Читать книгу Passages From Antiquity to Feudalism - Perry Anderson - Страница 10

4

Rome

The rise of Rome marked a new cycle of urban–imperial expansion, which represented not just a geographical shift in the centre of gravity of the Ancient world to Italy, but a socio-economic development of the mode of production which had been pioneered in Greece that rendered possible a far greater and more lasting dynamism than that which had produced the Hellenistic epoch. The early growth of the Roman Republic followed the normal course of any ascendant classical city-state: local wars with rival cities, annexation of land, subjection of ‘allies’, foundation of colonies. In one critical respect, however, Roman expansionism distinguished itself at the outset from Greek experience. The constitutional evolution of the city conserved aristocratic political power right down and into the classical phase of its urban civilization. Archaic monarchy was overthrown by a nobility in the earliest epoch of its existence, at the end of the 6th century B.C., in a change strictly comparable to the Hellenic pattern. But thereafter, unlike the Greek cities, Rome never knew the upheaval of tyrant rule, breaking aristocratic dominance and leading to a subsequent democratization of the city, based on a secure small and medium agriculture. Instead a hereditary nobility kept unbroken power through an extremely complex civic constitution, which underwent important popular modifications through the course of a prolonged and fierce social struggle within the city, but was never abrogated or replaced. The Republic was dominated by the Senate, which was controlled for the first two centuries of its existence by a small group of patrician clans; membership of the Senate, which was cooptive, was for life. Annual magistrates, of which the two highest were consuls, were elected by ‘assemblies of the people’, comprising the whole citizenry of Rome, but organized into unequal ‘centuriate’ units weighted to ensure a majority of the propertied classes. The consulates were the supreme executive offices of the State, and were legally monopolized by the closed order of patricians down to 366 B.C.

This original structure embodied the political dominion of the traditional aristocracy pure and simple. It was subsequently altered and qualified in two important respects, after successive struggles which provide the nearest Roman equivalent of the Greek phases of ‘tyranny’ and ‘democracy’, but which each time fell decisively short of the comparable outcome in Greece. Firstly, recently enriched ‘plebeians’ forced the ‘patrician’ nobility to concede access to one of the two annual consular offices from 366 B.C. onwards; although it was not until nearly two hundred years later, in 172 B.C., that both consuls were for the first time plebeians. This slow change led to a broadening of the composition of the Senate itself, since former consuls automatically became senators. The result was the social formation of a widened nobility, including both ‘patrician’ and ‘plebeian’ families, rather than the political overthrow of the system of aristocratic rule itself, such as had occurred during the age of tyrants in Greece. Chronologically and sociologically overlapping this contest within the wealthiest strata of the Republic, was the struggle of the poorer classes to gain increased rights within it. Their pressure quickly resulted in the creation of the tribunate of the plebs, a corporate representation of the popular mass of the citizenry. The tribunes were elected each year by a ‘tribal’ assembly which, unlike the ‘centuriate’ assembly, was in principle genuinely egalitarian: the ‘tribes’ were actually, as in Archaic Greece, territorial rather than kin divisions of the population, four in the city itself and seventeen outside it (an indication of the degree of urbanization at that date). The tribunate formed a secondary and parallel executive agency, designed to protect the poor from the oppression of the rich. Eventually, in the early 3rd century, the tribal assemblies which elected the tribunes gained legislative powers, and the tribunes themselves nominal rights of veto over the acts of the consuls and the decrees of the Senate.

The direction of this evolution corresponded to the process that had led to the democratic polis in Greece. But here too, the process was arrested before it could impend a new political constitution for the city. The tribunate and tribal assembly were simply added to the central existing institutions of the Senate, Consulates and Centuriate Assembly: they did not signify an internal abolition of the oligarchic complex of power which guided the Republic, but external accretions to it; whose practical significance was often much less than their formal potential. For the struggle of the poorer classes had generally been led by wealthy plebeians, who championed the popular cause to further their own parvenu interests: and this continued to be true even after the newly rich had gained access to the ranks of the senatorial order itself. The tribunes, normally men of considerable fortunes, thus became for long periods docile instruments of the Senate itself.¹ Aristocratic supremacy within the Republic was not seriously shaken. A plutocracy of wealth now merely enlarged a nobility of birth, both using extensive ‘clientage’ systems to ensure a deferent following among the urban masses, and lavish customary bribes to secure election to the annual magistracies through the centuriate assembly. The Roman Republic thus retained traditional oligarchic rule, through a composite constitution, down to the classical epoch of its history.

The resultant social structure of the Roman citizenry was thus inevitably distinct from that which had been typical of classical Greece. The patrician nobility had early on striven to concentrate landed property in its hands, reducing the poorer free peasantry to debt bondage (as in Greece), and appropriating the ager publicus or common lands which they used for pasturage and cultivation. The tendency to abase the peasantry by debt bondage to the condition of dependent tenants was checked, although the problem of debts themselves persisted:² but the expropriation of the ager publicus and the depression of medium and small farmers was not. There was no economic or political upheaval to stabilize the rural property of the ordinary citizenry of Rome, comparable to that which had occurred in Athens or, in a different way, Sparta. When the Gracchi eventually attempted to follow the path of Solon and Peisistratus, it was too late: by then, the 2nd century B.C., much more radical measures than those enacted in Athens were necessary to save the situation of the poor – nothing less than a redistribution of land, demanded by both Gracchi brothers – with correspondingly less chance of them ever being implemented over aristocratic opposition. In fact, no durable or substantial agrarian reform ever occurred in the Republic, despite constant agitation and turbulence over the question in the final epoch of its existence. The political dominance of the nobility blocked all efforts to reverse the relentless social polarization of property on the land. The result was a steady erosion of the modest farmer class that had provided the backbone of the Greek polis. The Roman equivalent of the hoplite category – men who could equip themselves with armour and weaponry necessary for infantry service in the legions – were the assidui or ‘those settled on the land’, who possessed the necessary property qualification to bear their own arms. Below them were the proletarii, propertyless citizens, whose service to the State was merely to rear children (proles). The increasing monopolization of land by the aristocracy was thus translated into a steady decline in the numbers of the assidui, and an inexorable increase in the size of the proletarii class. Moreover, Roman military expansionism also tended to thin the ranks of the assidui who provided the conscripts and casualties for the armies with which it was conducted. The result was that by the end of the 3rd century B.C., the proletarii were probably already an absolute majority of citizens, and had to be themselves called up to deal with the emergency of Hannibal’s invasion of Italy; while the property qualification for the assidui was twice reduced, until in the next century it had sunk below a subsistence minimum on the land.³

Small-holders never disappeared generally or completely in Italy; but they were increasingly driven into the more remote and precarious recesses of the country, in marshy or mountainous regions unappealing to engrossing landowners. The structure of the Roman polity in the Republican epoch thus came to diverge sharply from any Greek precedent. For while the countryside became chequered with large noble domains, the city conversely became populated with a proletarianized mass, deprived of land or any other property. Once fully urbanized, this large and desperate underclass lost any will to return to a small-holder condition, and could often be manipulated by aristocratic cliques against projects for agrarian reform backed by the assidui farmers.⁴ Its strategic position in the capital of an expanding empire ultimately obliged the Roman ruling class to pacify its immediate material interests with public grain distributions. These were, in effect, a cheap substitute for the land distribution which never occurred: a passive and consumer proletariat was preferable to a recalcitrant and producer peasantry, for the senatorial oligarchy which controlled the Republic.

