Empires: Perspectives from Archaeology and History

EMPIRES Perspectives [roIn Archaeology and History Edited by

Susan E. Alcock Terence N. D'Altroy I(athleen D. Morrison Carla M. Sinopoli

•

V

CAMBRIDGE UNIVERSITY PRESS

]-C ~SC\

L((b

ANTHROPOLOGY LlBRARt

?CC I

PUBLISHED BY THE PRESS SYNDICATE OF THE UNIVERSITY OF CAMBRIDGE

t if i

The Pitt Building, Trumpington Street, Cambridge, United Kingdom CAMBRIDGE UNIVERSITY PRESS

The Edinburgh Building, Cambridge CB2 2RU, UK 40 West 20th Street, New York, NY 10011-4211, USA 10 Stamford Road, Oakleigh, VIC 3166, Australia Ruiz de Alarc6n 13,28014 Madrid, Spain Dock House, The Waterfront, Cape Town 8001, South Africa http://www.cambridge.org © Cambridge University Press 2001

This book is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place witholll the written permission of Cambridge Unil.ersity Press. First published 2001 Printed in the United Kingdom at the University Press, Cambridge

Typcfllcc Galliard 10113pt

System QuarkXPress'"

[SE

I

A catllloglle reco/'d fo/' tbis book is /JI1IJi/llble fi'olll tbe n"itisb Libl'll/,Y Libl'll/,Y ofCollg/'css Clltlllogllillg ill Pllblicntioll dlltll Empires 1 edited by Susan E. Alcock ... [et Ill. J p. cm. Includes bibliographical references. ISBN 0521 770203 (hb) 1. Imperialism - History. I. Alcock, Susan E. JC359.E46 2001

930-fate formation along the Chinese-Nomad frontier

In addition to grain, ordinary nomads were also interested in other goods for use, particularly cloth and metals. Linen clotll (from hemp, ramie, or kudzu) and cotton (of Central Asian origin) were a useful supplement to tlle felts and woolens produced by the nomads. Perhaps not absolutely necessary, tlley nevertheless made life easier and were much in demand (anyone who has ever experienced tlle discomfort of trying to dry steaming wet wool after a rainstorm can appreciate tlle value of a change of faster-drying clothes!). Metals were tlle most strategic of tlle common goods needed by tlle nomads. Iron was in particularly high demand on tlle steppe for tools and weapons, but China was also a good source of bronze and copper. The nomads had tlleir own tradition of metallurgy, of course, but the Chinese produced bronze and iron in such quantity tllat it was much simpler to acquire metal 6:om tllem and rework it on tlle steppe tllan to have to mine and refine tlle ore tllemselves. Chinese coins were a particularly good source of metal because when tlley went out of circulation they could be purchased in bulk even tllOugh tllis trade was often prohibited. In Han times trade in iron to tlle nomads was absolutely banned and violators were punished by death, a penalty inflicted on 500 merchants in tlle capital in 121 B CE when tlley (in ignorance oftlle law) sold iron pots to a delegation of visiting nomads (Yi.i 1967: 119). Border markets where the nomad trade flourished were characteristic of China's frontier areas. The relationship was a natural one and profitable to both sides, but the Chinese insisted that these markets be closely regulated and during many periods they prohibited all trade with the steppe peoples. This was because native dynasties feared they would lose control of their frontier areas iftlley were too closely integrated into the steppe economy. Banning trade, however, only caused tlle nomads to get what they wanted by force and turned the frontier into a zone of endemic raiding. Jagchid and Symons (1989) go so far as to argue that peace or war between China and the steppe depended entirely on whether China permitted such markets to be opened. While this argument is too simplistic, it does capture the fact that the status of border trade in ordinary goods was a key point of dispute between the nomads of Mongolia and China over the centuries.

Long-distance trade and tribute in luxury goods If trade in ordinary goods made up the bulk of exchanges, it was the acquisition of luxury goods (silk, gold, wine, etc.) that is the focus of most of the written histories. Why the nomad preoccupation with these luxury goods, especially silk? First, silk was a valuable luxury good. Although not well suited for steppe life, it was a highly valued sign of elite status (Fig. 1.5). Perhaps more important, it was a store of wealth, light in weight and high in value, that could be traded for more utilitarian objects or luxury goods from elsewhere. Silk was not the only product the nomads demanded. Gold, satin, precious metals, bronze mirrors, and even musical instruments appear on lists of gifts. And alcohol (mostly as rice wine), or yeast to make it, was a common trade item and given in large quantities as gifts

19

20

J,5 17)c Xio1JlJlIII i1Jcorporatcd IIl1l11y jiJ/'Cigll goods into the stCjljlC, II'''crc they beClllllc part (if' cl'cI'yda), I~f'c, 77)e cOlllbination {if' Chincsc sill, i1Jto 11 shccpsldn IJllt lI'om by t1'l1diti01w/ly d/'c.ued [(azl1l, 1IICtI in 1I00,tlJll'cstc/'ll Chinl1 today is nil cXllmplc lif' thc IOlllJ-tcl'/IJ "istoric relatioll bctll'een Chinll Ilnd t"c steppe 1101IJIlds, (Altlli MOlllltains, XilljianlJ, Pcoples Republic lif' Chinll, JJI~)' J 987.)

Thomas J. Bm/icld

Iwperial .Itate /imlt atio 11 tllolllf the Chinese-Nolllad /hmtier to nomad leaders, Throughout steppe history, the nomads' appetite t(J(' alcohol was legendary and was associated with excess (Jagchid and Hyer 1979: 42-3), which led the Greeks to refer to drinking wine unmixed with water as "Scythian style" (Herodotus, The Hi.,tor)" 6,84,3), The Chinese had strong reservations about the export of luxury goods generated by tributary gifts because the Han court lost revenue in these transactions, Even private trade with distant foreign nations, it was argued, only served to drain China's real wealth t(J(' the acquisition of expensive luxury items whose only value was their rarity, This bias against international trade was closely tied to a Confucian ideology that condemned merchants as a class t(J(' being leeches on the hard work of peasants and artisans, vVhy, they argued, should a merchant deserve any more than the cost of transport t(X movi ng goods from onc place to another? That they could make phenomenal profits by moving merchandise from areas of surplus to areas of shortage was not an entrepreneurial virtue but a defect in the government administration that had allowed such shortages to arise in the first place, This is not to say that trade was unimportant in China, Many individuals, including government officials, profited li'om trade and throughout the Han period there were wealthy merchant f:1I11ilies that controlled large enterprises, However, at the elite level such enterprises wcre deemed illegitimate, and merchants were thereforc t()I'bidden /i'om competing in the imperial examinations which led to high officc and their wealth was subjcct to arbitrary confiscation, (A modern analogy that captures the flavor of this official distaste might be the public attitude toward I:lbulously wealthy drug slllugglers whose businesses and morals arc condemned evcn as their money is welcomcd,) Ideally China should be autarkic: sell~sullicient and sell~contained, The existcnce of a merchant class was evidence that it had I:liled to reach this idcal. Such attitudes stood in sharp contrast to those o( the nomad elites in Mongolia, They actively encouraged trade and attempted to attract merchants into their territories because the pastoral cconomy was not sell~sunicie11l except in terms o(sheep 01' horses, Far /i'om viewing export trade as a drain on national wealth, nomads saw it as a source o( prosperity and stability, Thc dcmand I(lr trade and the extortion of luxury items incrcased exponen-tially with the unification o(the steppe, As a I'cdistributive chieftain, thc Xiongnu Shrt1I)'II'S powcr was securcd in large part by his ability to generate revenue fl'om China and secure trading privileges there, But the nomads' increasing demands ICll' such luxury goods, particularly silk, were not simply I(lr their own use, Once they had acquired a surplus o( these valuable commodities the nomads in Mongolia became thc center o( an international reexport trade which attracted traders, especially fl'om thc oases of Central Asia, who became wealthy middlemen linking the economics o( China and the West. Indeed, when China under Emperor Wudi did expand into Central Asia it was to "cut off thc right ann of thc Xiongnu)) by stopping the revcnue the Xiongnu derivcd fl'om the city-states of Turkestan (Hulsewc 1979: 217), The wealth of items ffate formation along the Chinese-Nomad frontier raiding China] and restore our old pact, that the peoples ofthe border may have the peace such as they enjoyed in former times, that the young may grow to manhood, the old live out their lives in security, and generation after generation enjoy peace and comfort. (Ssu-ma 1993,2: 140-1)

The Han court decided the Xiongnu were far too powerful to attack and so agreed to renew the treaty and open border markets. Maodun died peacefully in 174 BCE, leaving his large steppe empire to his son. Mter Maodun's death, the Xiongnu made greater access to these regular markets their key demand, for the hoqin subsidy payments, although very profitable for the political elite, could not adequately compensate the much larger number of ordinary nomads who were forced to forgo raiding. Without the guarantee of regular access to border markets where ordinary nomads could trade live animals or other pastoral products for grain, cloth, or metal, the Shanyu could not expect his people to observe the peace. Since the Chinese feared that such widespread economic links between their own frontier people and the nomads would lead to political subversion, the Han court was opposed to any increase in the size or number of border markets. The Xiongnu were therefore forced to extort increased trade privileges the same way they extorted increased subsidies: by raiding or threatening to raid China. Loot from such raids kept the Xiongnu tribesmen supplied until China finally agreed to liberalize its trade policy. Once established, these border markets quickly became important trade centers to which the Xiongnu flocked, exchanging pastoral products for Chinese goods. Now instead of prohibiting trade, the Han court attempted to control it by regulating what items could be sold, as well as the location and timing of trade fairs. The whole relationship between China and the nomads became more stable and old hostilities were t()rgotten: "From the Shanyu on down, all the Xiongnu grew friendly with the Han, coming and going along the Great Wall" (Ssu-ma 1993,2: 148). This situation lasted until 133 BeE when, under the aggressive leadership of Emperor Wudi, the "Martial Emperor," the Han court abruptly abandoned the hoqin policy and mounted a surprise attack on the nomads, beginning more than a half~century of f1'ontier warfare. Although the hoqin policy had successfully preserved the peace for three generations, it was always unpopular at the Han court because treating the Xiongnu as an equal state violated the very essence of a Sinocentric world order, a view well expressed earlier by Jia Vi, an official at the court of Emperor Wen (r. 179-157 BCE): The situation of the empire may be described just like a person hanging upside down. The Son off-leaven is at the head of the empire. Why? Because he should be placed at the top. The barbarians are at the tcet of the empire. Why? Because they should be placed at the bonom ... To command the barbarian is a power vested in the Emperor at the top, and to present tribute to the Son of Heaven is a ril"llal to be perlormed by vassals at the bottom. Now the tcet are pm on the top and the head at the bottom. Hanging upside down is something beyond comprehension. (Yii 1967: 11)

25

26

Thomas J. Barfield

Emperor Wudi was susceptible to such criticism. Although his predecessors considered it expedient to appease the Xiongnu to avoid trouble, he now considered such a policy demeaning. China was an empire unrivaled in wealth and military power that would invade the steppe, defeat the Xiongnu and destroy their power forever. Wudi's wars failed. Although China sent a series of massive expeditionary forces against the Xiongnu, they found nothing to conquer but empty land. Lack of supplies forced Chinese armies to retreat within a few months of each campaign. Although they occasionally defeated the nomads in battle and engineered the defection of some of tile Xiongnu empire's component tribes, tile cost of these wars in men, horses, and money was so high that tile dynasty practically bankrupted itself. Nor did tile attacks destroy tile stability of tile Xiongnu empire; ironically the invasions only reinforced tile Shanyu's position as protector of tile nomads against Chinese aggression. Mter decades of war, tile Han court reluctantly concluded China had no more chance of ruling the nomads of the steppe tllan tlley had of governing the fish in tile sea. By 90 BC E they had abandoned tlleir attacks on the steppe and adopted a completely defensive position of cutting off trade while repulsing raids (Loewe 1974a). The Xiongnu had long understood tllat the disruption of peaceful relations over tile long term worked to tlleir disadvantage so tlU'oughout the war they had sent envoys to China requesting a resumption of the hoqin treaties as a way to restore tile status quo ante. But China had rejected such peace offers, insisting tllat any new peace agreement take place within tile new framework of a "tributary system" in which, tlley told the nomads, the Xiongnu would be required to pay homage to tile Han emperor, send a hostage to court, and pay tribute to China. It was a relationship the Xiongnu considered unacceptable and explicitly rejected in 107 BC E: "That is not the way things were done under the old alliance!" the ShanYIJ objected, "Under the old alliance the Han always sent us an imperial princess, as well as allotments of silks, foodstuffs, and other goods, in order to secure peace, while we for our part refrained from making trouble at the border. Now you want to go against the old ways and make me send my son as hostage. I have no use for such proposals!" (Ssu-ma 1993,2: 157)

Yet with no subsidies, no trade, and borders too strong to raid, successive Xiongnu Shanyus found tlleir political positions undermined. In 60 BC E a succession dispute split tile Xiongnu elite into rival f.1ctions who warred upon onc another. The losing Shanyu, Huhanyeh, decided that his only chance of political survival was to come to terms Witll China and so he broke with Xiongnu tradition in 53 B CE by agreeing to accept the Chinese demands for peace under tile terms of the tributary system. Surprisingly, the tributary system proved a sham. In return for formal compliance, the Xiongnu received even larger gifts and better border markets. During his first visit to the Han court in 51 BC E, Huhanyeh received twenty jin of gold, 200,000 cash, seventy-seven suits of

Imperial state formation along the Chinese-Nomad frontier

clotlles, 8000 pieces of silk, and 6000 jin of silk floss; his followers were supplied with 34,000 hu ofrice2 (Wylie 1875: 44-7). The discovery of tile true nature of tile Han tributary system allowed Huhanyeh to implement a new "inner frontier" strategy in steppe politics. In essence he used Han wealtll and military protection to re establish unity witllin tile Xiongnu empire. This strategy differed from outright surrender to China, in which a tribal leader accepted Chinese titles and entered tile Han administrative framework, disappearing from tile steppe political scene. Instead Huhanyeh maintained his autonomy and avoided direct Chinese control while demanding foreign aid and even military assistance to defeat rival Xiongnu leaders. The Chinese were eager to support contenders in a civil war ("using barbarians to fight barbarians"), a policy always popular at tile Han court witll the expectation tllat by aiding tile winning side they would be able dominate their ally in the future. While in tile short term such goals could be realized, in the long term Chinese aid simply enabled the nomads to rebuild tlleir empire and return to tlleir aggressive outer frontier strategy once again. In 43 BC E, after a decade of receiving Chinese aid, Huhanyeh did just this and returned nortll to his homeland as supreme ruler of the Xiongnu, "and his people all gradually came together from various quarters, so that the old country again became settled and tranquil" (Wylie 1875: 47-8). Even though they regained their unity and power, the Xiongnu never again objected to the structure of the tributary system. Instead, they actively set about exploiting it for their own ends. They continually demanded the right to present "tribute" and send hostages to court because they profited so handsomely. Indeed, they threatened invasion if their tributary missions were not received and appropriately rewarded. For the remainder of the Former Han dynasty Xiongnu regularly visited the Chinese court, with each Shanyu generally making at least one visit during his reign. And with each visit the amount of gifts increased (Table 1.2; also Yii 1967: 47). The policy oflavish tributary payments continued into the Later Han dynasty (25-220 CE) and expanded to include other newly powerful fi'ontier tribal groups like the Xianbei, Wuhuan, and Qiang. By 50 C E when the system was regularized, it is estimated that the annual cost of direct subsidies to the nomads amounted to one-third of the Han government payroll or 7 percent of all the empire's revenue, goods to the value of $130 million dollars in modern terms (Yi.i 1967: 61-4). This fragmentation of the tributary system in the Later Han was a consequence of a second civil war that permanently divided the Xiongnu into northern and southern branches beginning in 47 CE. As in the previous civil war, the southern Shanyu allied himselfwith China and employed the inner frontier strategy, using China's wealth to defeat his rival. However, because this civil war lasted more than forty years, the victorious lineage of the southern Shanyu was unable to reassert its control over northern Mongolia which fell into the hands of the rival Xianbei nomads. Instead, the southern Shanyu maintained his close

27

28

Imperial state formation along the Chinese-Nomad frontier

Thomas]. Barfield Table 1.2. Xiongnu ),isits to the Han Chinese court Year of visit BeE

Silk floss(jill)

Silk f.1bric (pieces)

51 49

6,000 8,000 16,000 20,000 30,000

8,000 9,000 16,000 20,000 30,000

33 25 1

connection with China and was content to leave the steppe fragmented, maintaining control only around the immediate frontier in order to dominate the flow of goods to the steppe and keep the less well-organized nomads from gaining access to the system. By the end of the second century CE, the relationship between the Han dynasty and the southern Xiongnu had become so close that they acted as "fi'ontier guarding barbarians," protecting China from attacks by other tribes on the steppe and, not coincidentally, milking the dynasty for more subsidies. Although dming the second century the southern Xiongnu became so closely tied to the Han court that they fell under the indirect control ofHan frontier officials who could determine succession to leadership by supporting favored candidates, they never lost their identity as an independent state. So important was this relationship that it was the nomads who provided the last bulwark against domestic rebels when China fell into civil war in 180 CE. But because nomad empires were dependent on a prosperous and stable Chinese dynasty, they could not survive its collapse. When the Han dynasty finally dissolved in 220 CE, China's economy and population were devastated. The nomads no longer had any rich provinces to loot, had no dependable border markets in which to trade, and saw their subsidy payments disappear. Under such conditions centralization proved impossible and the tribes in Mongolia reverted to anarchy. Thus an empire as powerful and centralized as that of the Xiongnu would not reemerge for another three hundred years until the Turks were able to exploit a reunifying China under the Sui and Tang dynasties in the sixth century to establish a relationship stl'l\cturally analogous to that of the Han and Xiongnu. Like that relationship, unity on the steppe also disappeared with the collapse of the Tang dynasty at the end of the ninth century (Barfield 1989).

