vase with engraved medallions of the goddess Athena, supposedly found in a Chinese tomb in Honan province, has now been recognized as the product of a nineteenth-century Bohemian glass factory. Similarly, a well-known hoard of sixteen Roman bronze coins from Ling-shih hsien in Shansi, ranging in date from Tiberius (AD 14–37) to Aurelian (270–75), has been shown to have been the private collection of a modern western missionary. Very sporadic Roman diplomatic contacts with China are known: Chinese sources record the arrival of a delegation from the emperor An-tun (Marcus Aurelius Antoninus) in AD 166, and in 226 a Roman merchant named ‘Ts’in Lun’ reached the court of the emperor Wu at Nanking. All this does not really add up to very much. The two greatest empires of the ancient world, imperial Rome and Han China, essentially proceeded on entirely separate paths, all but unaware of the other’s existence.

Several Roman towns in the upper Euphrates region certainly grew rich from the overland trade with the east. Foremost among them was the great caravan city of Palmyra, lying deep in the Syrian desert between Antioch and the Euphrates, where fragments of Chinese silk have been discovered (Plate 22). However, this need not necessarily mean that there was an overland ‘Silk Road’ running directly between Syria and China. In AD 18/19, Germanicus, the nephew of the emperor Tiberius, sent a certain Alexandros on an official embassy from Palmyra to the head of the Persian gulf. To all appearances, Germanicus was attempting to formalize a caravan route between Palmyra and the harbour-towns at the mouths of the Euphrates and Tigris rivers. In the first and second centuries AD, Palmyrene trading colonies were established in several parts of southern Mesopotamia, and tombs of Palmyrene traders have been found as far south as the island of Kharg, off the Iranian coast in the Persian gulf; we have a long series of inscriptions from Palmyra honouring individuals who helped to conduct the camel caravans from Spasinou Charax (near the mouth of the Tigris) to Palmyra. It seems more likely, then, that Chinese silk reached Palmyra via the ordinary southern route from India, being carried by sea from the Indus delta to the head of the Persian gulf, where it was loaded onto Palmyrene camels for the overland journey up the Euphrates valley to Syria.

This crucial caravan route, from Palmyra down the Euphrates to the Persian gulf, was utterly dependent on stable relations with Rome’s immediate neighbour to the east, the huge empire of the Parthians. The Parthians were originally a nomadic people, who migrated south from the steppes of central Asia into northern Iran in the early third century BC. Over the course of the third and second centuries BC, the Parthians expanded south and westwards into Seleucid territory; by the end of the reign of the Parthian king Mithradates II (124/3–88/7 BC), the whole of modern Iran and most of Mesopotamia, including Babylon, had fallen under Parthian control. The westward expansion of the Parthians inevitably brought them into conflict with Rome. Over the course of the first century BC Rome repeatedly provoked war with Parthia, with limited success and enormous loss of manpower; in 53 BC the Roman general Crassus lost 30,000 soldiers at the battle of Carrhae, in the course of an ill-judged invasion of northern Mesopotamia. Finally in 20 BC the emperor Augustus struck a treaty with the Parthians, fixing the Euphrates as the boundary between the two empires. As part of the deal, the Parthians agreed to return the military standards captured from Crassus thirty years earlier. For Rome, the treaty with Parthia was essentially a recognition of fifty years of military failure in the east. Nonetheless, Augustus cheerfully represented the treaty as a heroic Roman victory. A triumphal arch was erected in the Roman Forum, along with a temple of Mars the Avenger in Augustus’ new Forum Augustum (see above, Figure 25); coins were struck depicting the Parthian king kneeling in submission, meekly handing back the Roman standards captured at Carrhae (see Plate 31e).

Augustus and his successors consistently represented their exploits against the Parthians, which were not especially glorious, as the continuation of the endless struggle between European civilization and the eastern barbarian. In AD 61/2 the Athenians set up a huge honorific inscription to the emperor Nero, in gilded letters, on the east wall of the Parthenon, to celebrate Nero’s indecisive campaigns in the Parthian client kingdom of Armenia (AD 54–63). The location is significant: the Parthenon itself was a memorial to the Greek victory in the Persian Wars of the early fifth century BC, and Nero’s inscription overlooked the late-third-century BC monument of Attalus I commemorating his victories over the Galatians (the ‘New Persians’ of the day: above, Figure 20). Similarly, at the Sebasteion of Aphrodisias in Caria, Nero was portrayed supporting a slumped personification of ‘barbarian’ Armenia, depicted in the character of an oriental Amazon.