It is now possible to consider the implications of this configuration for the specific course of Roman expansionism. For the growth of Roman civic power was consequently distinguished from Greek examples in two fundamental respects, both related directly to the internal structure of the city. Firstly, Rome proved able to widen its own political system to include the Italian cities it subjugated in the course of its peninsular expansion. From the start, it had – unlike Athens – exacted troops for its armies, not money for its treasury, from its allies; thereby lightening the burden of its domination in peace, and binding them solidly to it in time of war. In this, it followed the example of Sparta, although its central military control of allied troops was always much greater. But Rome was also able to achieve an ultimate integration of these allies into its own polity which no Greek city had ever envisaged. It was the peculiar social structure of Rome which permitted this. Even the most oligarchic Greek polis of the classical epoch basically rested on a median body of propertied citizens, and precluded extreme economic disparities of wealth and poverty within the city. The political authoritarianism of Sparta – the exemplar of Hellenic oligarchy – did not mean a class polarization within the citizenry: in fact, as we have seen, it was accompanied by marked economic egalitarianism in the classical epoch, probably including allocation of inalienable state holdings to each Spartiate precisely to ensure hoplites against the type of ‘proletarianization’ which overtook them in Rome.⁵ The classical Greek polis, whatever its degree of relative democracy and oligarchy, retained a civic unity rooted in the rural property of its immediate locality: it was for the same reason territorially inelastic – incapable of an extension without loss of identity. The Roman Constitution, by contrast, was not merely oligarchic in form: it was much more deeply aristocratic in content, because behind it lay an economic stratification of Roman society of quite another order. It was this which rendered possible an extension of Republican citizenship outwards to comparable ruling classes in the allied cities of Italy, who were socially akin to the Roman nobility itself, and had benefited from Roman conquests overseas. The Italian cities finally revolted against Rome in 91 B.C., when their demand for the Roman franchise was refused – something no Athenian or Spartan ally had ever requested. Even then, their war aim was a peninsular Italian state with a capital and Senate, in avowed imitation of the unitary Roman order itself, rather than any return to scattered municipal independence.⁶ The Italian rebellion was militarily defeated in the long and bitter struggle of the so-called Social War. But amidst the subsequent turmoil of the Civil Wars between the Marian and Sullan factions within the Republic, the Senate could concede to the basic political programme of the allies, because the character of the Roman governing class and its Constitution facilitated a viable extension of citizenship to the other Italian cities, ruled by an urban gentry similar in character to the Senatorial class itself, with the wealth and leisure to participate in the political system of the Republic, even from a distance. The Italian gentry by no means consummated its political aspirations for central office within the Roman State immediately, and its ulterior ambitions after the grant of citizenship were to be a powerful force for social transformations at a later date. But their civic integration nevertheless represented a decisive step for the future structure of the Roman Empire as a whole. The relative institutional flexibility which it demonstrated gave Rome a signal advantage in its imperial ascent: it meant an avoidance of either of the two poles between which Greek expansion had divided and foundered – premature and impotent closure of the city-state or meteoric royal triumphalism at the expense of it. The political formula of Republican Rome marked a notable advance in comparative efficacy.

Yet the decisive innovation of Roman expansion was ultimately economic: it was the introduction of the large-scale slave latifundium for the first time in Antiquity. Greek agriculture had, as we have seen, employed slaves widely; but it was itself confined to small areas, with a meagre population, for Greek civilization always remained precariously coastal and insular in character. Moreover, and above all, the slave-tilled farms of Attica or Messenia were usually of very modest size – perhaps an average of some 30 to 60 acres, at most. This rural pattern was, of course, linked to the social structure of the Greek polis, with its absence of huge concentrations of wealth. Hellenistic civilization had, by contrast, witnessed enormous accumulations of landed property in the hands of dynasties and nobles, but no widespread agricultural slavery. It was the Roman Republic which first united large agrarian property with gang-slavery in the countryside on a major scale. The advent of slavery as an organized mode of production inaugurated, as it had in Greece, the classical phase proper of Roman civilization, the apogee of its power and culture. But whereas in Greece it had coincided with the stabilization of small farms and a compact citizen corps, in Rome it was systematized by an urban aristocracy which already enjoyed social and economic dominion over the city. The result was the new rural institution of the extensive slave latifundium. The manpower for the enormous holdings which emerged from the late 3rd century onwards was supplied by the spectacular series of campaigns which won Rome its mastery of the Mediterranean world: the Punic, Macedonian, Jugurthine, Mithridatic and Gallic wars, which poured military captives into Italy to the profit of the Roman ruling class. At the same time, successive ferocious struggles fought on the soil of the peninsula itself – the Hannibalic, Social and Civil Wars – delivered into the grasp of the senatorial oligarchy or its victorious factions large territories expropriated from the defeated victims of these conflicts, especially in Southern Italy.⁷ Moreover, these same external and internal wars dramatically accentuated the decline of the Roman peasantry, which had once formed the robust small-holder base of the city’s social pyramid. Constant warfare involved endless mobilization; the assidui citizenry called to the legions year after year died in thousands under their standards, while those who survived them were unable to maintain their farms at home, which were increasingly absorbed by the nobility. From 200 to 167 B.C., 10 per cent or more of all free Roman adult males were permanently conscripted: this gigantic military effort was only possible because the civilian economy behind it could be manned to such an extent by slave-labour, releasing corresponding manpower reserves for the armies of the Republic.⁸ Victorious wars in their turn provided more slave-captives to pump back into the towns and estates of Italy.