EMPIRES AND SHADOW EMPIRES When comparing the Xiongnu empire with Han China, an immediate question arises. Can states that are so different both be empires? They were certainly politically comparable and both ruled over vast territories. But in most other ways (sophistication of administration, political centralization, urbanization, size of population, economic specialization, etc.), they bore almost no similarities. This

was because steppe empires were secondary phenomena, arising in response to imperial state formation in China and largely dependent on exploiting a unified China for their existence. However, the Xiongnu empire is not a unique case. Looking at the other examples of empires discussed in this volume, we find a similar sharp divide between those that most scholars would easily accept as empires and those that are in some way problematic. The "primary empires" would include Assyria, Achaemenid Persia, Rome, China, Inka, Aztec (though this was still in its formative stage when destroyed by the Spaniards), Spanish, and Ottoman. The problematic cases are the Xiongnu, Nubia, Carolingian Europe, and the Portuguese Indies. But (to paraphrase Tolstoy) while all primary empires are alike, each problematic empire is problematic in its own way. In some respects, they are out\vardly empires, but each is missing something vital that sets it apart and makes it a shadow empire.

Primary empires What is an empire and how does it differ from other types of polities? An empire is a state established by conquest that has sovereignty over subcontinental or continental sized territories and incorporates millions or tens of millions of people within a unified and centralized administrative system. The state supports itself through a system of tribute or direct taxation of its component parts and maintains a large permanent military force to protect its marked fi'ontiers and preserve internal order. Empires also share a set of five common internal characteristics:

First, empires are ot;ganized both to administer and exploit diJ1ersity, whether economic, political, religious, or ethnic. While empires may begin with the hegemony of a single region or ethnic group, they all grow more cosmopolitan over time with the incorporation of new territories and people very different from themselves. Indeed it is characteristic that, once established, the elite of an empire may change or be replaced without the necessary collapse of the state structure. Egypt's many dynasties are a notable example of this, as was the tendency of Roman emperors to be drawn from non-Italian regions after the end of Augustus' line. Even in China one finds that dynasties that drew their elite from one region during the formation of a new empire moved to broaden their base to countrywide recruitment within a couple of generations. It is their ability to incorporate large numbers of different ethnic, regional, and religious groups that makes empires so different fi'om tribes and locality-based polities such as citystates that organize themselves on the basis of some common similarity. Empires are comfortable with, and even thrive on, diversity. The f.1mous frieze at Persepolis with all the Persian empire's many component satrapies lined up before the Great King in their native dress to present tribute of distinctive local products is a physical representation of this diversity. Of course this was not because empires thought well of peoples different from themselves, but rather that their policies were designed to make all groups integral parts of the empire

29

30

Thomas]. Barfield

and share in a common political order. It was no missionary enterprise. The first steps were often brutal: to conquer and destroy any groups that opposed them militarily or politically. This was accomplished by wholesale population transfers of conquered populations to distant parts of the empire, settlement of imperial troops and immigrants in new territories, and tlle destruction (or incorporation) oftlle group's indigenous political elite. Once rival groups were politically neutralized, empires displayed a high tolerance for local variation (as long of course as it did not interfere Witll good order and tlle collection of taxes). The success of tllis policy can be seen in tlle decline of troop concentrations and military fortifications within tlle empire and tlleir transfer to tlle frontier to confront outsiders. Second, empires established transportation systems designed to serve the imperial center militarily and economically. Sophisticated and well-maintained transportation systems were characteristic of all great empires. No place in the empire was beyond the center's military reach and supplies needed for military operations were stored tllfoughout tlle empire. The Persian, Roman, Chinese, and Inka road systems were particularly impressive in size. In maritime or riverine environments, empires made large investments in port f.'lcilities to exploit sea trade or canal systems to link rivers for bulk inland transport. In arid landlocked areas tlley created and maintained systems of caravansarai and river crossings that facilitated caravan trade. In all cases, the amount of investment in these infrastructures was well beyond the capacity of any single part of the empire to finance on its own, particularly in sparsely populated marginal territories. Economically this transportation network and centralized revenue collection allowed imperial centers to support a population and level of sophistication well beyond the capacity of its local hinterland. Where bulk transportation of commodities was possible, the empire's capital (or sometimes, dual capitals) became a mega-center that dwarfed all other urban centers. Here water transport was key. Roman imports of grain from Egypt and North Mrica were vital to the capital's survival. China invested heavily in a series of canals that linked the north and the south of the country. Aztec success in conquering the lake regions of central Mexico underlay the rapid growth of their capital to a size that astonished the Spanish. Areas that did not have access to such water transport focused on overland transport such as caravan routes using pack animals. This allowed the circulation of high quantities of goods (particularly luxury items), but was not enough to support the mega-city capitals that had access to maritime or canal commerce. Capitals such as the Inka's Cuzco or Assyria's Nineveh were architecturally impressive as imperial centers and covered large areas, but they were subject to much more strict upper limits on their size because they were supported mostly by the agricultural surplus of their local hinterlands. In economic terms, empires provided large-scale economic integration tllat permitted the expansion of trade and production. AltllOugh arising out of bloodshed and conquest, an imperial state was most admired for tlle peace it gave its component parts. Areas that had suffered endemic warfare could now turn their

Imperial state formation along the Chinese-Nomad frontier

attention to trade, investment, and production of goods. The market for such products was also greatly increased. The level of production and distribution of goods reached unprecedented scales, including large-scale foreign trade with distant peoples for luxury goods. These economic links bound the empire's component parts together far more strongly than any army. Roads and ports designed to ensure the rapid transport of troops and weapons to put down possible rebellions eventually became the lifelines of civilian commerce tllat superseded military force as the key to maintaining unity. Third, empires had sophisticated systems of communication that allowed them to administer all subject areas from the center directly. If trade and taxation were the material lifeblood tllat flowed through the empire's veins, then the development of a communications system could be likened to its central nervous system. AlIimportant policy decisions in empires were made at the center, so tile expedited flow of information was critical to imperial survival. Roads that transported armies also sped information from the very edges of empire to the center for analysis and disposition. Although an army on the march was a more highly visible sign of imperial power, the quieter flow of information to the center and instructions from tile center was perhaps more impressive. All empires had some sort of official postal system (often surprisingly swift in moving important information) that variously included horse relay stations, runners, or fast boats. In admiration of the Persian empire's messenger service, Herodotus (8.98.2) observed: Than this system of messengers there is nothing of mortal origin that is quicker. This is how the Persians arranged it: they say that for as many days as the whole journey consists in, that many horses and men arc stationed at intervals of a day's journey, onc horse and onc man assigned to each day. And him neither snow nor rain nor heat nor night holds back ti·oll1 for the accomplishment of the course that has been assigned to him, as quickly as he may.

But a communications system was much more than just a way to transmit information physically. It required a sophisticated record keeping system (most often writing, but also recording devices like Inkan Ithipu, see D'Altroy, p. 203), a permanent bureaucracy that managed the information flow, and a system for dealing with information in a timely manner. Underlying this was a common language of administration and education that all members of the imperial elite shared. This could be accomplished in a number of ways: with a common writing or symbolic system that transcended language differences (Chinese characters, Inkan Izhipu), using a common language of imperial administration that all members of the elite shared (Latin and Greek in the Roman world, Arabic in the Islamic world), or by employing a cultural lingua franca of administration adopted by new rulers even when it was not their own native tongue (L'ltin in the post-Roman West, or Persian in Central Asia and Mughal India). Less noted (but equally important) were empires' impositions of standardized measures (for weights, distance, volume, troop size, and money) and common numeral

31

32

Thomas J. Barfield systems for recording numbers and accounting. Such systems did not necessarily displace localized customary measures, but they did serve as a common denominator understood throughout the empire and meant that imperial officials did not have to keep track of myriad local variations.

Fourth, empires proclaimed a monopoly offorce within the territories they ruled and projected their military force outward. This may be a baseline requirement of all states, but because they were founded on the basis of conquest and had to administer very large territories, empires faced problems of scale. Preservation of internal order required the maintaining of a permanent imperial army that could control disturbances in any part of the empire. But to the extent that empires regularly relied on such armies to preserve internal order, they were short-lived polities. As a Chinese adage had it, "You can conquer an empire on horseback, but cannot rule it from there." Successful empires therefore maintained internal order by effective administration, including implementation of a common legal system that could be relied upon throughout its territory and a centralized system of government appointees responsible for carrying out imperial policy. Local officials were often quite constrained about what they could do without permission from the center. In successful empires, internal rebellions were relatively rare and considered a sign of administrative failure. Externally, empires sought to expand well beyond the limits of other types of states. In their expansive phases they generally stopped only for three reasons: (1) when they reached the frontiers of another empire of similar power to their own (Rome/Iran); (2) when they reached an ecological frontier (desert, steppe, mountains, jungle) that they could not effectively occupy (China and Mongolian steppe, Rome and the Sahara and Arabian deserts); or (3) where an advance that could be accomplished militarily was foregone as part of a strategic policy designed to create a defensible frontier (Rome on the Rhine), or where cost of administration was deemed to exceed the benefit of running it (China in Central Asia). Once established, however, imperial borders displayed remarkable persistence and continuity. In part this was because empires stationed the bulk of their troops on the frontiers, not at their centers. Even in decline, empires sought to defend the whole of their territory. Very few (notably the Byzantine) retreated gradually to the core from which they started when pressured on the ft·ontier. Instead empires tended to direct ever scarcer resources from the center to the frontier to hold the line even at the risk of collapsing the whole system. Fifth, empires had an ((imperial project» that i'mposed some type ofunity throughout the syste'l'n. One of the reasons that empires were so tolerant of diversity was that they expected that their own cultural system would create a common core of values that would override local variation. This was often reflected in such material attributes as common systems of measurement, architectural styles, cosmology, ritual, art, and fashion that marked an imperial presence, even at the margins. It was a vision of unity that extended beyond force and created what we often identifY as a civilization (see Woolf, this volume). It was a project that

Imperial state formation along the Chinese-Nomad frontier allowed wide participation and was implemented in a series of stages that moved from coercion through cooptation to cooperation and identification. Those empires that did not seek to develop long-term cooperation and identification were generally shorter lived than those that did. In such long-lived empires this legacy of belonging often survived its political collapse and remained alive as a historic model to be emulated in the future (see Moreland, MacCormack, this volume).

From empires to lat;ge states? It could be argued that these definitions could be equally applied to large states, not just empires. This should not be surprising because from an archaeological perspective it appears that empires were the templates for large states, and not the reverse. Historically, empires were the crucibles in which the possibility of large states was realized. Indeed, it is difficult to find examples of large states in areas that were not first united by an empire. Pre-imperial political structures were generally small and parochial: city-states, regional kingdoms, and tribes. States that did expand in this environment either grew to become empires themselves or collapsed into their component parts. It was the experience of empire that changed the political and social environment and created the capacity to rule large areas and populations in the states that followed. Imperial methods of government administration, military organization, and ideology were modified and used as models on a smaller scale. Thus large states were most common in areas where empires broke up and the imperial pieces became large states. It is only as we approach the modern period that large states rather than empires become common. This may have to do more with the expansion of western Europe, where this process was most pronounced, than with any changes in other parts of the world where empires as political structures ruled supreme (if not always independently) in South Asia, China, and the Near East until the twentieth century. It is also true that with the expansion of the world's population, states commonly found themselves ruling over large populations (40 million plus) that had been previously found only in the biggest empires of the premodern period. For most of history, there was an order of magnitude difference between the size of primary empires and any other type of state.

Shadow empires If classic empires were primary polities, that is, the product of an indigenous internal development, shadow empires were secondary phenomena. They came into existence as a response to imperial state formation that was initiated some place else. They could not exist except as part of an interaction with an imperial state because they lacked most of the essential characteristics of primary empires described above. However, in terms of power and historic influence, they mimicked empires in their actions and policies. They were "shadows" in that they took on the form of empires without all of their substance. They were in some

33

34

Thomas J. Barfield way parasitic on a larger systems, although under exceptional conditions they could transform themselves into self-sustaining primary empires. The cases in tlus volume illustrate at least four types of secondary empires: mirror empires, maritime trade empires, vulture empires, and empires of nostalgia. Mirror empires Mirror empires emerged in direct response to imperial state formation of tl1eir neighbors. They rose and fell in tandem witl1 their rivals because tl1ey were responses to tl1e challenges presented by a neighbor's imperial centralization. Nomadic imperial confederacies are prime examples of mirror empire. The Xiongnu empire, and tl1e steppe empires tl1at followed it in Mongolia, rose and fell in tandem witl1 native Chinese dynasties because they were parasitically attached to them. Such empires lacked tl1e internal sophistication of a primary empire: no distinct economic classes, no standing armies, little literacy witl1 a milumal bureaucracy, few craft specialists and (most important) a subject population that could not be easily exploited. But externally, the steppe ruler's monopoly on foreign relations and command of a very effective centralized military force enabled him to impose his will on tl1e steppe as a whole and to create an imperial structure tl1at could deal as an equal witl1 even very large empires. The creation and maintenance of a unified empire as a response to a neighbor's imperial expansion was found mostly among nomadic people, and then primarily on tl1e Eurasian steppe. The reason for this was that the nomads' political structure was determined more by their relations with the outside world than by any internal dynamic. Drawing on cases from southwestern Asia where polities were much smaller, Irons (1979: 362) observed that, "among pastoral nomadic societies hierarchical political institutions are generated only by external relations witl1 state societies and never develop purely as a result of internal dynamics in such societies." As a corollary one might add that the degree of hierarchy mirrored the level of organization of their opponents. In gross terms there was historically an arc of growing political centralization that ran ft'om East Africa to the steppes of Mongolia with four increasingly complex types of tribal organization (Barfield 1993: 17): (1) age sets and acephalous segmentary lineages in subSaharan Africa where tribal societies encountered few state societies until the colonial era; (2) lineages with permanent leaders but no regular supra-tribal organization in North Africa and Arabia where tribal societies f.1ced regional states with which they had symbiotic relations; (3) supra-tribal confederations with powerful leaders who were part of a regional political network within large empires distributed throughout the Iranian and Anatolian plateaus linking tribes to states as conquerors or subjects; and (4) centralized tribal states ruling over vast distances on the steppes of central Eurasia, north of China and Iran, supported by predatory relationships with neigh boring sedentary empires. In Mongolia, f.1cing a united China, the steppe nomads regularly reached an empire level of organization. In part this was because their horse cavalry gave them the military means to challenge China. But it was also because their social