In the second century AD a succession of Roman emperors attempted once again to push Rome’s eastern borders beyond the Euphrates. Septimius Severus (193–211) finally succeeded in creating two new Roman provinces beyond the Euphrates in northern Mesopotamia, the first lasting gains ever achieved by Rome against the Parthians. His son Caracalla (211–217), encouraged by these victories, entertained grand dreams of emulating the conquests of Alexander the Great in the east. In 214 he enlisted a phalanx of Macedonian soldiers, and set out to follow Alexander’s path from Macedonia to Persia. The expedition began farcically and ended in disaster. Like Alexander, he crossed from Europe to Asia via the Hellespont; unfortunately, his boat capsized halfway across and the emperor had to be rescued by one of his officers. After tracing Alexander’s route as far as Egypt, he provoked a gratuitous and humiliatingly ineffective war with Parthia (216–18); Caracalla himself was murdered by his own court in 217, and the war ended with Rome paying a huge indemnity to the Parthians, having won not a single inch of new territory.

It may ultimately have been the loss of northern Mesopotamia to Septimius Severus which led to the downfall of the Parthians. In the early third century AD the old Persian heartlands of southern Iran rose up against Parthian rule, under the leadership of an energetic new Iranian dynasty, the Sasanians. In AD 224 the Sasanian king Ardashir I slew the last Parthian king in battle, and by 226 the whole of the former Parthian empire was under Sasanian control. From the outset, the new Sasanian dynasty took a much more aggressive attitude towards the west. Ardashir promptly launched the first of a series of dramatic raids on the Roman fortresses of upper Mesopotamia; perhaps more significantly, his seizure of the ports at the head of the Persian gulf brought a sudden end to the profitable Roman overland caravan trade from Palmyra. His successor, Shapur I (240–72), continued his father’s policy of uncompromising expansion in the west. Shapur swept the Romans out of northern Mesopotamia, and even captured the great city of Antioch, a stone’s throw from the coast of the Mediterranean itself. In 243, a Roman emperor, Gordian III, was killed in battle against Shapur, and in 260, Shapur succeeded in capturing alive the Roman emperor Valerian. The Roman prisoners of war were set to work building a new royal city, Bishapur, high in the Zagros mountains of western Iran; the city’s fine coloured mosaics were almost certainly laid by captured Roman craftsmen.

The Sasanian victories of the mid-third century were the most spectacular achievements of a foreign power against Rome since the wars with Carthage almost half a millennium previously. It is particularly striking that Ardashir and Shapur chose to commemorate their victories in ways which linked them explicitly with Iranian memories of the Achaemenid Persian world empire of the sixth and fifth centuries BC. The most dramatic monuments of the Persian imperial era to survive into the Sasanian period were the royal tombs at Naqsh-i Rustam, near modern Shiraz. The colossal tombs of the Achaemenid kings Darius I, Xerxes, Artaxerxes I and Darius II are still visible today, carved into a huge cliff-face overlooking the Marv Dasht plain. Both Ardashir and Shapur added rock-cut reliefs of their own, slotted in neatly between the Achaemenid royal tombs, commemorating Ardashir’s accession to the throne (the re-establishment of the true Achaemenid royal line) and Shapur’s defeat of the emperor Valerian (the modern counterpart of the Achaemenid conquest of the west). In front of the cliff with the royal tombs stands a great stone tower, the Kaaba-i Zardusht, probably an early Persian fire-temple. It was on the base of this Achaemenid-era monument that Shapur carved a long trilingual inscription in Parthian, Middle Persian and Greek, recounting his victories in the west. Shapur’s victories over three successive Roman emperors are thus set on a level with the conquests of his Achaemenid ancestors. It is still very unclear exactly how much the Sasanians actually knew about the Achaemenids and their empire, but there is some evidence that they went so far as to lay claim to all the lands once ruled by the Persians. Shapur II (309–79), in a letter to the Roman emperor Constantius II (337–61), declared that ‘my ancestors’ empire stretched to the river Strymon and the borders of Macedonia, as even your own historical records attest; it is right that I should demand these lands, inasmuch as I surpass even those ancient kings in magnificence’. The Sasanian empire, no less than the Roman empire, was profoundly shaped by its memories of the past.