The final outcome was the emergence of slave-worked agrarian properties of a hitherto unknown immensity. Prominent nobles like Lucius Domitius Ahenobarbus could own over 200,000 acres in the 1st century B.C. These latifundia represented a new social phenomenon, which transformed the Italian countryside. They did not, of course, necessarily or invariably form consolidated blocks of land, farmed as single units.⁹ The typical pattern was for the latifundist to possess a large number of medium-sized villa estates, sometimes contiguous but perhaps equally often distributed across the country, designed for optimal surveillance by various bailiffs and agents. Even such dispersed holdings, however, were notably larger than their Greek predecessors, often exceeding 300 acres (500 iugera) in extent; while consolidated estates like the Younger Pliny’s seat in Tuscany might be 3,000 acres or more in size.¹⁰ The rise of the Italian latifundia led to a great extension of pastoral ranching, and the inter-cropping of wine and olives with cereal cultivation. The influx of slave-labour was so great that by the late Republic, not only was Italian agriculture recast by it, but trade and industry were overwhelmingly invaded by it too: perhaps 90 per cent of the artisans in Rome itself were of slave origin.¹¹ The nature of the gigantic social upheaval that Roman imperial expansion involved, and the basic motor-force that sustained it, can be seen from the sheer demographic transformation that it wrought. Brunt estimates that in 225 B.C. there were some 4,400,000 free persons in Italy, to 600,000 slaves; by 43 B.C., there were perhaps 4,500,000 free to 3,000,000 slave inhabitants – indeed there may actually have been a net decline in the total size of the free population, while the slave population quintupled.¹² Nothing like this had ever been seen in the Ancient World before. The full potential of the slave mode of production was for the first time unfolded by Rome, which organized and took it to a logical conclusion that Greece had never experienced. The predatory militarism of the Roman Republic was its main lever of economic accumulation. War brought lands, tributes and slaves; slaves, tributes and lands supplied the materiél for war.

But the historical significance of the Roman conquests in the Mediterranean basin was, of course, by no means reducible simply to the spectacular fortunes of the senatorial oligarchy. The march of the legions accomplished a change much deeper than this, for the whole history of Antiquity. Roman power integrated the Western Mediterranean and its northern hinterlands into the classical world. This was the decisive achievement of the Republic, which in contrast with its diplomatic caution in the East, from the outset unleashed its annexationist drive essentially in the West. Greek colonial expansion in the Eastern Mediterranean, as has been seen, took the form of a proliferation of urban foundations, first created from above by the Macedonian rulers themselves, then soon imitated from below by the local gentry of the region; and it occurred in a zone with an extremely long prior history of developed civilization, stretching back much further than that of Greece itself. Roman colonial expansion in the Western Mediterranean differed basically in both context and character. Spain and Gaul – later Noricum, Rhaetia and Britain – were remote and primitive lands peopled by Celtic tribal communities – many with no history of contact at all with the classical world. Their integration into it posed problems of an altogether different order from that of the Hellenization of the Near East. For not only were they socially and culturally backward: they also represented interior land-masses of a type that classical Antiquity had hitherto never been able to organize economically. The original matrix of the city-state was the narrow littoral and the sea, and classical Greece had never relinquished it. The Hellenistic epoch had seen the intensive urbanization of the riparian cultures of the Near East, long based on fluvial irrigation and now partly reoriented towards the sea (a modification symbolized by the shift from Memphis to Alexandria). But the desert lay close behind the whole Southern and Eastern coast-line of the Mediterranean, so that the depth of settlement was never very great either in the Levant or North Africa. The Western Mediterranean, however, offered neither a littoral nor an irrigative system to the advancing Roman frontiers. Here, for the first time, classical Antiquity was confronted with great interior land-masses, devoid of previous urban civilization. It was the Roman city-state, that had developed the rural slave-latifundium, that proved capable of mastering them. The river-routes of Spain or Gaul assisted this penetration. But the irresistible impetus which carried the legions to the Tagus, the Loire, the Thames and the Rhine, was that of the slave mode of production fully unleashed on the land, without constrictions or impediments. It was in this epoch that probably the major single advance in the agrarian technology of classical Antiquity was registered: the discovery of the rotary mill for grinding corn, which in its two main forms was first attested in Italy and Spain in the mid 2nd century B.C.,¹³ concomitant with Roman expansion in the Western Mediterranean, and a sign of its rural dynamism. The successful organization of large-scale agrarian production by slave-labour was the precondition of the permanent conquest and colonization of the great Western and Northern hinterlands. Spain and Gaul remained, with Italy, the Roman provinces most deeply marked by slavery down to the final end of the Empire.¹⁴ Greek trade had permeated the East: Latin agriculture ‘opened up’ the West. Naturally, towns were founded by the Romans in the Western Mediterranean too, characteristically built along the banks of navigable rivers. The creation of a slave-worked rural economy itself depended on the implantation of a prosperous network of cities which represented the terminal points for its surplus produce, and its structural principle of articulation and control. Cordoba, Lyon, Amiens, Trier and hundreds of other towns were constructed. Their number never equalled that of the much older and more densely populated Eastern Mediterranean society, but it was far larger than that of the cities founded by Rome in the East.

For Roman expansion into the Hellenistic zone followed a very different course from its pattern in the Celtic backlands of the West. It was for a long time much more hesitant and uncertain, oriented towards blocking interventions to check major disruptions of the existing state system (Philip V, Antiochus III), and creating client realms rather than conquered provinces.¹⁵ Thus it was characteristic than even after the rout of the last great Seleucid army at Magnesia in 198, no Eastern territory was annexed for another fifty years; and it was not until 129 B.C. that Pergamum passed peacefully into Roman administration, by the testament of its loyal monarch rather than by a senatorial volition, to become the first Asian province of the Empire. Thereafter, once the immense riches available in the East were fully realized in Rome, and army commanders gained escalating imperial powers abroad, aggression became more rapid and systematic, in the 1st century B.C. But the Republican regimes generally administered the profitable Asian provinces which their generals now seized from their Hellenistic rulers, with a minimum of social change or political interference, professing to have ‘liberated’ them from their royal despots, and contenting themselves with the lush tax returns of the region, There was no widespread introduction of agrarian slavery in the Eastern Mediterranean; the numerous war prisoners enslaved there were shipped westwards for employment in Italy itself. Royal estates were appropriated by Roman administrators and adventurers, but their labour systems were left effectively intact. The main innovation of Roman rule in the East occurred in the Greek cities throughout the region, where property qualifications were now imposed for municipal office – to align them more closely with the oligarchic norms of the Eternal City itself; in practice, this merely gave juridical codification to the de facto power of the local notabilities who dominated these towns already.¹⁶ A few specifically Roman urban colonies were created in the East by Caesar and Augustus, to settle Latin proletarians and veterans in Asia. But these left very little mark. Significantly, when a new wave of cities were built under the Principate (above all, in the Antonine epoch), they were essentially Greek foundations, consonant with the previous cultural character of the region. There was never any attempt to Romanize the Eastern provinces; it was the West which underwent the full brunt of Latinization. The language frontier – running from Illyricum to Cyrenaica – demarcated the two basic zones of the new imperial order.