Imperial state formation along the Chinese-Nomad frontier structure (unlike tllOse of nomads in the Near East or East Africa) was ruerarcrucal and culturally predisposed to accept tl1e claims of imperial lineages tl1at proclaimed an exclusive right to rule (Barfield 1991). As we have seen, steppe empires like the Xiongnu were not interested in conquest but extortion, and once tl1ey had established tl1eir lucrative relationship witl1 China (peace for goods) wealth poured into tl1e steppe and maintained tl1e imperial structure. For tl1e most part, mirror empires displayed little potential to transform tl1emselves into primary empires. The only case of this on tl1e steppe was the Mongol empire established by Chinggis Khan in the thirteentl1 century. He too started out intending only to extort his neighbors, but the technical skill and power of his army ended up destroying them and creating a true empire (t11e world's largest) t11at ruled directly from a center and relied on direct conquest and taxation of sedentary areas. For a time t11eir capital, Karakorom, became tl1e center of power for all of Eurasia, but within a few generations it broke down into its component parts. Maritime trade empires Maritime trade empires held only the minimum territory (usually at the margins) needed to extract economic benefits through trade from anotl1er polity run by someone else who was responsible for production and political organization. This form of empire, at least in the premodern period, was problematic because it attempted to extract wealth without the occupation or extortion of a large territory. To do this they attempted a monopoly, or at least direct control, over the means of exchange and transport rather than the means of production. For this reason all trade empires were maritime because it was possible, with adequate naval forces and fortified bases, to maintain control of shipping over large areas of ocean, particularly at the choke points of sea routes or at key embarkation points. By contrast, it was not possible to institute such controls over a land route without also dominating the territory surrounding it. Such control allowed imperial centers to amass the degree of wealth needed to finance the military power and act as empires in their dealings with other states. If the ultimate goal of empire was the centralization and control of resources to be used by the center to enrich itself and control the periphery, then such maritime empires could accomplish this task without the need to rule over large territories and populations. However, like the equally parasitic steppe empires, such polities were in many respects hollow (for a different perspective, see Subrahmanyam, this volume). They were vulnerable to collapse if their source of wealth, over which they had no direct control, was cut off. They were also vulnerable to attack ft'om surrounding states and rival naval powers. The solution to tl1ese problems involved either devolution (to become the center of a trading network that was not autonomous or in private hands) or evolution (expand into a true empire by conquering neighboring hinterlands and the populations that produce tl1e goods traded). In the ancient world, maritime trading empires arose most commonly as the

35

36

Thomas]. Barfield consequence of the growing power of a city-state in a world of city-states. The trading network of Phoenician city-states, or perhaps Bronze Age Crete, may have been the first examples. The ancient Athenian empire of the fifth century BC E, with its collection of colonies and clients from whom it benefited but which it did not rule directly was a classic example of an empire based on trade ratl1er t1un conquest. It took advantage of Atl1ens' superior navy and network of alliances t11at had emerged during the Persian Wars and used its position to become a center of wealth and power. The weakness of such a structure was also clear when, during the Peloponnesian War, the Athenians found themselves vulnerable to land attack and trade disruption by the Spartans and t11eir allies. The rise of centralization in the Mediterranean under Alexander and his successors, and then Rome, put an end to such polities. In t11e medieval Mediterranean world, city-states such as Venice and Genoa also created trade networks but they were unable to achieve a degree of autonomy that would lead anyone to call them empires. Later examples were on t11e cusp of the modern age and involved the expansion oftl1e Atlantic European powers into the Indian and Pacific Oceans. These included t11e Portuguese in the Indies, as well as the later ventures by t11e Dutch and British. As Subrahmanyam's analysis (t11is volume) of the Portuguese empire in Asia during the sixteenth and seventeenth centuries makes clear, they dominated a lucrative trading network with a surprisingly small numbers of men and boats. This domination was not due entirely to technical superiority (gunpowder weapons were quickly acquired by states across the region), but to the fact that the Indian Ocean had no local regional hegemony (but see Morrison, this volume). When the region's neighboring land-based states decided that they were losing something important, they turned on the Portuguese. The Ottoman Turks built a navy and increased their control of the Red Sea region, the Saf.wids in Iran retook the Persian Gulfport ofHormuz, and the Mughal empire in India united much of the subcontinent and excluded the Portuguese from many of their strongholds. In addition, the profits of the sea trade attracted the Dutch and English. The latter two straddled the boundary between state and private company, with the state taking predominance when they decided to occupy and directly administer the territory they exploited. Over time, these trade empires therefore evolved into primary empires in which colonial dependencies were ruled directly from London and Amsterdam. Vulture e'mpires Vulture empires were created by leaders of frontier provinces or client states or by allies that sat on the geographic and cultural periphery of an empire. They were formed after the internal collapse of their imperial neighbors, when peripheral leaders were able to seize the imperial center and form a new empire. Frontier regions on the edge of empires were always considered to be places apart culturally and administratively even though they had strong economic and political ties to the empire. At some periods, they were formally incorporated

Imperial state formation along the Chinese-Nomad frontier into an empire and occupied by imperial garrisons; at other times, they were ruled indirectly as client states or allies. In either case, empires typically compromised their usual policies of direct incorporation and cultural assimilation in favor of t11e cooptation of local e1ites who maintained much autonomy and preserved t11eir indigenous social structure. They were f.'lmiliar witl1 the dominant imperial culture but still strongly linked to t11eir coetlmics in t11e hinterland who maintained older indigenous ways of life. Long experience witl1 the dominant empire transformed local culture and political organization. During times of centralization, participation in t11e imperial system allowed cooperating local leaders to use their links witl1 t11e empire to overcome rivals with economic and military aid from the center. However, when the center weakened or witl1drew, these areas became independent. TillS led to a struggle between t110se local leaders who wished to return to a more autonomous and less hierarchical political structure, and those who wished to employ the tools of the witl1drawn imperium to create their own centralized state. This contest often took generations to resolve. If it resulted in the emergel1Ce of a new centralized state, it was one that combined elements of local indigenous culture with an administrative system modeled on the old empire (and sometimes employing its leftover personnel). If the center was weak enough, particularly if civil wars had created anarchy, leaders from such marginal territories could seize control of the center t11emselves and found new imperial dynasties. Although the new dynasty might be named after its new elite's frontier homeland, its success depended on maintaining domination over the existing administration of an older empire that it ran as a dominant minority. Such dynasties were, if you will, the tails that wagged the imperial dog occasionally, but even when dressed in imperial regalia they were always recognizable as tails and never dogs. The best example of clients at the margins in this volume would be Nubia, which Egypt treated as part of its political and economic sphere. Nubia absorbed much from Egypt but remained distinct culturally fi·om the lower part of the Nile Valley. As MOl·kot (this volume) points out, Nubia was an Egyptian frontier on the upper Nile, where Egypt initially established bases for direct control but also depended on the incorporation of a local elite as their agents. Over the course of time, interaction with Egypt had an enormous influence on the Nubian population. They were drawn into Egyptian politics and culture without becoming fully Egyptian. After Egypt withdrew, a process of state formation based on Egyptian models took place, its details largely undocumented, that resulted in the establishment ofNubian dynasties that later ruled Egypt. This was not really a Nubian empire so much as Nubians taking control of a declining Egyptian empire as a new dominant group. The population and economic locus always remained in Egypt proper, even though Nubia became more of a center politically than it was before. Similar examples can be drawn from other parts of the world, particularly frontier regions like Manchuria that play a similar role in Chinese history (Barfield

37

38

Thomas J. Barfield

1989). Like Nubia, Manchuria was a thinly populated border region that came under imperial control, directly and indirectly, when China was unified under native dynasties like the Han, Tang, and Ming. What parts escaped China's control fell under the imperial sway of tile Mongolian steppe empires. During these periods the Manchmian tribes were very weak and disorganized. However, upon tile collapse of imperial autllority within China (and tile consequent collapse of unity of tile Mongolian steppe), Manchmian tribes became autonomous. Rival tribes struggled for dominance in tllis vacuum and rulers became dominant by creating a new system of dual administrations Witll one branch (staffed by tribesmen) in charge of tribal affairs and war while tile otller branch (staffed by Chinese bureaucrats) handled civil affairs. Using tllis as a base, tlley expanded into areas of nortllern China weakened by civil war. When successful, tlley established dynasties in which tile Manchurians were a dominant elite in an empire still ruled by Cllinese officials under Chinese law. It was a new empire but built on an existing foundation. Table 1.1 shows tllat this pattern occurred regularly after tile collapse of central order in China. The final, and most successful, dynasty in tllis line was tile Qing (1644--1912), tllat eventually unified all of China and ruled it for almost tluee centmies. In tile process it underwent significant Sinification so that after a few generations it was transformed into a powerful primary empire in which Manchuria itself once again became a backwater. Empires of nostalgia The last group of shadow empires, empires of nostalgia, were based on remembrance of organizations past. They claimed an imperial tradition and the outward trappings of an extinct empire, but could not themselves meet basic requirements of true empire, such as centralized rule, direct control of territory, a significant imperial center, or enough territory to make tile imperial grade (i.e., a province of former empire). Long-lived empires left lasting marks on the regions they ruled (but see Sinopoli, this volume). When empires collapsed (particularly if that collapse resulted in political anarchy, population decline, and economic depression) the old imperial structure was often imbued with the aura of a former "golden age" now lost. More importantly the memory of the lost empire retained such an ideological hold over a region's population that it could be used as a powerful tool by rulers attempting to build new states and empires. It also provided a template for building a large-scale administration. As Yates (tllis volume) has shown, this tendency was strongest in China where the cosmological myth of a necessary emperor who linked heaven and earth was so influential that it served as the ideological foundation used by all succeeding states. The memory of empires past provided the foundations for empires of the future, and a united empire itselfwas seen as the highest political goal. A divided China was considered so incomplete that any group that could reunite it, including foreign invaders like the Mongols, was perceived as legitimate rulers when tlley accomplished this task. These efforts were so successful that, unlike the

Imperial state formation along the Chinese-Nomad frontier

West, primary empires came to reunite Cllina after each period of state collapse. In many time periods, the reestablishment of a primary empire was, however, more often a distant hope tllan an achievable goal. The idea of empire provided the ideological basis and adnlinistrative model for the creation of a new empire, but in most cases was little more tllan an influential fantasy because tile material base was lacking and tile ability of leaders to centralize power was opposed by powerful local elites. The Carolingian empire fell into tllis category, as did tile small Cllinese states tllat arose in periods of disunion tllere (third tluough Sixtll centuries, tentll tlu·ough twelfth centuries), or tile many dynasties tllat claimed tile mantle of empire in Etlliopia. The Carolingian empire (Moreland, tllis volume) is a particularly good example because it meets almost none of tile criteria listed above for primary empires. It did not have a permanent standing army, a true central administration, or effective transportation and communication systems. It parceled out territory on a feudal basis and local notables felt under little obligation to carry out its orders. Its capitals were small and temporary; its soldiers and officials moved to where the food was because tlley could not transport it. Imperial architecture was impressive only when compared to the peasant huts tllat surrounded it. Still, it was recognized as an empire at the time, and it continues to hold a hallowed place in European history. Carolingian influence was so great because, as the first major attempt to bring back the idea of empire modeled on Rome in the West, it struck a powerful cultural chord and served as a potent ideological weapon in the (eventually unsuccessful) drive to centralize the petty states that occupied the former provinces of the Western Roman empire. In empires of nostalgia, rulers tied their own legitimacy to something that no longer existed but that had the power to command loyalty, a source of power tllat has been much underrated by scholars who have focused only on material conditions. Petty struggles for power and supremacy were tied to loftier goals. Cooperation was easier to achieve and the recognition of the fiction of a new empire and emperor provided the basis on which to build a new structure. For this reason, empires of nostalgia need to be seen as secondary phenomena that drew their power largely from the symbolic world of cultural memory. The only thing that could kill it was the inculcation of a new cultural order. Thus, while the idea of Rome remained strong in the West, the tradition was lost in Nortll Mrica after the Islamic conquest. From that point on, the empires of nostalgia there were drawn from the Islamic tradition. Similarly the long-lived Byzantine tradition did not survive the Ottoman conquest and its cultural transformation into the Turkish world, although its legacy did survive in far-off Russia, which saw itself as the inheritor of the Byzantine legacy of Orthodox Christianity.

CONCLUSIONS Understanding the dynamics of empires demands coming to grips Witll what tlley took for granted, issues tllat often require us to examine our own assumptions

39

40

Thomas]. Barfield about what made them work. When we compare primary and secondary empires we can see that they relied on their ability both to centralize power and to maintain control at a distance. Primary empires did this by creating complex structures that were self-sustaining and self-reproducing. Secondary empires achieved the same ends but without the complexity because they did not generate their own revenue directly. Two of our secondary types, tlle mirror empires of tlle steppe and tlle maritime trade empires of the seas, developed specific military technologies tllat magnified their power and disguised tlleir weakness in numbers. For the steppe empires, this was horse cavalry, for maritime empires, a naval force. BOtll permitted military strikes at a distance and allowed for a small but concentrated force to attack and dominate an enemy at a chosen point. The steppe empires used tllis power to establish empires financed by extortion and subsidies drawn from China. The maritime empires used tlleir power not so much to extort revenue directly as to dominate tlle international terms of trade to tlleir benefit. In both cases, military power did have the potential to transmute a secondary empire into a primary one ifit began to conquer distant lands directly. This rarely happened because these empires were run by people whose numbers were few and scattered. Their power also lay in their ability to retreat into tlle steppe or over tlle seas WitllOut putting tlleir key military assets at risk when confronted by a powerful counter-force. Extortion and trade were their strategic tools of choice. Both of tllese types of empires depended on the existence of more complex states to survive and existed simultaneously with powerful primary empires. Vulture empires and empires of nostalgia, by contrast, existed only in the wake of the collapse of empire. If there was enough of the structure intact, then vulture empires were able to keep them going (usually with a simpler organizational structure) by creating new empires or dynasties on the remains of the old. If an imperial system had collapsed more completciy, or its structural integrity had completely disappeared, then attempts to recreate it had a provisional quality to them, more symbolic than real. It was the myth of empire that ran at the heart of empires of nostalgia, but it was a powerful myth indeed, capable of inspiring people to recreate what was lost even when they lacked the means to do it. For archaeologists in particular, secondary empires like the Xiongnu provide an example of powerful long-lived imperial polities that break out of the normal mold. And, as this examination has shown, different lypes of polities, which I have classified as primary and secondary, can both produce imperial structures, but of quite different types. Perhaps because empires were so large they have been treated for too long as isolated phenomena and not as spheres of interaction (see also Smith, this volume). Once the process of empire was established it had an enormous impact internally (a process that has been well studied), but also externally (a process that has been neglected). Empires influenced one another and surrounding peoples even at long distances. Whether creating a frontier with tribal peoples (Rome and Germany, China and Mongolia) or confronting rival empires (Rome and Iran, Ottoman and Spanish), the dynamics of

Imperial state formation along the Chinese-Nomad frontier empire shaped tlleir development. Much of this development was not documented in written sources but did leave an archaeological record. It is to this archaeological record tllat we should turn to better understand how empires worked and the impact they had at home and abroad.