By the time of the accession of Constantius II in 337, the Roman world had undergone a full-scale religious revolution, with momentous consequences for the history of Europe down to the present day. For Constantius’ empire was now a Christian empire. InAD326, Constantius’ father, Constantine I (306–37), had written to the young Shapur II to commend the Christian faith to him, and to ask him to protect the Christians living under his rule. After three centuries of persecution and martyrdom, the tiny Messianic movement born under Tiberius among the fishermen on the shores of lake Galilee could now claim the public adherence of the Roman emperor himself.

The Church began as a reform movement within Judaism. Jesus’ own ministry in Galilee and Jerusalem (c.AD 28–30) was directed primarily at the Jewish people, and long after the crucifixion many Jewish converts remained deeply hostile to the idea of ‘gentile Christianity’. The crucial turning point, which set Christianity on a different path from other Jewish Messianic sects, came in AD 48 or 49, with the decision of the apostolic conference at Jerusalem that while gentile converts to Christianity must refrain from idolatory and unchastity and observe the main Jewish dietary laws, they need not undergo circumcision. This decision set the early Church firmly on a path of openness: gentiles who were unwilling to accept the rigid limits of Rabbinic Judaism were nonetheless welcomed into the Christian religion.

Over the second half of the first century AD the boundaries between Jew and Christian gradually became sharper. By AD 112, when the provincial governor Pliny the Younger wrote to the emperor Trajan to report the measures he had taken against a group of Christians brought to trial before him in Pontus (north-central Asia Minor), Christianity was clearly recognized by the Roman state as a sect distinct from Judaism. It is significant that Christians seem always to have been exempt from the Jewish tax, which had been imposed on all the Jews of the empire after the fall of Jerusalem in AD 70. Indeed, it is hardly surprising that some early Christians – most famously the adherents of the heretical bishop Marcion (c. AD 85–160) – came to reject Jewish scripture altogether. For the Marcionites, Christ was the son of a previously unknown god, quite different from the Jewish God of the Old Testament. Christianity, on this line, was an entirely new religion, with no past stretching back earlier than Jesus’ lifetime. Unsurprisingly, this radical denial of the Church’s Jewish roots met with ferocious opposition among ordinary Christians. Most Christians recognized the authority of the Jewish scripture; their attitude to the Old Testament differed from that of the Jews only in their belief that the scriptural prophecies had now been fulfilled in the person of Christ.

One of the key debates within the early Church was the question of how far the Christian faith was compatible with the Roman world order. The First Epistle of Peter, composed at an uncertain date in the late first century AD, is quite clear on the matter: ‘Be subject for the Lord’s sake to every human institution, whether it be to the emperor as supreme, or to governors as sent by him to punish those who do wrong and to praise those who do right … Fear God. Honour the emperor.’ There was, it is true, a more militant strand in the Church, best represented for us by the Revelation of St John of Patmos. John’s vision of Rome as the whore of Babylon, soon to be laid low by pestilence and mourning and famine, was hardly designed to promote friendly relations between Christians and pagans. This militant vision was fuelled by the increasingly fierce persecution of Christians in the second and third centuries AD by the imperial authorities. It is clear from Pliny’s prosecution of the Christians of Pontus in 112 that profession of Christianity was a capital crime; it is equally clear that both Pliny and Trajan were fairly relaxed about enforcing the law. In the second and early third centuries, only the most outspoken fanatics could expect to face the death penalty. It therefore came as a real shock to the empire’s Christians when, in 249, the emperor Decius (249–51) passed an edict requiring all the inhabitants of the empire to offer sacrifice to the pagan gods. The persecution which followed led to the execution of the bishops of Rome and Alexandria, among many others. Although the mass trials and martyrdoms eased off in the 260s, the experience of centrally directed persecution had shown quite how difficult it was to be a good Christian and a good Roman at the same time.