The Roman conquest of the Mediterranean in the last two centuries of the Republic, and the massive expansion of the senatorial economy which it promoted, was accompanied at home by a superstructural development without precedent in the Ancient world. For it was in this period that Roman civil law emerged in all its unity and singularity. Gradually developed from 300 B.C. onwards, the Roman legal system became essentially concerned with regulation of informal relationships of contract and exchange between private citizens. Its fundamental orientation lay in economic transactions – purchase, sale, hire, lease, inheritance, security – and their familial concomitants – matrimonial or testamentary. The public relationship of the citizen to the State, and the patriarchal relationship of the head of the family to his dependants, were marginal to the central development of legal theory and practice; the first was considered too mutable to be subject to systematic jurisprudence, while the second covered most of the inferior domain of crime.¹⁷ The real thrust of Republican jurisprudence was concerned with neither of these: it was not public or criminal law, but civil law governing suits between disputing parties over property, that formed the peculiar province of its remarkable advance. The development of a general legal theory as such was wholly new in Antiquity. It was the creation, not of State functionaries or of practising lawyers, but of specialized and aristocratic jurists who remained outside the process of litigation itself, furnishing opinions to the judiciary in actual court-cases only on questions of legal principle rather than matters of fact. Republican jurists, who had no official status, evolved a series of abstract ‘contractual figures’ applicable to the analysis of particular acts of commercial and social intercourse. Their intellectual bent was analytic rather than systematic, but the cumulative result of their work was the appearance, for the first time in history, of an organized body of civil jurisprudence as such. The economic growth of commodity exchange in Italy attendant on the construction of the Roman imperial system, founded on the extensive use of slavery, thus found its juridical reflection in the creation of an unexampled commercial law in the later Republic. The great, decisive accomplishment of the new Roman law was thus, appropriately enough, its invention of the concept of ‘absolute property’ – dominium ex jure Quiritium.¹⁸ No prior legal system had ever known the notion of an unqualified private property: ownership in Greece, Persia or Egypt had always been ‘relative’, in other words conditional on superior or collateral rights of other authorities and parties, or obligations to them. It was Roman jurisprudence that for the first time emancipated private ownership from any extrinsic qualifications or restraints, by developing the novel distinction between mere ‘possession’ – factual control of goods, and ‘property’ – full legal title to them. The Roman law of property, of which an extremely substantial sector was naturally devoted to ownership of slaves, represented the pristine conceptual distillation of the commercialized production and exchange of commodities within an enlarged State system, which Republican imperialism had made possible. Just as Greek civilization had been the first to disengage the absolute pole of ‘liberty’ from the political continuum of relative conditions and rights that had always prevailed before it, so Roman civilization was the first to separate the pure colour of ‘property’ from the economic spectrum of opaque and indeterminate possession that had typically preceded it. Quiritary ownership, the legal consummation of the extensive slave economy of Rome, was a momentous arrival, destined to outlive the world and age that had given birth to it.

The Republic had won Rome its Empire: it was rendered anachronistic by its own victories. The oligarchy of a single city could not hold the Mediterranean together in a unitary polity – it had been outgrown by the very scale of its success. The final century of Republican conquest, which took the legions to the Euphrates and the Channel, was accompanied by spiralling social tensions within Roman society itself – the direct outcome of the very triumphs that were being regularly won abroad. Peasant agitation for land had been stifled by the suppression of the Cracchi. But it now reappeared in new and menacing forms, within the army itself. Constant conscription had steadily weakened and reduced the whole small-holder class as such: but its economic aspirations lived on and now found expression in the mounting pressures from the time of Marius onwards for allocations of land to discharged veterans – the bitter survivors of the military duties that lay so heavily on the Roman peasantry. The senatorial aristocracy profited enormously from the financial sacking of the Mediterranean that succeeded progressive annexations by Rome, making boundless fortunes in tribute, extortion, land and slaves: but it was utterly unwilling to provide even a modicum of compensation to the soldiery whose fighting yielded these unheard-of gains to it. Legionaries were meanly paid and brusquely dismissed, without any solatium for long periods of service in which they not only risked their lives but often lost their property at home too. To have paid them bounties on discharge would have meant taxing the possessing classes, however slightly, and this the ruling aristocracy refused to consider. The result was to create an inherent tendency within the later Republican armies to a deflection of military loyalty away from the State, towards successful generals who could guarantee their soldiers plunder or donatives by their personal power. The bond between legionary and commander came increasingly to resemble that between patron and client in civilian life: from the epoch of Marius and Sulla onwards, soldiers looked to their generals for economic rehabilitation, and generals used their soldiers for political advancement. Armies came to be instruments of popular commanders, and wars started to become private ventures of ambitious consuls: Pompey in Bithynia, Crassus in Parthia, Caesar in Gaul, determined their own strategic plans of conquest or aggression.¹⁹ The factional rivalries which had traditionally rent municipal politics were consequently transferred onto a military stage, much vaster than the narrow confines of Rome itself. The inevitable result was to be the outbreak of full-scale civil wars.

At the same time, if peasant distress was the subsoil of the military turbulence and disorder of the late Republic, the plight of the urban masses acutely sharpened the crisis of senatorial power. With the extension of the Empire, the capital city of Rome itself increased uncontrollably in size. Growing rural drift from the land was combined with massive imports of slaves, to produce a vast metropolis. By the time of Caesar, Rome probably contained a population of some 750,000 – surpassing even the largest cities of the Hellenistic world. Hunger, disease and poverty squeezed the crowded slums of the capital, filled with artisans, labourers and petty-shopkeepers, whether slave, manumitted or freeborn.²⁰ The urban mob had been skilfully mobilized by noble manoeuvres against agrarian reformers in the 2nd century – an operation repeated once again with the abandonment of Catiline by the Roman plebs, which succumbed in time-honoured fashion to oligarchic propaganda against an ‘incendiary’ enemy of the State, to whom only Etrurian small-holders remained faithful to the end. But this was the last such episode. Thereafter, the Roman proletariat seems to have broken away irreversibly from senatorial tutelage; its mood became increasingly threatening and hostile to the traditional political order in the closing years of the Republic. Given the virtual absence of any solid or serious police force in a teeming city of three-quarters of a million inhabitants, the immediate mass pressure which urban riots could bring to bear in crises of the Republic was considerable. Orchestrated by the tribune Clodius, who armed sections of the city’s poor in the 50’s, the urban proletariat obtained a free grain dole for the first time in 53 B.C. – henceforward a permanent fact of Roman political life: the number of its recipients had risen to 320,000 by 46 B.C. Moreover, it was popular clamour that gave Pompey the extraordinary army commands which set in motion the final military disintegration of the senatorial state; popular enthusiasm for Caesar that rendered him so menacing to the aristocracy a decade later; and popular welcome that ensured him his triumphal reception in Rome after the crossing of the Rubicon. After Caesar’s death, it was once again the popular uproar in the streets of Rome at the absence of his heir that forced the Senate to plead for Augustus’s acceptance of renewed consular and dictatorial powers in 22–19 B.C., the definitive entombment of the Republic.