41

Designs and dynamics in the Portuguese Estado da India

2

Written on water: designs and dynamics in the Portugese Estado da India /

Sanjay Subrahmanyam

Many physicists are working very hard to put together a grand picture that unifies everything into one super-duper model. It's a delightful game, but at the present time none of the speculators agree with any of the other speculators as to what the grand picture is. (Feynman 1985: 150)

PROBLEMS OF DEFINITION

42

Comparing empires obviously requires some reflection on the nature of the domain itself, and more particularly, in this context, on the vexed question of the criteria for inclusion and exclusion (Duverger 1980). Empires are states, but not all states are empires; we may begin with this simple dictum. But going beyond such rather evident wisdom, it appears clear that there are at least two quite different ways of thinking about empires. One approach, relatively structuralist in orientation, would require a very large number of criteria to be met by any state that has pretensions to being an empire. The "symptoms" of empire would include elaborate hierarchical systems of administration, extensive military power (and the fiscal mechanisms that go with it), the control over extensive landmasses, a large subject population, and substantial revenues. Not only might all of these be taken into consideration, but some of our colleagues would even wish to define quantitative thresholds to be met in each of these cases. I must confess, at the outset, to considerable skepticism with regard to this quite peremptory manner of proceeding. If one is to use the concept of empire as much for antique political formations, when the world population was a mere fraction of what it was in, for example, around 1700, and the technologies of state-building rather different from those of the medieval and modern periods, it is obvious that the same demographic, or fiscal, or even military thresholds could simply not work for, say, the Achaemenids and the sixteenth-century Ottomans. The problem is roughly the same as defining an idea such as the "city": many quite minor settlements in South Asia today would exceed in population what were considered cities in antiquity, but tlley do not function as cities. It would hence appear to me far more sensible to propose a rather more minimal, contingent, and conjunctural definition of empire, one that controls,

.

:.

inter alia, for tile passage of time; thus, one ought not anachronistically to demand from empires of tile centuries before tile Common Era tile same tilat one does from early modern empires, for instance. But, equally, I would propose that even within a synchronic sample, such as one from the early modern period, tile notion of "empire" be nuanced somewhat. In the period between 1400 CE and 1750, some clear candidates exist for the status of empire. Amongst these are the Ottomans, tile Mughals in South Asia, the Mings and Qings in China, tile Spaniards (or tile Spanish Habsburgs) after about 1520, tile British and tile French in tile eighteenth century and later. But many more cases exist whose status was indeterminate even to contemporary observers. Were the Mexica in Tenochtitlan (or the Aztecs as they are sometimes called in the literature; see chapters by Brumfiel and Smitll, tllis volume) at the head of an empire, as tile size of tlleir domains might suggest? What of the Burmese monarchy of the Toungoo Dynasty in the years from 1530 to 1600? In brief, what is one to do with seemingly "imperial" states tllat do not fit into the mold and heritage defined in antiquity by four classic, and generally admitted, precedents: the empire of Alexander, tllat of the Achaemenids in Persia, the Roman Empire, and China? This is tile problem that one f.1ces with the Portuguese in the sixteenth century. For while it is clear that the Portuguese monarch in tile years before 1580 (the year when tile Portuguese and Spanish crowns were united) did not term himself "emperor," bOtll the Italians and the northern European rivals of the Portuguese saw the overseas possessions controlled from Lisbon more or less in those terms. But, ironically, it was only far later, in 1822, that Dom Pedro of tile House of Bragan~a officially accepted in his Brazilian exile the title of "constitutiona
emperor." Thus, f.1r from the "maximal" approach summarized above, I would propose a "minimal" approach, in which empires would be rather modestly described as follows: (1) as states with an extensive geographical spread, embracing more than one cultural domain and ecozone; (2) as states powered by an ideological motor that claimed extensive, at times even universal, forms of dominance, rather tllan the mere control of a compact domain; (3) as states where the idea of suzerainty was a crucial component of political articulation, and where the monarch was defined not merely as king, but as "king over kings," with an explicit notion of hierarchy in which various levels of sovereignty, both "from above" and "from below," were involved. Writing in 1985, in a classic essay on the administrative and political structure of the Portuguese state in sixteenth- and seventeenth-century Asia, the most innovative Portuguese historian of that domain, LUls Filipe Thomaz (1994 [1985]: 208), wrote presciently of his subject matter: "When we confi'ont it with the clll'rent notion of empire, the Portuguese Estado da India may appear to us to be somewhat original, and even disconcerting." He nevertheless proposed to include the Portuguese presence in Asia within the category of empire, while equally making a studious effort at contrasting the Portuguese with tile Spaniards in the Americas in the same period. If the model for the Spaniards was Rome (see MacCormack, this volume), it was argued by Thomaz, the precedent

43

44

Sanjay Subrahmanyam for the Portuguese may well have been the Phoenicians. In marked contrast, some years later, another study, also by a quite well-known authority on the Portuguese Estado da india ("State of India"), concluded with the following remarks, of some relevance for the reflections brought together in this volume: It has to be extremely doubtful that what the Portuguese constructed in Asia in the sixteenth and seventeenth centuries should really bear this title [of empire 1 at all ... Given tile su·ong image conjured up by the use of tile word "empire," and tile difficulty of reconciling tile actual content of the Estado da i"dia for most ofthe sixteentll century to tile real meaning of the term, it might be more appropriate to drop it altogetller. (Disney 1995: 34-5)

The same author then went on to suggest 1:\'10 alternatives: first, the quite vague idea of "network," to which one could immediately raise myriad objections on account of its imprecision, and absolute lack of reference to the idea of power; and second, the idea of a sort of proto-empire, a presence that was arguably in the process of becoming an empire, which to all intents and purposes merely pushes the problem to one remove. Some other writers, more imbued perhaps with the spirit of political correctness, have gone even further since then. The mere mention of words like "empire" in a Portuguese context, it is argued by historians like M. N. Pearson, means that suspicion should fall on the political motives of the authors themselves, whose use of such terminology "come[ s] close to invalidating their work" (Pearson 1996: 7). At the core of this, admittedly at times overwrought, discussion is a serious matter, an issue that needs to be aired: namely, that various sorts of imperial polities have existed in the past, which do not all conform to a single profile, with a contiguous landmass, centralized fiscal and cadastral organizations, and a powerful and continuous imperial military presence in peripheries that are rigorously controlled from a well-defined center. Did the Mongol "empire of the steppes" really constitute an empire in the same sense as, say, the Mughal empire in South Asia? Is the fiscal and bureaucratic structure of Vi jayan agar a at all comparable in its extent and depth to that of the Ottomans? This matter is made even more complex in the case of the Portuguese Estado da india, a structure that was - to employ and adapt a popular Indian expression of the epoch - "written on water." It should be made clear then that the idea of a Portuguese empire in Asia in the early modern period will necessarily be a more contested one than, say, the proposition that there was indeed a Roman empire in antiquity, which no historian at present seems noticeably interested in denying.

DIMENSIONS AND MOTIVATIONS This chapter presents a concise and somewhat schematic narrative of the Portuguese presence in Asia and the Indian Ocean between about 1505 - the year when Dom Francisco de Almeida was sent out as the first viceroy of Portuguese India - and 1665, when hostilities between the Portuguese and their

Designs and dynamics in the Portuguese Estado da India main European rivals, the Dutch, had nearly come to an end in Asia. My narrative will be broadly chronological, and its main analytical purposes are three in number. First and foremost, I wish to argue that unlike an earlier historiography on the Portuguese, which had focused on an Age of Consolidation in the first half of the sixteenth century, followed by an Age of Decadence (reaching its nadir under the later Habsburgs in the 1630s), it is in fact necessary to look to the Portuguese presence in Asia as a constantly evolving one, responding to a multiplicity of stimuli in both Asia and Europe. This itself thus requires a direct disavowal of some of the more structural methods used to define tlle essence oftllis or tllat empire, which one so commonly finds in the literature. Second, it is an implicit argument tllat it is necessary to move away from an exclusive focus on tlle western Indian Ocean (Kerala, Goa, and Gujarat) to a view that allows a far greater importance to the Portuguese presence east of Cape Comorin, namely around tlle Bay of Bengal, in Southeast Asia, and the Far East. Thus, my second objective is to propose a geographical opening out in our general conception of what constituted Portuguese expansion in Asia, and thus an exploration of the interface between "formal" and "informal" empire (as well as tlle shades of gray in between; Coates 1998). Third, there is a question of the social groups to be addressed in defining a history of Portuguese expansion in Asia. Rather than focus exclusively on the official hierarchy of viceroys, governors, and aristocrats (fidalgos), we need to look at other social categories, ranging from trader-settlers (or casados), to renegades, to mixed-blood (mestiro), and to other groups such as Christian converts who participated willy-nilly in building the edifice of the Portuguese presence in Asia. In sum, if many empires that we recognize as such were agrarian ones, we should in my view also permit the possibility of a commercial and frequently semi-formal empire within our range of possibilities. The Portuguese presence in Asia extended by the middle years of the sixteenth century from East Africa to Japan, via the Persian Gulf, India, Sri Lanka, the Malay Peninsula, and island Southeast Asia (Fig. 2.1). There is thus no doubt about its sheer physical spread, with problems being posed, on the contrary, by the density of the presence in each of these areas. What then were the human dimensions of this structure? Around 1540, one frequently cited contemporary estimate puts the number of Portuguese between Sofa la and China at 6000 to 7000; in the 1570s, the chronicler Diogo do Couto puts the total number of Portuguese in Asia at about 16,000. But with the process of miscegenation, and acculturation, the numbers of people within the official settlements of Portuguese Asia were obviously rather higher. These included the casados or casados moradores, which is to say married settler-colonizers; soldados or soldiers; and religiosos or ecclesiastics, as the main categories. The account of Ant6nio Bocarro from the mid-1630s provides us a breakdown by settlement of the casados, showing their presence to have been a vigorous one still at that time, when the Portuguese presence had entered somewhat into decline. Interestingly, he insists on listing not only the casados brancos (white casados) as was usually the case in official compendia, but also the native Christian settlers of the towns,

45

DCS{lJltS and dYllamics ill the Portugucse Estado da India Table 2.1. Casado settlements i1l Port1lguese Asia) 1635 Settlcmcllt Macall Goa Daman Basscin Collll11bo Cochin Mclaka Challl Nagapattinam Jafflla Sao Tome Thana Galle Others Total

White

Rtflell

850 800 400 400 350 300 250 200 140 140 120 80 70 800

850 2200

4900

600 2000 200 50 360 270 200 100

130 525 7485

tu

co i= z

~CI)

Cl)w

I'I11s were cacao beans I;')r small purchases, and cotton textiles ()r most costly items. The imperial tribute system greatly enriched Tenochtitian, with large quantities of both raw materials and manu(:lCtured goods arriving several times a year, and merchants brought even greater quantities of goods !i'om distant areas (Berdan 1987b). Although onc sometimes reads that the Aztec empire had begun to decline bcl()re the arrival of Cones, there is no evidence to support such a notion.

AZTEC IMPERIAL STRATEGIES Scholars have been slow to identity the strategies employed by the Aztec rulers in the expansion, operation, and organization of their empire. Contemporary descriptions of the specific motives or plans of ancient imperial rulers arc few and

139

140

Miclme! E. Smith far between, even for heavily documented empires such as the Roman or Chinese. For the Aztecs, a bias against hegemonic empires has hindered the study of strategies; this is exemplified by statements that the Aztec polity was "not really an empire)) (sce discussion in Berdan et al. 1996: 6). Onc popular account of the Aztec empire (Conrad and Demarest 1984) has even asserted that the Mexica had no plans or strategies at all, and merely sent out armies of religious fanatics to conquer haphazardly in search of captives fClr rituals of human sacrifice. The authors of a recent book, Aztec Imperial Strrtt(lJies (Berdan et al. 1996), employed a "bottom-up)) approach to reconstructing the territory and economics of the empire, resulting in a significant modification of Robert Barlow's (1949) classic "top-down)) approach. By reconstructing the empire from the ground up, the Aztec Imjlerial Strategies project was able to: (1) identif)' a variety of strategies or principles that guided imperial expansion and administration; (2) produce a new and more accurate map of the empire; and (3) identif)' a new principle of territorial organization that characterized lllany of the provincial towns not listed in the Codex Mendozrt (sec discussion of the fl'ontier strategy below). The resulting principles and strategies can be organized under fCllIr labels: the economic, political, frontier, and elite strategies.

The ecortornic strate)]), The basic motives of Aztec imperial expansion were economic. Starting with the M PC Aztec city-states, the goal of conquest was to secure tribute payments fl'om subordinate politics. As the empire expanded, the volul11e and geographical range of tribute payments expanded greatly and the Mexica put considerable effllrt into establishing and operating a system of regular imperial tribute payl11el1\:S. But the collection of imperial tribute was only onc of three types of economic strategy employed by the Aztec rulers. A second economic strategy (llCused on the pre-existing commercial networks of the world system, and a third strategy involved stimulatioll and manipulation of the Basin of Mexico market system in ways that both benefited the economy and kept rival city-state I'lIlers in check.

Imperial tribllte In Doyle'5 terms, the Aztec tributc system was the major institution of the transnational system. Although we have considerable inflJl'll1ation about what kinds of goods were paid in what quantities fl'om specific provinces (Berdan and Anawalt 1992), the organization of tribute collection in the outer provinces is poorly understood. Because provincial city-states each had their own internal tribute systems in effect long bctllre they were conquered, it would have been easy (cJr city-state rulers to pass on the imperial tribute quota to their subjects by simply raising taxes. A quantitative reconstruction of' tribute in relation to demography in the Cuauhnahuac and Huaxtepec provinces of Morelos shows

141

The Aztec empire alld the lvIesoaJlleriCl1l1 lI'orld system that imperial tribute was not a particularly heavy burden when spread among the commoner households of the provinces (M. E. Smith 1994). Aztec imperial tribute included fClOdstulIs, raw materials, utilitarian goods, and luxuries. Some provinces were assessed fCl!' tribute in local specialties, while in other cases 1110re generic goods were demanded fl'ol11 a province and the Mexica obtained local specialties through trade. Prances Berdan (Berdan et al. 1996: ch. 5) has remarked upon the very low quantities of tribute in war materiel, in spite of the fact that the Mexica sometimes provided politics on imperial fl'ontiers with arms. The key raw material fCl!' Aztec swords and spears, obsidian, was not part of the imperial tribute system, probably because it was more efficiently obtained through preexisting commercial networks. I n many provinces, imperial tribute acted as an economic stimulus beyond simply requiring increased production fllr the tribute goods. Two of the main tribute items warrior costumes and textiles - were manuf~lctured goods whose raw materials were not widely available. Both required goods ft'om the tropical lowlands (exotic bird fCathers (CH' the elaborate costumes, and cotton fllr textiles) and many city-states had to engage in commerce to obtain them. Although part of the motivation fllr requiring these goods was probably ideological (subordination to the empire was emphasized by the warrior costumes), I believe that this was part of a deliberate strategy fllr stimulating commerce in the outer provi nces.

5.3 Azh'C profcssi01lal pociltcca mcrclmllls: left: lIIac/mll/s 01/ tllc road; right: !IUTe/mll/s sdlill.!! IlIxlIl'Y.!!oodJ III

a

Wodd .1),stClII C01Il1I1Cl'ciailletwor/ts The commercial net works of the 1,1'I'I11 than in their early stages. We still have only

223

, Farm or estate IfII Storage: 100-299 structures

dtlI Storage: 300-500 structures

t........................ ... I

./

N

(I) Storage: 1000 + structures Q Storage mentioned historically

r imperial religion; the warrior ideal was not often celebrated in monthly household rituals. Instead, these rituals focused on commoners' own concerns: agricultmal fertility, successful childbirth, and good health. Provincial clites may have embraced the cult of warfare and sacrifice. As Michacl Smith (1986: 75) observes, the Aztec rulers invited provincial clites to witness rituals in Tenochtitlan that involved massive human sacrifice. At these festivals, provincial clites were given sumptuous gifts of clothing, jewels, and military insignia (Oman 1967: I1, 174-5,279,293,308-10, 325, 345,415-16, 437, and 483). It is possible, as M. E. Smith (1986) suggests, that the rituals and gifts won over the regional elites and created a cultmally cohesive ruling class. But if regional elites supported the cult ofwarf.lre, its influence is not apparent in the rituals that local rulers sponsored in provincial capitals. According to the ethnohistoric record, rituals in provincial capitals joined many different sectors of the local population in activities to ensure rain, fertility, and bountiful

309

310

Elizabeth M. Brumfiel

harvests. Archaeological investigations of peasant houses in the provinces are consistent with the ethnohistoric record; they yield few indications of militarist ideology (BrumfielI996; M. E. Smith in press). This variation in ideology across divisions of age, gender, class, and locality may be typical of ancient empires. Given the mosaic of economic, political, and cultural circumstances in ancient empires (D'Altroy 1994; Morrison 1997a; Sinopoli 1994a; Smith and Hodge 1994; G. Stein 1994) and the difficulties of communication in the ancient world, it is probably unrealistic to expect that whole empires could be integrated by a single dominant ideology. More likely, dominant ideologies were specifically tailored and selectively deployed in ancient empires to win the support of strategically important groups. In Tenochtitlan, tllis was young men of fighting age, tlle core of the Aztec army. ~,

...