Those aspects of the third-century Church which seem strangest and most foreign to us today – the rejection of marriage and embrace of perpetual virginity, martyrdom, asceticism and self-abasement – can be seen in large part as an attempt to reject, as violently as possible, all the established norms and values of Roman society. Some of the internal conflict and anguish this provoked for individual converts can be seen in the remarkable account which has come down to us of the martyrdom of a Christian woman by the name of Perpetua, executed at Carthage in 203. Perpetua was, in most respects, a model member of provincial Roman society: a Roman citizen of the local Carthaginian nobility, married with a young child. The night before her death, Perpetua dreamed that she was wrestling with a black Egyptian in the amphitheatre at Carthage; the bout was presided over by a man clad in a purple robe, holding a branch with golden apples as a prize for the victor. On waking, she realized that her Egyptian opponent represented the devil, and the purple-clad master of ceremonies Christ. Perpetua’s dream gives us a precious insight into an early Christian martyr’s subconscious. From our perspective, the interesting thing is that she subconsciously interpreted her martyrdom in purely civic and pagan terms: the amphitheatre, the games, the master of ceremonies with his prize of golden apples (the chief prize at the pagan festival of the Pythia at Carthage in the early third century AD). The early Christians could not help but conceive their faith in terms of the Roman world they claimed to have put aside. At the very moment of her death, Perpetua ‘pulled down her tunic, which was torn down the side, to cover her thighs, thinking more of her modesty than her pain’; one of her last acts was to ask for a hairpin to fasten her loosened hair. Before the crowds in the amphitheatre at Carthage, Perpetua’s first instinct was still to try to look and behave like an appropriately modest and decorous Roman wife.

By the third century the Church had become a highly organized, empire-wide state within the Roman state. As a result of the struggle with heretical sects such as the Marcionites, the Church had developed a strict internal hierarchy, its own legal system, and a procedure for deciding doctrinal problems through universal (ecumenical) Church councils. Christians’ relations with their non-Christian neighbours also showed an increasing sophistication and openness. Alongside the use made by the Church of the Jewish past, particularly in terms of Old Testament exegesis, Christians also began to negotiate with and tap into the civic cults and ideology of the cities of the Roman provinces. From the reign of Septimius Severus (193–211) onwards, the city of Apamea in Phrygia minted large numbers of bronze coins depicting the Judaeo-Christian story of Noah and the Ark. Apamea had long been informally known as Apamea ‘Kibotos’, the Chest, thanks to its status as the main Roman trading post in inland Asia Minor. However, the Greek wordkibotos also means ‘Ark’. The Noah coins of Apamea are apparently the result of a highly subtle and effective act of negotiation between the Christian and pagan communities at Apamea. The Christians of Apamea had successfully argued that Apamea’s nickname derived from its status as the resting place of the Ark of Noah after the Flood, the true location of Mt. Ararat being unknown at this point. We have seen already in this chapter that the cities of inland Asia Minor were desperately keen to present themselves as communities of great antiquity, founded deep in the legendary past before the Trojan War. The claim that the city of Apamea was the first city to be founded after the biblical Flood pushed all the right buttons for the pagans of Apamea; the town’s Christian community had found the perfect way of legitimizing and publicizing their religion in terms which could be easily understood and appreciated by non-Christians.

We have no way of telling how many Christians there were in the Roman empire in the first three centuries AD. Most probably, the size of the Christian community differed radically from one region, even from one city, to another. According to a reliable source (Eusebius’ Church History), there were 155 members of the Christian clergy in the city of Rome in AD 251; by this point the Christian population of Rome must have numbered in the thousands. The only area where we have real evidence for the spread of Christianity is, once again, the Phrygian highlands of inland Asia Minor. Christianity struck deep roots in Phrygia at an early date. This isolated rural area had had a large Jewish population since the Hellenistic period, and the Phrygian brand of paganism had strong monotheistic tendencies even before the coming of the Church. In the upper Tembris valley, a remote part of northern Phrygia, around 20 per cent of the population was openly professing the Christian faith on their tombstones by 230; by the end of the third century the proportion has risen to more than 80 per cent. In the late third and early fourth centuries AD Egyptian papyri show a sudden explosion of characteristically Christian personal names like David, Matthew, Johannes. The proportion of identifiably Christian names in Egypt rose from around 10–15 per cent of the population in AD 280 to c. 50 per cent at the death of Constantine in 337. By 425, a century later, the proportion had reached 80 per cent. At least in the eastern provinces, then, the late third and early fourth centuries mark the crucial turning point; by the end of Constantine’s reign, the victory of the Church was assured. It is this new, Christian empire of the fourth century AD which will be the subject of our final chapter.