Finally, and perhaps most fundamentally of all, the self-protective immobilism and haphazard misgovernment of the Roman nobility in the conduct of its rule over the provinces rendered it increasingly unfit to manage a cosmopolitan empire. Its exclusive privileges were incompatible with any progressive unification of its overseas conquests. The provinces as such were still helpless to put up any serious resistance to its rapacious egoism. But Italy itself, the first province to achieve formal civic parity after a violent rebellion in the previous generation, was not. The Italian gentry had won juridical integration into the Roman community, but had not hitherto broken into the inner circle of senatorial office and power. With the eruption of the final round of civil wars between the Triumvirs, its opportunity for decisive political intervention had come. The provincial gentry of Italy flocked to Augustus, self-proclaimed defender of its traditions and prerogatives against the ominous and outlandish orientalism of Marcus Antonius and his camp.²¹ It was their adhesion to his cause, with the famous oath of allegiance sworn by ‘tota Italia’ in 32, which ensured the victory of Actium. It is significant that each of the three civil wars which determined the fate of the Republic followed the same geographical pattern: they were all won by the side which controlled the West, and lost by the party based on the East, despite its much greater wealth and resources. Pharsala, Philippi and Actium were all fought out in Greece, the advance-post of the defeated hemisphere. The dynamic centre of the Roman imperial system was shown once again to be in the Western Mediterranean. But whereas Caesar’s original territorial base had been the barbarian provinces of Gaul, Octavian forged his political bloc in Italy itself – and his victory proved in consequence less praetorian and more lasting.

The new Augustus garnered supreme power by uniting behind him the multiple forces of discontent and disintegration within the later Republic. He was able to rally the desperate urban plebs and weary peasant conscripts against a small and hated governing elite, whose opulent conservatism exposed it to ever greater popular contumely: and above all, he relied on the Italian provincial gentry who now sought their share of office and honour in the system which they had helped to build up. A stable, universal monarchy emerged from Actium, because it alone could transcend the narrow municipalism of the senatorial oligarchy in Rome. The Macedonian monarchy had been suddenly superimposed on a vast, alien continent and had failed to produce a unified ruling class to govern it post facto, despite Alexander’s possible awareness that this was the central structural problem it faced. The Roman monarchy of Augustus, by contrast, punctually arrived when its hour struck, neither too early nor too late: the critical passage from city-state to universal empire – the familiar cyclical transition of classical Antiquity – was accomplished with signal success under the Principate.

The most dangerous tensions of the later Republic were now lowered by a series of astute policies, designed to restabilize the whole Roman social order. First and foremost, Augustus provided allotments of land for the thousands of soldiers demobilized after the civil wars, financing many of them out of his personal fortune. These grants – like those of Sulla before them – were probably mostly at the expense of other small-holders, who were evicted to make room for home-coming veterans, and hence did little to improve the social situation of the peasantry as a whole or alter the general pattern of agrarian property in Italy;²² but they did effectively pacify the demands of the critical minority of the peasant class in arms, the key section of the rural population. Pay on active service had already been doubled by Caesar, an increase maintained under the Principate. More important still, from A.D. 6 onwards, veterans received regular cash bounties on discharge, worth thirteen years’ wages, which were paid out of a specially created military treasury financed by modest sales and inheritance taxes on the propertied classes of Italy. Such measures had been resisted to the death by the senatorial oligarchy, to their undoing: with the inauguration of the new system, discipline and loyalty returned to the army, which was trimmed from 50 to 28 legions, and converted into a permanent, professional force.²³ The result was to make possible the most significant change of all: conscription was lifted by the time of Tiberius, thereby relieving the Italian small-holders of the secular burden that had provoked such widespread suffering under the Republic – probably a more tangible benefit than any of the land allotment schemes.

In the capital, the urban proletariat was calmed with distributions of corn that were allowed to rise again from their Caesarian levels, and which were now better assured since the incorporation of the granary of Egypt into the Empire. An ambitious building programme was launched, which provided considerable plebeian employment, and the municipal services of the city were greatly improved, by the creation of effective fire-brigades and water-supplies. At the same time, praetorian cohorts and urban police were henceforward always stationed in Rome to quell tumults. In the provinces, meanwhile, the random and unbridled extortions of the Republican tax-farmers – one of the worst abuses of the old regime – were phased out, and a uniform fiscal system composed of a land-tax and poll-tax, and based on accurate censuses, was instituted: the revenues of the central state were increased, while peripheral regions no longer suffered the pillage of publicans. Provincial governors were henceforward paid regular salaries. The judicial system was overhauled to expand its appellate facilities against arbitrary decisions very greatly, both for Italians and provincials. An imperial postal service was created to link the far-flung provinces of the Empire together for the first time with a regular communications system.²⁴ Roman colonies and municipalities and Latin communities were planted in outlying zones, with a heavy concentration in the Western provinces. Domestic peace was restored after a generation of destructive civil strife, and with it provincial prosperity. On the frontiers, the successful conquest and integration of the critical corridors between East and West – Rhaetia, Noricum, Pannonia and Illyria – achieved the final geo-strategic unification of the Empire. Illyria, in particular, was henceforward the central military link of the imperial system in the Mediterranean.²⁵