::-" , .... ,~

J l,..

-....

,...I

)

12

Inventing empire in ancient Rome GregWoolf

INTRODUCTION Ancient empires may be characterized as geographically extensive political entities. Most were ruled by elites who were often centrally located and internally divided, and whose power was severely limited by low levels of surplus production and preindustrial communications and technology. Imperial aristocracies thus depended on devolving a great part of their running costs on local elites of various sorts: creoles, descendants of traditional rulers, new cadres recruited from subject societies, and the like. Ancient empires were therefore of necessity tolerant of regional diversity and their rulers set themselves modest goals, often little more than maintaining their security against internal and external threats and extracting sufficient profit to reward those on whom imperial power depended. Despite, or perhaps because of, the resultant lack of homogeneity, many ancient empires were very long lasting. The early Roman empire of the first three centuries of the Common Era broadly conforms to this pattern (Fig. 12.1; for general accounts see Garnsey and Sailer 1987; Jacques and Scheid 1990; Goodman 1997). At the center an imperial court coexisted with a senior (senatorial) and a junior (equestrian) aristocracy. Tensions existed between the emperors and the senate, within both the senior and junior aristocracies and between aristocracies and the court. The emperors' power was based on their control of the empire's finances, of its administration, and of the standing army. But they also cultivated a reputation for generosity and virtue with all sectors of society: especially with the army and with the privileged populations of the capital and of Italy (Veyne 1976; Millar 1977). In the course of their conquest of the Mediterranean basin and its continental hinterlands, the Romans had absorbed kingdoms, cities and temple states, federal leagues, and sedentary and semi-nomadic tribal societies ofvarious types. The early empire was divided into provinces, but each governor presided over a large number of semi-autonomous communities, most notionally organized as republics, but in f.1Ct differing from one another widely except in tlleir duties to the Roman state, principally the maintenance of civil, social, and religious order. The communities of the empire had their own cults, constitutions, generally their own laws, and often their own monetary systems (Lepelley 1998).

311

312

blPC1ttill;[J empire ill ancient Rome

Greg Woo/f

Atlantic Ocean

Mediterranean Sea

\

I

I

km

12.1 '[hc ROIl/ail elllpirc at thc dcath of' A "ill/stllS, 14 CH. Solid lilies indicate illlpcrial bOlllldarics; dashcd fillcs illdiCll/c jJl'ol'incial bOllndarics; pray shaded arms indicatc principal efic/lt stall'S.

Again, despite this diversity, the empire was very long lasting. Most ofit was eonquered in the last two centuries BeE, with expansion being most rapid in the period 70 BC E to 10 C E. That period was also marked by a series of civil wars that ushered in the monarchical system described above. Despite a brief crisis in the third century C E, the emperors ruled the entire Mediterranean basin tCll' five hundred years, and parts of it {c)r much longer. Rome was more than simply a typical early empire: in some senses it was an archetypal one. The repeated use of Rome as a model of empire, fi'om the Christian and Islamic middle ages, through the early modern period into more recent times, is well known. Moreland and MacCormack treat some of these issues in their chapters in this volume, and a fldl account of the afterlife of Rome would extend well into the nineteenth and twentieth centuries, to include the overtly Roman styles of Napoleonic, Victorian, and lod he should ingest, in what activities he should engage, and what orders he should promulgate: all throughout the course of the year in order to harmonize himselfwith the ehanging rhythms of the (cHlr seasons as manifCsted in the alternation of the Five Phases (Wu Xing) dynamic powers that were used to categorize and classify all things in the universe and with the ever-fluctuating flows of the cosmic powers, Yin and Yang. The emperor was Man pm' excellencc within the tripartite unity of Heaven, Earth, and Man which composed the cosmos. Irregularities in the ruler's behavior would immediately affect the harmony of the natural world, and it would manilCst aberrations, resonating in response, such as generating snowlill1s in summer. Thus the macrocosm and the mierocosm humans, the state, and the natural order were bound together into a single, complex, organic whole through the medium of the body of' the ruler. When the First Emperor finally achieved the conquest of his rivals, he assumed the elevated title of Hllt1llJJd i (August Celestial Deity, which we translate "Emperor"), putting himself' on a par with the Folll' Cosmic Deities of the Folll' Quarters, and adopting Water as his ruling Phase. This Phase was correlated with the color blaek, and so he changed the color of his vestments and all imperial symbols, such as flags, to black. Six was the correlated number, so he ordered that the gauge of all chariots in the empire should be 6 (eet, and the number of regional administrative commanderies be thirty-six.

367

368

Robill D. S. Yates Thus he sought to harmonize the change in the nature of his polity with what he perceived to be current dominant Phase in nature. 31

CONCLUSION From what I have described above, the Qin imperial design was cosmographic and cosmologic. Huge amounts of labor and wealth resources were expended on bringing this vision to reality. However, as has been indicated, the Qin empire itself lasted only a few years beyond the death of the First Emperor, principally because the cornerstone of the entire structure was the emperor himself. When he was replaced with a weak, greedy, suspicious, lazy, incompetent younger son - the Second Emperor it took only a few years to subvert the dynasty. Nevertheless, the Qin had succeeded in inscribing their cosmic vision on the bodies of its people, in altering the language that they used, in subverting the structures of their families and their sense of time and of social practice, and in laying out over the landscape a normative hierarchy of administrative centers, flows ofcoml11unication, and a division of land. These practices then became the heart of the l11yth of the imperial unity of the Chinese people, a myth that has survived to this day and still infc>I'll1S the sense of self-identity of the Chinese people and the policies of their government.

PART V THE AFTERLIFE OF EMPIRES Susan E. Alcock

To poets, the end of empire has offered a metaphor through which to bewail the inevitable ending of all things - a compelling trope, expressed countless times, in ways that range fi'om the sublime to the ridiculous: "My name is Ozymandias, king of kings: Look on my works, ye Mighty, and despair!" Nothing beside remains. Round the decay Of that colossal wreck, boundless and bare The lone and level sands stretch 1;11' away. (Percy Bysshe Shelley, OZYlIIlIlIdillS) The jackals prowl, the serpents hiss In what was once Persepolis. Proud Bahylon is but a trace Upon the desert's dusty I:ICe ... And all the oligarchies, kings, And potentates th,1t ruled these things Are gone! But cheer up; dOIl't be sad; Think what a lovely time they had! (Arthlll' Guiterman, Elegy)

The death of empires seems to exert a lurid t:lscination, underlining, as it so pointedly does, the mutability and transience of human lite. Great men live, great deeds are done, great wealth is amassed, great power is wielded - but, in the end, it all comes to nothing. In this conception, empires are truly "written on water" (Subrahmanyam, this volume).

369

370

Susan E. Alcock Ironically, one of the most striking, and certainly one of the most persistent, messages of this volume is that old empires never entirely die, and some strongly resist fading away. In other words, empires possess a potent afterlife, albeit one that is decidedly stronger in some cases than in others. Such an imperial afterlife can take on many forms, but two particularly dominant manifestations will be highlighted here. First is the interplay between imperial systems themselves. Processes of emulation and rejection, copying and capping, of the mores and memories of one empire by another appear and reappear throughout the volume. Second is the long-lived and often insidious t:lshion in which understandings of certain imperial f()l'Il1ations have shaped subsequent scholarly paradigms. Rome, f()l' example, emerges in several chapters as an archetypal f<Jrce in the debate over "what is an empire?" (from different perspectives, see Barfield, Subrahmanyam, and vVoolf). However one examines this particular system, it seems indubitably imperial that word reminding us how the very vocabulary f()J' sllch things (leading to such phrases as jlax Aztcca or pax Sntrtl'ahalla!) is in part derived from the world of Rome. The case of Rome offers a good springboard, then, f()r exploring aspects of the imperial afterlife. Despite its ultimate "decline and fall," subsequent western empires clearly had Rome "in their bones," or at least on their minds. Its legacy became fertile ground t()r the construction of new imperial ideologies, right lip to the twentieth centLll'y and indeed even extending into the futLll'istic world of science fiction, as the Stal' Wars saga makes clear. Such reworkings are touched on in more than one chapter (I()r example, see Deagan, Wool!'), but are discussed at greatest length in Moreland's chapter on the Carolingian empire. Rome rebom (Roma l'etJascem) was an integral element in Carolingian conceptions of their rule, offering a necessary tactic of legitimation f(x a new and, in some ways, usurping elite. The Carolingians played admirably upon the notion of themselves as worthy successors, reinvigorating the "glory that was Rome" and highlighting links not least through the use of art, and architectural and monumental 1(>1'1115 of display. Certainly, this sell~representation largely convinced later scholars, overriding the quieter claims of other contemporary groups who likewise attempted to summon "the shades of the Roman ancestors" (Moreland, this volume, p. 415). MOl'eland makes clear, however, that this cannot be left as too simple a picture. The governing imperial strategy of the Carolingians was lllore the goal of I'etJIIPntio I'clfni F1'allC01'llTfl the restitution of the Frankish realm than the recreation of the Roman empire. Other potentially "easy" connections arc also subverted; ()r example, the Carolingians may have called on Roman I()rms of display, but their intended audiences were quite diff{.'rent. If Roman authorities had rendered their visions of imperial success broadly accessible and visible, to reach elite and non-elite alike (sce Alcock, Woolf), the Carolingians aimed rather at a narrower, elite consumption of imperial symbols, an objective more in keeping with their modes of governance. Moreland thus offers a classic case in which emulation and "succession" may, at (irst blush, seem quite stTaight/()rward

Part V Thc aftcrlife ofempires - but which subsequent contextual study reveals as the complicated product of much manipulation and reinterpretation, as the Carolingians summoned their Roman ancestors but put them to work in their own ways, to their own ends. Imperial rhetorics of continuity should not blind us to such stratagems, here or in other cases where the legacy of onc empire is employed to legitimate another. Rome may offer, in this volume, the easiest place from which to observe such reformulations and recastings. Yet such "empires of nostalgia" (in Barfield's phrase) can be traced elsewhere and arc of growing interest, as scholars of empire shift their focus "'om more purely material conditions and contingencies. Sinopoli '$ chapter, ICH' example, introduces the image of a Mauryan "golden age," annexed by such diverse later empires as the Gupta and the British; Mauryan symbols appeal' on modern Indian currency and the dynasty's most renowned ruler, Asoka (a figure respected in antique classical culture), is still a name with which to conjure. Sinopoli's main concern, however, is with a successor empire that did not scramble to bask in that rosy afterglow; indeed the Satavahana appear to have largely bypassed the Mauryans as a reservoir of imperial legitimacy. Two more general and very cogent points arc made here. First, the Satavahana example insists upon the possibility of rejection of~ 01' indifference to, what might be considered "obvious" models to invoke: imperial emulation need follow no lineal', inevitable trajectory. Second, the study underlines how onc dynasty's dominance in cultural memory can overshadow, even warp, the study of others, thanks to over-ready assumptions t-hat later politics would naturally "1~111 heir" to the structures and I!'ameworks of their predecessors. Concentration on and admiration Icu' the Mauryans, Sinopoli argues, has long impacted, and impeded, interpretations of the Satavahana. This returns us to the second principal manifestation of the imperial "alterli/C" the manner in which the perception of onc empire can shape perceptions of another. Returning to Rome, the "Ur-empire" of MacCormack's chapter, her R01lJa flCtC1'1111 now plays the part of a more distant backdrop, or alternatively of a kind of navigational beacon. Her central subject is the encounter between the Peruvian Inka and the Spanish empire (which itselfhad been much influenced by the example of Rome), and Spanish attempts to explore and to comprehend these newly conquered peoples. What MacCormack reveals is how the Roman empire, as it was perceived in sixteenth-century Spain, served the Spanish as a kind of cognitive model against which they could map and measure the Inka, and with which they could understand the alien empire - in turn leal'lling "to do their own job better" (MacCormack, this volume, p. 419). Such recognition took various (>1'1115, ("om the material (the existence of a notable communications system, in which all roads might be thought to lead to Cuzco) to questions of moral and political order, be they positive (the lnka /Cd his people) or negative (Atahuallpa was a tyrant). Reading MacCormack's chapter, onc is left wondering what might have transpired if there had been "no Rome," no means by which the Spanish could begin to conceive of this imperial "other." By f()regrounding the historical lens adopted by the Spanish to scrutinize their conquered,

371

372

Susan E. Alcocll MacCormack quietly makes the case for just how much the resulting perspective mattered to the course of what followed in Peru. That explicit recognition of cognitive models, of course, introduces issues of preconceptions and misperceptions into the study of imperial systems. And that leads, inevitably, to considering the historiography of empire. Just as different imperial paradigms have been held up as yardsticks against which to measure other cultures, so what has been thought important about empires has varied in emphasis over time and from discipline to discipline - from a focus on their heroic leaders, to their economic policies, to their military stratagems, to their ideological systems. The stories empires have been made to tell emerge tt'om the particular context of their production. This volume's own relative lack of interest in previously urgent imperial questions, notably discussions of coercive power and military fl)l'ce, serves very much as a sign of our times, coupled as it is with a distinct shift in emphasis to other issues - on the best use of sources; on empires in their wider "global" contexts; on the integration of non-elite and elite elements; on ideologies and memories. The chapter that lingers longest on questions of historiography is that of Liverani discussing the collapse of the Assyrian empire. He views that "cnd" fl't'll1ation, Liverani never allows an unhealthy division to emerge between notions of the "ideal" and the "real." As with other papers in this volume, what people thought and said about empire is believed to be as valid in its construction and operation as other, more tangible indicators of imperial activity (sce also Brumfiel, Deagan, Woolf~ Yates). Finally, the chapter overtly links ancient and modern interpretations of the collapse of the same empire, exploring the intricate relation between the two or, actually, among the many, since versions of collapse take numerous fClI'Il1S, from the objective to the mythic. The tangled nature of the scholarly tradition engaged with empires where explanations from the past invariably influence interpretations in the present, and where modern paradigms inevitably govern the reading of original soUl'ces - is more or less explicitly recognized throughout this volume. Through their material traces, their legacies and their memories, empires can and do, in the idiom of popular culture, "strike back": at subsequent imperial t(H'lnations that

Part V The afterlife of empires respond to them, at later scholars who analyze them. A single verb there, however, is not quite sufficient. The afterlife of empires can strike, influence, transtC>rIl1, or deceive; it can be used, abused, absorbed, or rewritten. All of these processes arc to be witnessed in this collection, and all require further analysis in future investigations of empire.

373

Thefall of the ASJp'ian cmpire

375

15 The fall of the Assyrian en1pire: ancient and n10dern interpretations Mario Liverani

THE FALL OF ASSYIUA: A KURDISH LEGEND

374

The Assyrian empire reached its peak in the seventh century BC E. Under the great Ashurbanipal it dominated the entire Near East, including Egypt to the west and Elam to the east, even subduing the surrounding nomads the Arabs of the desert and the Medes of the highlands and demanding of them gifts and mercenaries (Figs. 15.1-15.4). The capital city, Nineveh, covering 750 hectares and with some 100,000 inhabitants, was the largest urban concentration ever seen in the world (sce most recently Stronach 1995). The Assyrian army was (or seemed) invincible, the royal treasury inexhaustible. Yet, just a dozen years after Ashurbanipal's death, the Assyrian empire collapsed, almost abruptly, under the attack of old rivals ddeated many times before: namely the Babylonians and Chaldeans ft'OI11 the marshes of lower Mesopotamia, and the Median mountaineers from the Zagros and the Iranian plateau. All Assyrian cities, including the capital, were completely destroyed; the court and state administration, the scribes and the army, disappeared f()rever. On the ruins of Nineveh, squatters established their precarious shelters, and when Xenophon crossed the region a couple of centuries later, he met only villagers and brigands (Dalley 1993). The sudden fall of Assyria was not unprecedented. It was merely the final instance in a long list of Near Eastern empires that collapsed shortly allTr reaching their peak of power and wealth. It came, therd()re, as a confirmation of long-lived meditations about ups-and-downs in the course of history: meditations by Mesopotamian scholars themselves, and later by western scholars rediscovering ancient empires and their f:ltes. But I want to start with a less f:lI11iliar and still problematic explanation of the f:l11 of the Assyrian empire, a popular explanation provided by the l(lUndation legend of the Kurdish people and celebrated every year in their NC)J!roz (New Year) festival (Minorsky 1986: 451, 479-80). As is well known, the modern Kurds claim to be descendants of the ancient Medes. The legend goes that there was once a despotic and "satanic" king, Zohak by name, who suffered from two tumors (in the shape of snakes) on his shoulders.

Mediterranean Sea

o

200

400 km

These he llsed to treat by the daily application of the brains of two children. The vizier in charge of this af1:1ir took pity on the children and let them (or rather, onc of them each day) flee to the mountains, providing the king with a sheep brain instead. Free on the mountains, the children increased in number and became the progenitors of the Kurdish people. Down in the city, Zohak continued his tyrannical rule, until a smith (Kawa by name), whose nine sons had been slain by the king, refused to tolerate his behavior any longer and acted. Hoisting his working apron like a flag, he summoned the escaped children lI'om the mountains, and then all together they attacked and fired the royal palace, killing the tyrant in its ruins. 'T'his all happened on 21 March, the date of the Ncwl'oz, in 612 BeE: the date the Medes entered into history with their destruction of Nineveh. The historical value of such an explanation is highly problematic, of course, since the link between Medes and Kurds seems more a literate than a popular identification, and the precise dating of'the event to 612 BeE depends on the modern rediscovery of the Babylonian chronicle on the f:l11 of Nineveh (published by Gadd 1923). Moreover, the story is just a variant of a well-known chapter in the Persian national epic, as made f:1I1lOUS by Firdausi's Shah mm/ch,

15.1 'Ihe ASlyriml fTNjlil'c, c. 860 nCE. 'I7JC dashcd lille indiClltcs tlJt: extc1It oj' the 1l1lciCllt shol't:iinc oj'thc Persilln GIt((

376

IVfario LiJlcralli

The fall of the AS5.I'J'iall empire

Mediterranean Sea

Mediterranean Sea

oL

15.2 71)( Ass)'rirttl emjlirc, c. 730 BeE. The drtshed litle indicrttes the extmt I!f' the tt1lciclIt shorelillc of thc Petsi rill G 11 (t:

377

---20LO---4~O

km

N

o'----2-'-oo---4~O

1

relating the despotic reign of the monster Zohak (Dahhak) and his defeat by Faridun (cf. Yarshater 1983: 426-9). Yet the story of Zohak and the liberated children had been connected with the origin of the Kurds at least ft'om the time of Masudi's historical work, written in 943. The most detailed treatment of the Kurds' "/c>LIndation legend" is then recorded in the Shm'(~f'111l111t:h, a Persian epic of the late sixteenth century (Charmoy 1868-70),1 long before any modem knowledge of the Median destruction of the Assyrian empire. 2 Above all, the Kurdish legend is able to evoke the secular struggle between city and highlands, between empire and mountain tribes. The legend speaks to the mountaineers' desire jc)r revolt and vengeance against unjust rule fl'om an imperial palace, and to the persistent dream of a spring when oppressed people will finally come down Il'om their refuges, punish the tyrant, and proclaim fi·eedom. Such might have been the kelings of the Median tribes when they descended from the mountains in order to I1ght against the empire of evil. On the other hand, it is not impossible that a decisive event such as the destruction of the Assyrian empire left some trace in the Iranian legendary corpus. In any case, the popular account sets the scene lix our problem, in terms that are those of/()Ik-tale and legend, but that could be easily translated into modem

km

N

t

and sophisticated historical language. We can now leave the legend, and survey the various explanations provided by ancient scribes and modern historians fClr the f:,1I of Assyria, and the fi,1I of ancient Near Eastern empires in general.

THE FALL OF STATES IN TI-IE ANCIENT NEAR EAST Normali~y (~f' collapse

People in ancient Mesopotamia seem to have been well accustomed to cohabiting with collapse. The origin ofthis mental habit can be looked t(lr in their most common building material mud-brick. Mud-brick walls can endure tell' a kw decades (provided they arc replastered every year), but will sooner or later unavoidably f:,1I down: certainly as soon as a building is abandoned and upkeep is interrupted, not to mention when it is subjected to floods, heavy rains, occasional flres, earthquakes, military attacks, and other shocks. The t:,1I of a wall or even of an emire building is, however, nothing exceptional, but would immediately be f()lIowed by Ieveling and rebuilding on the same spot, with the same

15.3 'flJt: AS.I),rillll clII/,il't:, c. 70S BeE. nu' dmhcd !hle illtiimtcs thc extellt I!f' I/;c IIlIcimt shorelillc I!f't/;C Pcrsiall GI/(I:

lvlario Lil'ertHli

378

Tbc fall ofthc AS.lyrit111 empire The same principle applied in the political sphere: kingdoms, dynasties, even empires, do fall down after a while (usually after a short while), soon to be replaced by other kingdoms, other dynasties, or other empires. The very arrangement of the famous and authoritative "Sumerian King List" was such as to underscore the obvious and repetitive event ofa dynasty collapsing, only to have its place filled by another (J acobsen 1939; Michalowski 1965; Wileke 1989). The usual word f()l' dynasty (Sumerian bala, Akkadian /laM; cf. Glassner 1993: 25-6) refers to temporary ofl-ices, to terms or turns in public duties; the dynastic sequence is just a major instance of the same change seen in a sequence of off1cials. The f:l11 of the most prosperoLlS ofall Mesopotamian dynasties, Ur Ill, was explained in exactly these terms:

LYOIA

The judgement of the assembly cannot be turned back, The word of' An and Enlil knows no overturning, Ur was indeed given kingship [but] it was not given an etel'llal reign. Prom time immemorial, since the land was fillll1ded, until the population multiplied,

Mediterranean Sea

Who has ever seen a reign of' kingship that would take precedence 11()I'ever p The reign of'its kingship had been long indeed but had to exhaust itself. (Michalowski 1989: 59) 'Ouma

o

15.4 'UJC A.')yril11l cmpire, c. 640 ne H. 'Un: dashed lille indiwtcs the exlm! l!f' thc Ilncient shore/ille I!f'the PCl'sill11 GII(t:

200

400 km

cheap material. Stories of temples and palaces, so common in middle- and lateAssyrian l'Oyal inscriptions, arc stories of sllccessive blls and restorations, and arc told by the last restorer-king, drawing upon his predecessors' f(lllndation inscriptions: The great wall of'the New City ... which l'utzur-j\shur, my li)rcl~llher, a king who preceded mc, had previously built - to that wall Ashur-bel-nisheshu, 1who was I also my li)rcl~llher applied a l~lcing; it again became dilapidated and 1the wall I together with its gates and towers Eriba-Adad, vice-regent of'the god Ashur, 1who was I also my li)rcl~llher, a king who preceded me, reconstructed in some places and applied a liKing to others; he rebuilt the ruined wall I/'om top to bottom. That wall had become dilapidated I again I and was in ruin. I, Adadnerari, vice-regent of the god Ashur, restored and reconstructed the dilapidated and weakened Iwalll. I rebuilt the ruined Iwalllli'om top to bottom. I made I it Ithe thickness of'li)urtecn bricks 1which were made I in my large brick mold and deposited my steles. I deposited my lill'cI~\thers' stcies with mine. In the f'uture may a later prince, when the wall becomes old and dilapidated, restore it. May he restore my inscribed name, my steles, and my clay inscriptions to their places. IThen I the god Ashur will listen to his prayers.

(Grayson 1972: 434)

Reverting to the building metaphor as applied to a political body, this is something much more than a simple image. Public buildings and entire settlements did lie in ruins as a consequence of the waning or respective politics, and a landscape of ruins offCrs the physical image of' political collapse. In the f:lt110US (just quoted) "Lamentation" over the 1:111 of the Ur III empire and the destruction of its capital city, physical and institutional aspects of the colbpse arc described together. In a sense, the polity is represented as an agency in charge of'the upkeep of' temples and cities, of the abodes or gods and people. In such a view which is quite traditional in Mcsopotamian history collapse becomes so much a part of'the basic structure of events as to present a fI'Il1 of continuity, in a cyclical pattern of time that applies to its minute texture (days and nights, weeks and months, seasons and years), as well as to its more comprehensive fCaturcs (reigns and dynasties). Since collapse is just a shift in of/lce, there is nothing traumatic about it.

Om books of' history usually cnd the chapter on the ancient Near East with its conquest by Alexander the Great, an event obviously considered as epochmaking, even as marking a paramount change in civilization. Yet the Babylonian "astronomical diaries" report thc event without any emphasis, recording it together with the price of basic commodities, and on the same level as other features of daily life (Saehs and Hunger 19~~: l76-9; cC Wise man 1985: Il6-20). \Ve do not have the "diaries" f()r the year in which Babylon was conquered by Cyrus, or the year ofNineveh's 1:\11, but wc can be sure that their recording was quite similar. In f:1Cl wc have the Babylonian chronicles Ire the rediscovery (beginning in the mid-nineteenth century) of the Assyrian and Babylonian empires, attentive travelers in the Near East had been impressed by the visible ruins of large cities and magnificent monuments placed within a deserted and decayed landscape. Meditations on ancient mins - their origin, their nature, their message - became an important factor in shaping philosophies of history in the eighteenth and nineteenth centuries. Most of the travelers were content with adopting the ancient explanation purveyed by the Bible: namely that god's will had wrecked the ancient empires of As syria and Babylonia in vengeance for their oppression ofIsrael (Silberman 1991). The idea of a "structural" or "natural" sequence of empires - with each onc arising from the ruins of its predecessor - was also available in the book of Daniel or in its reworking by later classical historians (et: Machinist 1997: 187, n. 28).

381

382

The fall of the Assyrian empire

Mario Liverani But the most acute among the travelers, of whom Volney is the best known, made use of the old paradigm by applying it to modern reality as well. In their eyes, the present waste was the result of Ottoman misgovernment, in particular corruption and over-taxation (Volney 1792, 1959 [1787]). Thanks to the rational attitude of the Enlightenment philosophers, the old "moral" and "religious" reasons became political reasons, and indeed were soon to become socioeconomic reasons in the materialistic historiography of the late nineteenth century. Such a historiographic evolution is important. Yet it must also be perceived as partly an evolution in terminology, and certainly in the availability of the conceptual tools needed to express one and the same idea: namely tllat empires collapse because of tlleir inner functioning, not just because of outer attack, or because of natural catastrophes. The modern "systemic" approach, started in tile early 1970s by tile Club of Rome with its modeling of the limits of development (Meadows et al. 1972), is just tile most sophisticated - and now current - application of such a line of reasoning. Parallel to tllis "inner" explanation, any account based on the notion of an "outer" shock could be partly alternative and partly complementary. Barbarian invasions or migrations of peoples were obviously put to the fore by "romantic" and later by "nationalistic" ideologies (Bronson 1988). Yet a compromise solution runs through all modern historiography: that an outer shock is just the death blow to a polity already enfeebled by internal, deep-running causes. Jacob Burckhardt put it succinctly: In nature, ruin comes about for external causes only: geological or climatic catastrophes, overcoming of feeble species by more aggressive ones, of nobler species by more vile ones. In history, the cnd is always prepared by an inner decadence, by waning of life. An outer shock is then sufficient to put an cnd to everything. (Burckhardt 1929 [1870])

This short statement already contains the essence of modern theories, in the vein of the Armalesschool, that point to the cumulative effect of longue dude factors, of mid-range factors, and of eJlenememiel factors, to explain the collapse of an empire or of a social order. It also contains an evident polemic against "natural" explanations (in particular climatic causes), which were becoming popular around the beginning of the twentieth century, and which were applied to the collapse of ancient Oriental empires and civilizations. As to the exact nature of this "inner crisis," every culture has its own proposal. The theological misconduct adduced by the ancient Oriental scribes, the moral corruption adduced by the romantic historians, and the socioeconomic contradictions adduced by the Marxist theorists, are all different wordings of the same concept - namely that a state or empire cannot fall down unless it contains within itself the germs of its own ruin. An outer shock, if directed against a sound organism, will be repulsed or absorbed; against a weak organism the same shock will pull it down, with no hope for future recovery.

THE FALL OF ASSYRlA

The protagonists) views Taken in tllis sense, tile fall of Assyria is paradigmatic, since tile available data offer evidence for both an inner decadence and an outer shock. The problem is not why and how the Babylonians and Medes won some battles or stormed some cities; the problem is why Assyria did not recover from tllese defeats. Previously, other battles and sieges eitller marked a temporary crisis, or tlley had simply paved the way for a new dynasty. But after tile shock of 612 BC E tile empire collapsed completely, leaving the core of Assyria depopulated and deurbanized for centuries. What had once been the center of tlle world became a border area between conflicting empires; what had once been tlle most intensively cultivated countryside reverted to tribal pastoralism; what had once been the core ofinternational trade was bypassed as a dangerous crossing-point. Let us begin witll tile protagonists' own views, and the reactions oftlleir contempo~·aries (cf. most recently Machinist 1997). Not surprisingly, the Babylonian rec~rdl11g of tile gL~eat event belongs to the "normalizing" or "continual" point ofvLew. The chrolllcle of the crucial years 614-612 adds no comment beyond a ba~·e an? concise narrative of events, merely underscoring the quantity of casualties wLth a rather stereotyped expression (Grayson 1975: 92-5).3 The victorious king, Nabopolassar, does not celebrate his victory - at least in the inscriptions that have reached us. Is this just a matter of style, or even of chance? Probably n.ot. In subseql~ent years, Nabopolassar's and Nebuchadnezzar's campaigns were a.lIned at ensunng that the new Babylonian kingdom possessed as much as posSible of former Assyrian territories, leaving to the Medes just the unproductive and ever-rebellious highlands, and excluding Egypt from the Levant. In their view, the new empire inherited the old one with no major break, apart from a change in capital city and ruling dynasty. ' Such a view probably also held true in the former Assyrian provinces. At Durkatlimmu, Assyrian scribes were still at work for some time (Kiihne 1993); at Harran, the mother of Nab on id us summarized her long life through a continuous line of Assyrian and Chaldean kings, not even marking the difference: Prom the 20th year of Ashurbanipal, king of As syria, when I was born, until the 4:nd year of Ashurbanipal, the 3rd year of his son Ashur·etil-ili, the 21st year ot Nabopolassar, the 43rd year of Nebuchadnezzar, the 2nd year of AwclMarduk, the 4th year of Neriglissar, during [all] these 95 years etc. (Oppenheim 1968: 561)

Interpretations by the other victorious people, the Medes, could of course have been quite different, provided that we can trust the (already mentioned) foundation legend of the Kurdish Nowruz. To Median eyes, the f.111 of Assyria was the end of a hateful, oppressive rule, an end that saw the granting offreedom to the mountain tribes.