Within the new borders, the advent of the Principate meant the promotion of Italian municipal families into the ranks of the senatorial order and upper administration, where they now formed one of the bastions of Augustus’s power. The Senate itself ceased any longer to be the central authority in the Roman State: it was not deprived of power or prestige, but it was henceforward a generally obedient and subordinate tool of successive emperors, reviving politically only during dynastic disputes or interregna. But while the Senate as an institution became a stately shell of its former self, the senatorial order itself – now purged and renovated by the reforms of the Principate – continued to be the ruling class of the Empire, largely dominating the imperial state machine even after equestrian appointments became normal to a wider range of positions within it. Its capacity for cultural and ideological assimilation of newcomers to its ranks was remarkable: no representative of the old patrician nobility of the Republic ever gave such powerful expression to its outlook on the world as a once modest provincial from Southern Gaul under Trajan, Tacitus. Senatorial oppositionism survived for centuries after the creation of the Empire, in quiescent reserve or refusal of the autocracy installed by the Principate. Athens, which had known the most untrammelled democracy of the Ancient World, produced no important theorists or defenders of it. Paradoxically yet logically, Rome, which had never experienced anything but a narrow and oppressive oligarchy, gave birth to the most eloquent threnodies for freedom in Antiquity. There was no real Greek equivalent to the Latin cult of Libertas, intense or ironic in the pages of Cicero or Tacitus.²⁶ The reason is evident from the contrasting structure of the two slave-owning societies. In Rome, there was no social conflict between literature and politics: power and culture were concentrated in a compact aristocracy under the Republic and the Empire. The narrower the circle that enjoyed the characteristic municipal freedom of Antiquity, the purer was the vindication of liberty it bequeathed to posterity, still memorable and formidable over fifteen hundred years later.

The senatorial ideal of libertas was, of course, suppressed and negated by the imperial autocracy of the Principate, and the resigned acquiescence of the propertied classes of Italy to the new dispensation, the alien visage of their own rule in the epoch to come. But it was never altogether cancelled, for the political structure of the Roman monarchy that now encompassed the whole Mediterranean world was never that of the Hellenistic monarchies of the Greek East which preceded it. The Roman imperial state rested on a system of civil laws, not mere royal caprice, and its public administration never interfered greatly with the basic legal framework handed down by the Republic. In fact, the Principate for the first time elevated Roman jurists to official positions within the State, when Augustus selected prominent jurisconsults for advisers and conferred imperial authority on their interpretations of the law. The Emperors themselves, on the other hand, were henceforward to legislate by edicts, adjudications and rescripts to questions or petitions from subjects. The development of an autocratic public law through imperial decretals, of course, rendered Roman legality much more complex and composite than it had been under the Republic. The political distance travelled from Cicero’s legum servi sumus ut liberi esse possimus (‘We obey laws in order to be free’) to Ulpian’s quod principi placuit legis habet vicem (‘The ruler’s will has force of law’) speaks for itself.²⁷ But the key tenets of civil law – above all, those governing economic transactions – were left substantially intact by this authoritarian evolution of public law, which by and large did not trespass upon the inter-citizen domain. The possessing classes continued to be juridically guaranteed in their property by the precepts established in the Republic. Beneath them, criminal law – essentially designed for the lower classes – remained as arbitrary and repressive as it had always been: a social safeguard for the whole ruling order. The Principate thus preserved the classical legal system of Rome, while superimposing on it the new innovatory powers of the Emperor in the realm of public law. Ulpian was later to formulate the distinction which articulated the whole juridical corpus under the Empire with characteristic clarity: private law – quod ad singulorum utilitatem pertinet – was in particular separate from public law – quod ad statum rei Romanae spectat. The former suffered no real eclipse from the extension of the latter.²⁸ It was, indeed, the Empire which produced the great systematizations of civil jurisprudence in the 3rd century, in the work of the Severan prefects Papinian, Ulpian and Paulus, that transmitted Roman law as a codified body to later ages. The solidity and stability of the Roman imperial State, so different from anything the Hellenistic world produced, was rooted in this heritage.

The subsequent history of the Principate was largely that of the increasing ‘provincialization’ of central power within the Empire. Once the monopoly of central political office enjoyed by the Roman aristocracy proper had been broken, a gradual process of diffusion integrated a wider and wider ambit of the Western landed classes outside Italy itself into the imperial system.²⁹ The origin of the successive dynasties of the Principate was a straightforward record of this evolution. The Roman patrician Julio-Claudian house (Augustus to Nero) was followed by the Italian municipal Flavian line (Vespasian to Domitian); succession then passed a series of Emperors with a provincial Spanish or Southern Gallic background (Trajan to Marcus Aurelius). Spain and Gallia Narbonensis were the oldest Roman conquests in the West, whose social structure was consequently closest to that of Italy itself. The composition of the Senate reflected much the same pattern, with a growing intake of rural dignitaries from Transpadane Italy, Southern Gaul and Mediterranean Spain. The imperial unification of which Alexander had once dreamed appeared symbolically accomplished by the epoch of Hadrian, the first Emperor to tour his whole immense domain from end to end in person. It was formally consummated by Caracalla’s decree of A.D. 212 granting Roman citizenship to nearly all free inhabitants throughout the Empire. Political and administrative unification was matched by external security and economic prosperity. The Dacian kingdom was conquered and its gold mines annexed; the Asian frontiers were extended and consolidated. Agricultural and artisanal techniques slightly improved: screw presses promoted oil production, kneading-machines facilitated the manufacture of bread, glass-blowing became widespread.³⁰ Above all, the new pax romana was accompanied by a buoyant surge of municipal rivalry and urban embellishments in virtually all the provinces of the Empire, exploiting the architectural discovery by Rome of the arch and the vault. The Antonine epoch was perhaps the peak period for city construction in Antiquity. Economic growth was accompanied by the flowering of Latin culture in the Principate, when poetry, history and philosophy blossomed after the relative intellectual and aesthetic austerity of the early Republic. To the Enlightenment this was a Golden Age, in Gibbon’s words ‘the period in the history of the world during which the condition of the human race was most happy and prosperous’.³¹