383

",.1

384

Mario Liverani

Although we can only speculate (for lack of positive data) about the Medes' view of the Assyrian collapse, the "hubris-nemesis" interpretation is in any case supported by the biblical recording of the event. In that context, the Hebrew prophets can be rightIy considered as representative of tIle general reaction to tIlis news among the oppressed people throughout tIle empire. Nahum's oracle on tIle fall of Nineveh (Machinist 1997: 181-6)4 is more insistent on Assyrian crimes than on Yahweh's vengeance, theology being largely overridden by the spontaneous expression of joy and relief, vengeance and hatred: Woe to the city soaked in blood, full of lies, stuffed with booty, whose plunderings know no end! There is no remedy for your wound, your injury is past healing. All who hear the news of you clap their hands at your downf.111. For who has not felt your unrelenting cruelty? (Nahum 3.1; 3.19)

It is understandable that any appreciation of the event by the Babylonian scribes, evaluating it as a routine change in dynasties, as natural and unavoidable as the course of the stars in the sky, and any appreciation by the dominated peoples, celebrating the end of oppression and of depredation, would be quite different, even opposed. Babylonian consciousness of this difference in political strategies and in historical interpretations seems demonstrated by their attempt to charge the Medes with all the destruction; I shall return to this point shortly. We do not have - and we badly miss - the Assyrian interpretation. Of course we do not have the Assyrian reaction following the collapse, but we do not even have many data from the last ten years of the empire. Whether such a scarcity of data is just a matter of archaeological chance, or else a symptom of the impending disaster, is a question to ponder. In order to look for the Assyrian attitude, we have to go back to the archives of the last great kings, Esarhaddon and Ashurbanipal. At the very apex of Assyrian power and glory, in a state of unchallenged hegemony, the rulers of the world yet seem to have spent their days secluded in their palaces, in constant fear and alarm. From letters (Parpola 1993) and reports forwarded by astrologers to the king (Hunger 1992; cf. also Starr 1990), we learn that the danger was not expected from outside the empire, but from inside the palace. The kings were tremendously worried about their personal safety, and only secondarily concerned about the well-being of the state. In the recent past, palace conspiracies or succession wars had marked the enthronement of Tiglatll-pileser Ill, of Sargon Il, of Sennacherib, of Esarhaddon himself: all with little or no harm to the survival of the empire. Esarhaddon and Ashurbanipal were therefore as much

The fall of the Assyrian empire

worried about a possible usurper as they were unconcerned about an unpredictable collapse. Nor were they helped by the cultural code in which danger and remedy were expressed, .namely the code of astrology and related disciplines. On tIle one hal~~, tIle tIme-honored lists of omens and tlleir meanings spoke of geopolitical entItIes tIlat ha.d become increasingly anachronistic as time passed, and were in ne~d of up-datIng. On the otller hand, tlle astI'ologers were afraid of communica~ng unf.1Vorable omens to the king, and tI'ied to decode tIlem in tlle most reassunng way: This eclipse wl~ch occurred in the month Tebet, afflicted Amurru [= the Westland]; the king of Amurru will die and his country decrease or else perish. P.erhaps tile scholars can tell some tiling about the [concept] "Amurru" to tile king, my lord. Amurru means tile Hatti [= Syria] and tile nomad land or else C1~aldea. S~meone of the kings of Hatti, Chaldea or Arabia will carry tills sign. Wltll tile king, my lord, all is well: tile king, my lord, will attain his desire, and the deeds of tile king, my lord, are acceptable to tile gods. (Parpola 1993: 287, no. 351)

When ~he omens were patently negative, recourse was made to magical and ritual remedies, the most extreme being the ritual killing of a "substitute king," used as a scapegoat for formal accomplishment of evil signs. Was such. a m.agical and ritual apparatus of forecasts and of remedies the only defel~se agall1st ll1ternal and external political dangers? Probably not, since more practlcalmeasures were taken as well. We can just mention here the traditional re~ourse to elll~uchs (who could not consolidate and transfer their power to hem) for the highest state offices or the later recourse to foreign bodyguards note~ ~e1ow (p'. 391). But such measures confirm that contemporary basic wornes ll1volved mternal and personal security and not the survival of the empire as sl~ch. C?mpared to the hundreds of extant letters sent or received by the king deal~ng wlth.personal and court aff.1irs, just a few deal with border defense and foreign relations, and not even one with tribute income or similar economic matters, or with the administrative arrangement of provinces. Ifwe c~n ela~ora~e a bi.t more upon these data, we get the impression that the I~st Assynan kmgs mhented from their predecessors (above all from Tiglathpllese.r III and Sargon I~) a c.omplex and efficient administrative and imperial ~l1~~I~ll1e, and t1~at they l111ag.ll1ed that such a machine could continue on by mC.I tla alone. It IS. symptomatIC that, while the earlier Assyrian kings had always pel~onally led their troops, Esarhaddon and Ashurbanipal preferred to remain in thclr palace, entrusting the responsibility for success, on the human level, to their officers, and in the ideological sphere, to the gods: Do not fear! - the goddess Ishtar is addressing to Ashurbanipal - Because of your hands which you have raised in prayer, and the eyes which were filled willi tears, I have had mercy upon you ... You shall remain here, where tile abode

385

If.

••

...

386

Mario LiJlerani of Nabu is. Eat food, drink wine, provide music, honor my divinity, until I go and carry out this work and cause you to attain unto your heart's desire. Your face shall not grow pale, )'our feet shall not be weary, your strength shall not f.1i1 in the midst of battle. (LuckenbilI1927: 860-1; cf. most recently Nissinen 1998: 44-5, 53-5)

Following divine counsel, the last Assyrian kings did not "fear" anything from beyond their borders, trusting in their army and their impressive city-walls. Both would prove ineffective on the critical day.

Modern interpretations We certainly could not expect the Assyrian kings (or their counselors) to offer us any clear analysis of the inner weakness of their empire; they would have even less to offer about the structural weakness of empires in general. We could expect, however, that such analyses would be in progress among contemporary scholars dealing with the fall of ancient Assyria. Yet this is not always the case. While the old theological and teleological pattern of empires following each other in linear sequence (still alive and kicking at the end of the nineteenth century! - see R.'1wlinson 1862) was finally dismissed by modern historiography of both the idealistic and positivistic schools, its successor is perhaps surprising. The easiest and most frequent solution to the problem is not to provide any explanation at all. The first edition of the Cambridge Ancient History has a paragraph on "The Fall of Assyria" by the late Sidney Smith (Smith 1925: 126-31); the latest edition of the same work has an entire chapter with the same title by Joan Oates (Oates 1991). Both treat the details of the event, but neither addresses the general problem it raises. The same holds true in the pages devoted to the f.'1ll of Assyria by the late Rene Labat in the Fischer Weltgeschichte (Labat 1967: 93-7), and in similar works. The best treatment is perhaps the one by Paul Garelli (Garelli and Nikiprowedzky 1974: 125-8,239-42), but this is extremely brief and thus cannot discuss the problem in all its complexity. The most recent monograph on the subject, The Fall of Assyria by Stefan Zawadzki (1988), is a lengthy discussion of Babylonian and Greek sources, and of the modern studies based upon them, but the theoretical issues involved are only briefly mentioned and not really confronted. The remark in passing ofGarelli, that the f.'1ll of Assyria remains a "historical scandal" (Garelli and Nikiprowetzky 1974: 125,240), remains valid. By comparison, two volumes have recently been devoted to the subject of "collapse," with special reference to ancient empires. One is the personal work of Joseph Tainter (1988), the other a collection of papers edited by Norman Yoffee and George Cowgill (1988). By pure chance both appeared in the same year, and not by chance both adopted a "systemic" approach, which Tainter develops more explicitly and analytically and which is left more in the background in the other volume. The fall of Assyria is not particularly analyzed in either book; while both include Assyria in their sample of historical cases, they

The fall of the Arsyrian empire devote to it only a short and rather vague treatment,S and one in which the proper "systemic" features of internal crisis are hardly mentioned. The most pertinent remark is Norman Yoffee's comment that the Assyrian collapse was total because "the state had systematically eradicated its own infrastructures" (1988: 67; see below, p. 389). Responsibility for such a situation does not rest on the authors or editors of tile se general and generalizing works. Ratller it rests on tile Assyriologists who failed to produce any valuable analysis on tile specific case of Assyria. It should, however, not be too difficult to set tlle old pattern of inner decline plus outer shock into modern terms. Far from being a "historical scandal," the fall of Assyria could and should become a classic, paradigmatic case. In comparison to the two most frequently considered cases of ancient collapse, the Roman empire and the lowland Maya, Assyria even has some advantages. First, it was much more sudden and total than the end of the Roman empire, and second, it is better documented (through both textual and archaeological data) than the vanishing of May a civilization. As to tile inner crisis, a proposal of Tainter's can be utilized, one suggesting that collapse be viewed through tile (mostly economic) issue of decreasing returns on complexity. In more practical terms, the most important factor to be dealt Witll, in a complex but fundamentally agrarian state like Assyria, is the quantitative relationship between producers (basically peasants) and non-producers (chiefly palace officials, bureaucrats, priests, and full-time soldiers). General trends in the evolution of the Assyrian empire seem evident: the growth of the palace apparatus was less and less sustained by parallel growth in revenues from internal production, provincial taxation, and war booty. The very same process of building and maintaining the empire simultaneously brought about an excessive exploitation of the conquered periphery, more and more depleted of its human and economic resources. The demographic, agricultural, and even cultural crises of the conquered (and formerly highly developed) territories of Syria, Palestine, and southeast Anatolia are well known.

The inner decline: looking for (%ard)) evidence But such general statements remain unconvincing unless they are supported by "hard" evidence. The most obvious parameter to be considered, of course namely the percentage of the Assyrian GNP devoted to maintaining the palace and its bureaucracy - is not available as such in the extant data. Nor is it possible to calculate tlle number of officials and soldiers in order to compare their size with that of the rural population. Yet some otller parameters or indices are indeed available, from texts and from archaeological data, which could provide a clearer idea of declining resources and increasing exploitation. The detailed collection and analysis of the relevant data have not yet been done, and could not be accomplished for this chapter given tile montlls and months of work necessary. I will limit myself, therefore, to tracing in general outline the possible analytical tools that we could use to study the collapse of Assyria.

387

388

Mario Liverani

The first obvious parameter is the empire's territorial growth. If put on a graph, it would certainly fit the characteristic outline reproduced by Tainter for the Roman and Ottoman empires (1988: 125), with tile later part of Ashurbanipal's reign already in tile descending parabola and soon to be followed by a vertical collapse. In Tainter's terms, the annexation of outer territories, which temporarily provided increasing resources for tile increasing costs of tile imperial machine, became more expensive as conquest pushed into more remote lands. This led inexorably to a groWtll of military expenditures well above tile economic returns of tile war activity, and to a groWtll of provincial administrative expenditures well above tile economic returns of tribute gathering. A second quantitative parameter tllat could be roughly calculated, at least in contrasting tile data of tile eightll century to tllOse of tile seventh century, is the groWtll of tile Assyrian capital in comparison to the decline of settlement in the provinces. Tony Wilkinson (1995: 148, fig. 7) has recently attempted such a calculation in a few sample areas of western Jezira. It is quite evident that the groWtll of tile Assyrian capital city (from 320 hectares at Dur-Sharrukin to 750 hectares at Nineveh), far fi'OJl1 being part and parcel of a general trend in neoAssyrian settlement, was ratller its reverse; the evidence suggests instead tllat tile countryside was becoming more and more depopulated. Of course tile capital city was inhabited primarily by non-producers (state officials of various kinds), while food-producers dwelled in the increasingly depopulated provinces. Anotller calculation could be made regarding the influx of deportees and booty from conquered lands to Assyria (Oded 1979: 20). In the empire's expansive phase (nintll and eighth centuries), such an influx compensated for the costs and casualties of conquest; while in its declining phase (seventh century) the influx came to an almost complete stop, while the costs remained as high as before. We could also easily put on a graph either the average, or the maximum, distance between military activities and either the Assyrian capital, or the border of the empire in successive periods. Quite clearly the maximum distance of the army from the core of Assyria was achieved with the campaigns of Esarhaddon and Ashurbanipal to Egypt, to Arabian Bazu, and to Elam. This underwent a sudden contraction, however, with most battles in the final two decades before the empire's end taking place within its borders. A quite different but interesting indicator, also accessible through extant data, would be the amount of work done in public building (e.g., temples, palaces, armories, city-walls), tile most conspicuous investment of human and economic resources in ancient Oriental empires. The measurement could be based simply on square (or possibly cubic) meters of masonry, or else converted into the more significant standard of man-days of work. It is easy to predict that any graph of such data would roughly reproduce that of the territorial extent of the empire, Witll an accentuated rise-and-fall parabola. Apart from being a symptom of the availability of public resources, such "conspicuous consumption" (to use Veblen's term) is in itself a possible cause for the eventual decline.

The fall of the Assyrian empire

Finally, we could, without too much trouble, place the empire's growth on a set of maps in chronological sequence, in order to underscore one very interesting feature. The expansion took place basically in the lowlands, to the west and SOUtll, while tile nortllern and eastern borders of the empire remained almost stable or were extended only at very high costs and with very unstable results. The reason is quite obvious: the mountains were not easily conquered, nor were they easily exploited and controlled. But the consequence of such asymmetrical growth is also evident, leaving as it did the dangers of highland tribes only a couple days distant from tile Assyrian cities and the royal palace. This situation endured until tile very end of tile empire. The long-lasting fight - between the cities in tile plain and the tribes on the mountains - was, of course, unequal. The military, economic, technological, and demographic balance was all in favor of the lowland states. The mountaineers possessed, however, the advantages of tlleir geographical and socioeconomic mobility and elusiveness. In military conflicts, tile Assyrians did win dozens, hundreds of times, without being able to eliminate the problem; the mountaineers won just once - but that was enough to destroy the Assyrian empire.

The outer shock - coming from two opposed sides So we arrive at the problem of the outer shock, tile blow that led to the fall of a declining empire. The war between Assyria and a coalition of Medes and Babylonians has attracted the almost exclusive attention of Assyriologists, not least because it is explicitly recorded in the Babylonian chronicle (cf. Gadd 1923: 1-26). In current interpretations (as recently exemplified by Zawadzki 1988), military defeat is explained by the enfeeblement of Assyria during the succession war between Ashur-etil-ilani and Sin-shar-ishkun, following the death of Ashurbanipal. Such a diagnosis is certainly valuable. The rise of Nabopolassar took place in the framework of this succession war, with the internal Assyrian dispute offering a f.worable occasion for the Chaldean chief to free Babylonia and eventually to attack Assyria itself. Ashur-etil-ilani, already enfeebled by the long contest with his rival, proved unable to resist the new opponent and was defeated. In this way, not only is the final shock, but also the previous decline, reduced to short-term events; no place is left to longue duree or middle-range explanations. But a crisis connected to a war of succession was by no means new to Assyria, and in the past it had always been followed with the state's reorganization by the winning party. This recurring, almost structural, kind of crisis could not explain - by itself - a total collapse. That total collapse is basically to be explained by the previous long-term decline along the lines summarized above (pp. 387-8; Yoffee 1988).6 But I would add that the quite peculiar circumstance of a double outer shock, occurring at the same time from two directions, from both the Medes and the Babylonians, should also be considered. In fact the role played by the Babylonians was not too different from that of another contending party within

389

.

•

:.'I ~ ~I

..