For some two centuries, the tranquil magnificence of the urban civilization of the Roman Empire concealed the underlying limits and strains of the productive basis on which it rested. For, unlike the feudal economy which succeeded it, the slave mode of production of Antiquity possessed no natural, internal mechanism of self-reproduction, because its labour-force could never be homeostatically stabilized within the system. Traditionally, the supply of slaves largely depended on foreign conquests, since prisoners of war probably always provided the main source of servile labour in Antiquity. The Republic had sacked the whole Mediterranean for its manpower, to install the Roman imperial system. The Principate halted further expansion, in the three available remaining sectors of possible advance, Germany, Dacia and Mesopotamia. With the final closure of the imperial frontiers after Trajan, the well of war captives inevitably dried up. The commercial slave-trade could not make up for the shortages that resulted, since it had always itself been largely parasitic on military operations for its stocks. The barbarian periphery along the Empire continued to provide slaves, bought by dealers at the frontiers, but not in sufficient numbers to solve the supply problem in conditions of peace. The result was that prices started to climb sharply upwards; by the 1st and 2nd centuries A.D., they were 8 to 10 times the levels of the 2nd and 1st centuries B.C.³² This steep rise in costs increasingly exposed the contradictions and risks of slave labour for its proprietors. For each adult slave represented a perishable capital investment for the slave-owner, which had to be written off in toto at death, so that the renewal of forced labour (unlike wage-labour) demanded a heavy preliminary outlay in what had become an increasingly tight market. For, as Marx pointed out, ‘the capital paid for the purchase of a slave does not belong to the capital by means of which profit, surplus-labour, is extracted from him. On the contrary, it is capital which the slave-owner has parted with, it is a deduction from the capital which he has available for actual production.’³³ In addition, of course, the maintenance of slave offspring was an unproductive financial charge on the owner, which inevitably tended to be minimized or neglected. Agricultural slaves were housed in barrack-like ergastula, in conditions approximating to those of rural prisons. Female slaves were few, being generally unprofitable to owners because there was a lack of ready employment for them, beyond domestic tasks.³⁴ Hence the sexual composition of the rural slave-population was always drastically lopsided, and was compounded by the virtual absence of conjugality within it. The result must have been a customarily low rate of reproduction, which would have diminished the size of the labour-force from generation to generation.³⁵ To counteract this fall, slave-breeding seems to have been increasingly practised by landowners in the later Principate, who granted premia to women-slaves for child-bearing.³⁶ Although there is little evidence as to the scale of slave-breeding in the Empire, it may have been this resort which for a time mitigated the crisis in the whole mode of production after the closure of the frontiers: but it could not provide a long-term solution to it. Nor, meanwhile, was the rural free population increasing, to compensate for losses in the slave sector. Imperial anxieties about the demographic situation in the countryside were revealed as early as Trajan, who instituted public loans to landowners for the upkeep of local orphans, an omen of shortages to come.

Nor could the dwindling volume of labour be compensated by increases in its productivity. Slave agriculture in the late Republic and early Empire was more rational and profitable to landowners than any other form of exploitation of the soil, partly because slaves could be utilized full-time, where tenants were unproductive for considerable stretches of the year.³⁷ Cato and Columella carefully enumerate all the different indoor and out of season tasks to which they could be set when there were no fields to be tilled or crops to be gathered. Slave artisans were just as proficient as free craftsmen, since they tended to determine the general level of skills in any trade by their employment in it. On the other hand, not only did the efficiency of the latifundia depend on the quality of their vilicus bailiffs (always the weak link in the fundus), but supervision of slave-labourers was notoriously difficult in the more extensive cereal crops.³⁸ Above all, however, certain inherent limits of slave productivity could never be overcome. The slave mode of production was by no means devoid of technical progress; as we have seen, its extensive ascent in the West was marked by some significant agricultural innovations, in particular the introduction of the rotary mill and the screw press. But its dynamic was a very restricted one, since it rested essentially on the annexation of labour rather than the exploitation of land or the accumulation of capital; thus unlike either the feudal or capitalist modes of production which were to succeed it, the slave mode of production possessed very little objective impetus for technological advance, since its labour-additive type of growth constituted a structural field ultimately resistant to technical innovations, although not initially exclusive of them. Thus, while it is not wholly accurate to say that Alexandrine technology remained the unchanging basis of work-processes in the Roman Empire, or that no labour-saving devices of any kind were ever introduced in the four centuries of its existence, the boundaries of the Roman agrarian economy were soon reached, and rigidly fixed.

The insuperable social obstacles to further technical progress – and the fundamental limits of the slave mode of production – were, in fact, strikingly illustrated by the fate of the two major inventions which were recorded under the Principate: the water-mill (Palestine at the turn of the 1st century A.D.) and the reaping machine (Gaul in the 1st century A.D.). The immense potential of the water-mill – basic to later feudal agriculture – is evident enough: for it represented the first applied use of inorganic power in economic production: as Marx commented, with its appearance, ‘the Roman Empire handed down the elementary form of all machinery in the water-wheel’.³⁹ The Empire, however, made no general use of the invention itself. It was in practice ignored under the Principate; in the later Empire, its incidence was somewhat more frequent, but it never seems to have become a normal instrument of Ancient agriculture. Likewise, the wheeled harvester introduced to accelerate reaping in the rainy climate of the North, was never adopted on any scale outside Gaul.⁴⁰ Here, the lack of interest was a reflection of a wider failure to alter the methods of Mediterranean dry-soil agriculture – with its scratch-plough and two-field system – on the heavy, moist lands of Northern Europe, which needed new instruments of labour to be fully exploited. Both cases amply demonstrate that mere technique itself is never a prime mover of economic change: inventions by individuals can remain isolated for centuries, so long as the social relations have not emerged which alone can set them to work as a collective technology. The slave mode of production had little space or time for the mill or the harvester: Roman agriculture as a whole remained innocent of them to the end. Significantly, the only major treatises of applied invention or technique to have survived from the Roman Empire were military or architectural – designed essentially for its complex of weaponry and fortifications, and its repertoire of civic ornamentation.

There was no urban salvation for the malady of the countryside, however. The Principate presided over an unprecedented range of city-building in the Mediterranean. But the quantitative expansion in the number of large and medium towns in the first two centuries of the Empire was never accompanied by any qualitative modification of the structure of overall production within it. Neither industry nor trade could ever accumulate capital or experience growth beyond the strict limits set by the economy of classical Antiquity as a whole. The regionalization of manufactures, because of transport costs, thwarted any industrial concentration and development of a more advanced division of labour in manufactures. A population overwhelmingly composed of subsistence peasants, slave-labourers and urban paupers narrowed consumer markets down to a very slender scale. Apart from the tax-farms and public contracts of the Republican epoch (whose role declined greatly in the Principate, after the Augustan fiscal reforms), no commercial companies ever developed, and funded debts did not exist: the credit system remained rudimentary. The propertied classes maintained their traditional disdain for trade. Merchants were a despised category, frequently recruited from freedmen. For manumission of administrative and domestic slaves remained a widespread practice, regularly thinning the higher ranks of the urban slave population; while the contraction of external supplies must have gradually diminished the stock of servile artisans in the cities. The economic vitality of the towns was always limited and derivative: its course reflected rather than offset that of the countryside. There were no civic springs for a reversal of the relationship between the two. Moreover, once the Principate had been consolidated, the character of the imperial State apparatus itself tended to stunt commercial enterprise. For the State was by far the largest single consumer of the Empire, and was the one real focus for mass production of necessities, which might have given birth to a dynamic manufacturing sector. The provisioning policy and peculiar structure of the imperial State precluded this, however. Throughout classical Antiquity, ordinary public works – roads, buildings, aqueducts, drains – were typically executed by slave-labour. The Roman Empire, with its massively enlarged State machine, saw a corresponding extension of this principle: for the entire armaments and a considerable proportion of the procurements supply for its military and civilian apparatus came to be furnished autarchically by its own industries, manned by sub-military personnel or hereditary state slaves.⁴¹ The one truly large-scale manufacturing sector was thus to a great extent subtracted from commodity exchange altogether. The permanent and direct use by the Roman State of slave-labour – a structural feature which lasted right down and into the Byzantine Empire – was one of the central pillars of the political economy of late Antiquity. The infrastructure of slavery found one of its most concentrated expressions within the imperial superstructure itself. Thus the State could expand, but the urban economy received little benefit from its growth: if anything, its size and weight tended to suffocate private commercial initiative and entrepreneurial activity. There was thus no increase of production in either agriculture or industry within the imperial borders to offset the silent decline in its servile manpower, once external expansion had ceased.⁴²

The result was an incipient crisis in the whole economic and social system, by the early 3rd century, that soon developed into a pervasive breakdown of the traditional political order, amidst fierce external attacks on the Empire. The sudden dearth of sources, one of the symptoms of the crisis of the mid 3rd century, makes it very difficult retrospectively to trace its exact course or mechanisms.⁴³ It looks as if serious strains were already surfacing in the closing years of the Antonine age. Germanic pressure on the Danubian frontiers had led to the lengthy Marcomannic wars; the silver denarius had been devalued by 25 per cent by Marcus Aurelius; the first major outbreak of social brigandage had erupted, with the menacing seizure of large regions of Gaul and Spain by the armed bands of the deserter Maternus, who even sought to invade Italy itself during the disastrous reign of Commodus.⁴⁴ The accession, after a brief civil war, of the Severan house brought an African dynasty to power: regional rotation of the imperial office seemed to have functioned once again, as civic order and prosperity were apparently restored. But soon inflation became mysteriously rampant, as the currency was devalued again and again. By mid century, there was a complete collapse of the silver coinage, which reduced the denarius to 5 per cent of its traditional value, while by the end of the century, corn prices had rocketed to levels 200 times over their rates in the early Principate.⁴⁵ Political stability degenerated apace with monetary stability. In the chaotic fifty years from 235 to 284, there were no less than 20 Emperors, eighteen of whom died violent deaths, one a captive abroad, the other a victim of the plague – all fates expressive of the times. Civil wars and usurpations were virtually uninterrupted, from Maximinus Thrax to Diocletian. They were compounded by a devastating sequence of foreign invasions and attacks along the frontiers, stabbing deep into the interior. Franks and other Germanic tribes repeatedly ravaged Gaul, sacking their way into Spain; Alamanni and Iuthungi marched into Italy; Carpi raided into Dacia and Moesia; Heruli overran Thrace and Greece; Goths crossed the sea to pillage Asia Minor; Sassanid Persia occupied Cilicia, Cappadocia and Syria; Palmyra detached Egypt; Mauri and Blemmyes nomads harassed North Africa. Athens, Antioch and Alexandria at different moments fell into enemy hands; Paris and Tarragona were put to the torch; Rome itself had to be refortified. Domestic political turmoil and foreign invasions soon brought successive epidemics in their train, weakening and reducing the populations of the Empire, already diminished by the destructions of war. Lands were deserted, and supply shortages in agrarian output developed.⁴⁶ The tax system disintegrated with the depreciation of the currency, and fiscal dues reverted to deliveries in kind. City construction came to an abrupt halt, archaeologically attested throughout the Empire; in some regions, urban centres withered and contracted.⁴⁷ In Gaul, where a breakaway imperial state was maintained with a capital at Trier for fifteen years, there were full-scale rural uprisings by the exploited masses in 283–4, the first of the Bacaudae insurrections which were to recur in the history of the Western provinces. Under intense internal and external pressure, for some fifty years – from 235 to 284 – Roman society looked as if it might collapse.

By the late 3rd and early 4th century, however, the imperial state had altered and recovered. Military security was gradually restored by a martial series of Danubian and Balkan generals who successively seized the purple: Claudius II routed the Goths in Moesia, Aurelian drove back the Alamanni from Italy and subdued Palmyra, Probus annihilated the Germanic invaders of Gaul. These successes paved the way for the reorganization of the whole structure of the Roman State in the epoch of Diocletian, proclaimed Emperor in 284, which were to make possible the precarious revival of the next hundred years. First and foremost, the imperial armies were massively enlarged by the reintroduction of conscription: the number of legions was effectively doubled in the course of the century, bringing total troops strengths up to over 450,000 or so. From the late 2nd and early 3rd century onwards, increasing numbers of soldiers were stationed in guard-posts along highways to maintain internal security and police the countryside.⁴⁸ Later, from the time of Gallienus in the 260’s onwards, crack field armies were increasingly redeployed in depth behind the imperial frontiers for greater mobility against external attacks, leaving second-class limitanei units to guard the outer perimeter of the Empire. Large numbers of barbarian volunteers were incorporated into the Army, henceforward providing many of its elite regiments. More important, all top military commands were now entrusted to men of equestrian rank only; the senatorial aristocracy was thereby displaced from its traditionally pivotal role in the political system, as supreme imperial power increasingly passed to the professional officer corps of the army. Diocletian himself also systematically shut senators out of the civil administration.⁴⁹ Provinces multiplied twice over, as they were divided into smaller and more manageable units, and the officialdom set over them was increased proportionately, to create closer bureaucratic control. A new fiscal system was established, after the debacle of the mid century, fusing the principles of a land and poll tax into a single unit, calculated on the basis of new and comprehensive censuses. Annual budgetary estimates were introduced for the first time in the Ancient World, which could adjust tax levels to current expenditures – that predictably moved steeply upwards. The formidable material expansion of the State machine that was the result of all these measures inevitably contradicted the ideological attempts of Diocletian and his successors to stabilize the social structure of the later Empire beneath it. Decrees penning large groups of the population into caste-like hereditary guilds after the turbulence of the past half century can have had little practical efficacy;⁵⁰ social mobility probably increased somewhat, because of the enlargement of new military and bureaucratic avenues of promotion within the State.⁵¹ Transient efforts to fix administrative prices and wages throughout the Empire were even less realistic. On the other hand, the imperial autocracy itself successfully prised away all traditional restraints imposed by senatorial opinion and custom on the exercise of personal power. The ‘Principate’ gave way to the ‘Dominate’, as Emperors from Aurelian onwards styled themselves dominus et deus and enforced the Oriental ceremonial of full-length abasement before the royal presence, the proskynesis with which Alexander had once inaugurated the Hellenistic Empires of the Near East.