390

Mario Liverani

Assyria itself. The victory ofNabopolassar alone would have resulted in his assimilation of Assyria within the Babylonian kingdom, in his inheritance of the empire, and therefore in his reorganization - not his destruction - of it. Unfortunately for Assyria, the intervention of the Medes dramatically altered that scenario. The Medes had no interest in inheriting the empire; they did not possess the necessary cultural traditions and administrative competency to run an empire. They belonged to a quite different cultural tradition - that ofmountaineers lacking urban centers, centralized states, bureaucratic statecraft, or literacy. Well able to destroy, they were less able to reconstruct. The different, even opposed, strategies of the Babylonians and the Medes toward Assyria are clearly demonstrated by the former's subsequent attempt to blame all the destruction on the latter: (Marduk) provided him [= Nabopolassar] with helpers, made him acquire a friend and caused the king of the Manda-hordes who has no rival, to bow to his orders in submission and to come to his assistance. [And] he [= the king of the Manda-hordes] swept on like a flood storm, above and below, right and left, avenging Babylon in retaliation. The king ofthe Manda-hordes, without [religious] fear, demolished the sanctuaries of all the gods of Subartu [ = Assyria]. He also demolished the towns within the [present] territory of Akkad [ = Babylonia] which [at that time] had been hostile against the king of Akkad and had not come to his assistance [in his fight against Subartu]. None of their cult [ -centers] he omitted, laying waste their [sacred Jtowns worse than a flood storm. The king of BabyIon, however, for whom this sacrilegious action ofMarduk was horrible, did not raise his hand against the cult [ -places] of any of the great gods, but let his hair unkept and slept on the floor [to express his pious desperation]. (Oppenheim 1955: 309)

History repeats itself, with the role of the aggressive and oppressive empire, once typical of Akkad, now resting on Assyria. The role of the barbarians, summoned down from their mountains to punish hubris and pull down the empire, once placed on the Gutians, now rested on the Medes. And the role of the restorers of order and civilization (expressing scandal for the "dirty job" of destruction, but profiting from it) rested once on Ur, and now on Babylon. To revert to the traditional explanations of collapse of the peoples involved, the Babylonian version would be that of the "normal" fall of a dynasty, to be followed by another dynasty and another empire. This is clearly paralleled in actual Babylonian policy, which was aimed at inheriting the empire and continuing to run it along preexisting lines. On the other hand, the Median version, one of vengeance against oppressive exploitation, is well reflected in enraged and implacable forms of destruction. Seemingly underscoring this point is the fact that the tablets containing the text of treaties linking the Medes to the Assyrian king have been found purposefully crushed in small fragments (Mallowan 1958: i). And Carl Nylander (in an unpublished lecture; cf. Nylander 1980) has noticed in the sculptured reliefs of Nineveh the intentional mutilation of those personages involved in the Assyrian victory over Elam and other Iranian peoples.

The fall of the Assyrian empire

I have to add one new, but important, detail to the received scenario. Attentive analysis of the well-known treaty between Esarhaddon and the Medes reveals that Medes had been recruited to serve as a corps of palace guards, and in particular as the personal bodyguard of the crown-prince at the end ofEsarhaddon's reign (c. 675 BCE; Liverani 1995: 57-62). They continued to serve in this fashion under Ashurbanipal, probably until c. 650 when the Median tribes in the highlands decided to form a kind of confederation and to remove themselves from Assyrian overlordship. This detail- the existence of a Median guard corps in the Assyrian royal palace - is important, because it could have been a factor both in the modernization of the Median army, and in their acquaintance with the internal functioning of the Assyrian defense system. The sons or grandsons of those Medians who had served inside the Assyrian capital cities were the men who stormed the walls of Nineveh and destroyed the royal palace. Later historical interpretation by Greek and Roman authors conferred a different role on the Medes. Working within the framework of a preconceived sequence of empires, for them Media had to find a place after Assyria and before Persia, and to be placed on the same footing as its predecessors and successors. The existence of a "Median empire," long maintained by modern scholarship, has only recently been put in doubt by Helene Sancisi-Weerdenburg (1988: 201-3). Reality, in fact, could have been quite different from tile traditional view. The Median intervention produced a real break in political tradition, changes articulated under the banner offreedom, ethnicity, and a chiefdom level of polity. This is not the place to reconsider the true nature of Median hegemony in the highlands, but to me it clearly seems quite different from tile imperial tradition inherited by Babylonia. I think that we can evaluate the Median hegemony over the highlands - the fifty years from tile fall of Nineveh to tile fall of Hecbatana - as an interlude of anti-imperialistic f1avor, dominated not by the aggressive and exploitative attitude of the lowland states, but by the highlanders' rules of hospitality and gift-exchange, inter-marriage and alliance, chivalry and bravery. It was an evil f.1te for Assyria that the very boundary between Babylonian and Median territories, established after the destruction ofNineveh, passed through the core lands of Assyria. Thus, the area that had formerly been the center of the empire and of the world became a borderland between two different political orders and customs: the urbanized and bureaucratic polity of tile plains, and the tribal and pastoralist polities of the mountains. Continuity and total collapse are not alternative (contrasting) issues. In our case they were the different results in the two halves of Assyria, assigned to victors coming from two opposed political environments and traditions.

391

The Carolingian empire: Rome reborn?

16 The Carolingian empire: Rome reborn? John Moreland

King Charles [the Bald], hearing of his brother Louis's illness, invaded the kingdom of Lothar and disposed of it as he wished ... Furthermore, taking the advice of evil men, he had a crown set on his head in the city of Metz by the bishop of that city [Adventius] and ordered that he should be called emperor arid augustus as Olle ",ho ",as to possess two killgdoms. (The Atmals of FlIlda [869 CE]; cited in Reuter 1992: 61, emphasis added)

..il

INTRODUCTION

"

392

Charles the Bald's opportunistic seizure of his brother's kingdom was a particular event in the late history of the Carolingian empire and of the civil wars that characterize it (see Table 16.1; Nelson 1992). Charles' order that he should be called emperor because he "was to possess two kingdoms" nevertheless provides an insight not only into the Carolingian concept of empire, but into the idea of empire more generally. It is clear from this, and from earlier Carolingian history, that these early medieval rulers of Europe differentiated between kingdom and empire, and that they saw the latter as the product of the incorporation of several of the former under a single ruler. Although the immediate basis of Charles' claim to imperial status was the addition of his brother's kingdom to his own, we should bear in mind that his grandfather was Charlemagne - the greatest of the Carolingian emperors - and that his father, Louis the Pious, had been emperor of the whole Carolingian realm. Charles became king of a portion of that empire following Louis' division of it, in 839, between his two sons (Nelson 1992: 99-100). The desire to be called "emperor and augustus" can thus be seen as a desire to recreate the past. As we shall see, these factors - the incorporation of other kingdoms and emulation of the past - played a fundamental part in the construction of the Carolingian empire. The largest and most complex polity in western Europe since the fall of the Roman empire, at its height in the late eighth and early ninth centuries CE the Carolingian empire covered some 1,200,000 sq. km and stretched from Denmark to northern Spain, from northeastern Germany to Rome, and eastwards beyond the Danube (Einhard 1969: 15; James 1988: 230; Leyser 1994:

393

Table 16.1. Principal rulers mentioned in the text) with their period of rule and {(ethnic affiliation»j all dates AD Ruler

Affiliation

Dates

Constantine

Roman emperor

312-37

Theodosius

Roman emperor

379-95

Childeric

Merovingian king of the Franks

died 481

Clovis I

Merovingian king of the Franks

Charles (Carolus) Marte!

Carolingian mayor ofthe palace

481-511 714-41

Childeric III

Merovingian king of the Franks, deposed by Pippin Ill, mayor of the palace

743-51

Pippin Ill, son of Charles Marte!

Carolingian mayor of the palace king of the Franks

741-51 751-68

Charles (Charlemagne), son of Pippin III

Carolingian king of the Franks emperor

Louis (the Pious), son of Charlemagne

Carolingian emperor

768-814 800-14 814-40

Lothar, son of Louis the Pious

Carolingian co-emperor

817-55

Louis II, son of Lothar

Carolingian king ofItaly co-emperor

Charles Il (the Bald), son of Louis the Pious

Carolingian king emperor

840 850-75 840-77 875-7

7; see Fig. 16.1). It was created by the expansion of the Arnulfings, a major Frankish lineage whose heartland lay in the Rhine, Moselle, and Meuse region and who became known as the Carolingians after one of their most successful members Charles (Carolus) Martel. In the seventh century, they had held senior offices (as the major domus, or mayor of the palace) at the court of the sacred kings of the Franks, the Merovingians; by the early eighth century the Carolingians ran the Merovingian kingdom, and shortly afterwards took the throne in a "palace coup." Carolingian expansion initially involved the subjection of other Frankish groups, then other Germanic peoples or gentes (singular gens). The result was a heterogeneous empire: a mosaic of peoples and power. Many of these peoples tile Burgundians, Alemans, Bavarians, and Saxons - maintained their political, cultural, economic, and linguistic distinctiveness (McKitterick 1983: 18); some kept their own laws (Nelson 1995: 412). Non-Christian religious practices persisted, especially in tile eastern part of the empire (Reuter 1991: 51-69; Wood 1994: 319). Several factors (some common to all imperial systems, others more or less unique to this group of Franks) tllreatened Carolingian dominance. First, there was tile problem of succession. This refers not only to attempts to avoid the crises which seem often to accompany the transition from one ruler to

394

John MOI'eland

The

North Sea

SAXONY • Paderborn

.

Ingelheim •

SI. Denls.

,'"~ 7.:,:::::"h

• Orleans

Bay

BAVARIA

of

AQUITAINE

Biscay

BURGUNDYC:::>

GASCONY

o

Mediterranean Sea

16.1 The Carolillgial1 empire, c. 800 CB, including the border

regions

(~narcJJeJ).

[)

Rome reborn?

another (see MacCormack, this volume, p. 433), but also to the particular t:1Ct that the Carolingians had usurped power from a legitimate line of kings. Second, they had to confront the tact of the ethnic (and religious) diversity of the empire they had constructed. Such heterogeneity was not peculiar to the empire of the Carolingians, of course; in f:lct it may be said to be constitutive of empires, and the Carolingians certainly used it as an argument fi)l' their imperial status: "he should be called emperor and attIJltstus as one who was to possess two kingdoms" (see below, p. 415). Third, they had to deal with the fact that although they claimed imjlcrill11l over the peoples of the empire, their power was mediated through sub-kings (often their own sons) and nobles. These people held instrumental power and had to be bound to the imperial cause. The Carolingians were no doubt only too aware that they had emerged ft'om the nobility (Werner 1979: 147), and what they had done, others could do. The Carolingians (with varying degrees of intentionality ) implemented a series of strategies to cope with these problems, outlined in the tiJllowing section under three broad headings: administration, warf:ll'e, and the economy. Next I will argue that the Carolingian Renaissance, a major florescence in the arts, architecture, and learning, can be seen as another potentially integrative strategy. Two principal themes of this Renaissance - the rebirth of Rome (Roma re1lascens), and the Franks as the Chosen People of God (jlOjlllittJ Dei) - were embodied in material culture (defined to include texts), and the latter was designed to spread an imperial message abroad (cf. Moreland and van de Noort 1992; see a\so DeMarrais et al. 1996). This Renaissance can be seen as "art f(H' art's sake," but to present it solely in these terms is to impose modern notions of the neutrality of art 011 to the past. Here I contextualize the material culture of the Carolingian Renaissance and emphasize that it was designed to play an active role in the reproduction of the Carolingian empire (see Zanker 1988 fix similar ideas about the Roman empire; and Brumfiel, this volume, ()r the Aztecs). The strategies implemented by the Carolingians could be said to represent the ideal structuring of society and the world, intended to overcome the real divisions of people and power which were inherent aspects of empire. Prom time to time I will point to Ihat reality and to the weaknesses of the strategies of integration. J. Smith has recently argued that we can clarit)! our image of the Carolingian core by f()cusing on its peripheries (J. Smith 1995: 169). In the later sections of the chapter, I explore how the Carolingian ideal was played out in the midst of the reality of disputed claims to territory and lordship on the empire's southern borders in Italy. Notions of the rebirth of Rome might have possessed a greater resonance so close to the heart of the old empire. In conclusion, a detailed study of the art and architecture of the region suggests that earlier assertions of the efficacy of Carolingian ideology (I-lodges et al. 1985) not only overestimated its power, but also simplified the complex relationships which existed between the ninth-century present and the Roman past. The Carolingians were not the only people to reach across the ages to find legitimacy in the "glory that was Rome."

395

396

John More/and

The Cal'o/ingiall empire: Rome reborn?

397

CONSTRUCTING THE CAROLINGIAN EMPIRE

Administration The Carolingian empire was focused on a series of palaces which the emperors visited as part of their peripatetic routine (Fehring 1991: 126-32; Reuter 1991: 24). Those at Aachen, Paderborn, and Ingelheim (Figs. 16.2 and 16.5) were favored although, by the latter part of Charlemagne's reign, royal government was increasingly based at Aachen (Nelson 1995: 386; Innes 1997: 845). Frequently, it was at or near such palaces that the royal host was raised and assemblies met (McCormick 1986: 367; see below), and by contemporary standards these were truly impressive structures (see, for example, the poet Ermold's early-ninth-century description of Ingelheim, eited in Dutton 1993: 265). Equally impressive were the numerous monasteries and churches either constructed or enlarged (Einhard 1969: 17; Heitz 1980). It is clear that, along with the palaces, they constituted a major investment in the architectural aggrandizement of the empire. To paraphrase McConnick on the role of Roman architecture (1986: 25), we might see the palaces and monasteries as "Carolingianising the architectural landscape of the territories and relaying the message of imperial unity." Resident in the palace, or en route from one to another, the emperor stood at the heart of a system of patronage which to some extent tied the elites to him and which, on occasion, brought this nobility bct()re him. It was at the palace, and in front of the emperor, that the noble assembly was held. Assemblies could be called only by the emperor (Rosenthal 1964: 30, 39); the main reasons f()l' issuing such a call were "to gather the Frankish host ... to discuss political or ecclesiastical matters ... and to act as an assembly at which judgements were made" (McKitterick 1983: 97). Rosenthal (1964) stresses the essentially "passive" role of these assemblies and suggests that, rather than being a decisionmaking f()rum, they were occasions on which the emperor announced his own judgments (but see Innes and McKitterick 1994: 214). On the other hand, Nelson suggests that they constituted a "shared experience that surely reinforced participants' sense of themselves as a group" (Nelson 1995: 417-18; see also McCormick 1986: 367; T. Noble 1990b: 344), a cohesiveness perhaps encouraged by proximity to the emperor. It was at sllch "public" occasions that, through gesture, pronouncement, and speech, Carolingian emperors directly proclaimed their links with the imperial Roman past (lImes and McKitterick 1994: 214). In addition, the venue - the palace - provided visual reminders of glory and the exercise of power. The walls of the palatial hall at Ingelheim were decorated with scenes oLFrankish victories won under the Carolingians (Nelson 1995: 419; described in Dutton 1993: 265). Even if we accept Rosenthal's passive image of the Carolingian assemblies, it does not negate their potential importance for imperial integration; they could

still be f~1Ctors in creating a sense of belonging which transcended the level of the lJens. Indeed it has been suggested that gatherings, such as assemblies "at which a political collectivity was !(mlled ... Iwere Inot only a central agent in the creation of 'national identities', but also the context in which the germs of 'national history' can be detected" (Innes and McKitterick 1994: 214). However, we must always bear in mind the very "personalized," l~lCe-to-l~lCe nature of power relations. Just as the emperor stood at the center of an imperial system of patronage and power and attracted people to him, so sub-kings and nobles OCCllpied similar positions at the regional and local levels, and we can envisage similar centripetal tendencies operating to enhance their power and to threaten that of the imperial core. Even within imperial assemblies the potential f(l" disintegration was present. Some accounts of their structure suggest that those who attended did so as Franks, Saxons, Bavarians, Alemans, etc. (Reuter 1991: 91), and that when the host was summoned, ethnic groups maintained their separate identities within it (Nelson 1995: 4(7). vVhether or not the emperor announced his decisions at the assembly, or whether he consulted the assembled elites, in essence much of the Carolingian administrative system was in the hands of people, such as counts, who exercised effective power at the local and regional levels (McKitterick 1983: 87-97). Counts were meant to act in three governmental capacities: to keep order in their localities; to look after the royal estates; and to Sllmmon men to the king's host

16.2 Rcc(J1lstmctilJ1l (if/he palacc at lt1llc/bcim.

398

John Moreland (Nelson 1995: 411). Although theoretically official representatives, the counts held instrumental local power and "may not always have distinguished clearly between their 'official' and their personal power" (McKitterick 1983: 87-91; Reuter 1991: 27-8). Potential abuses were meant to be controlled by the missi dominici (royal agents) who were sent out to communicate information to the counts and to check that orders were being carried out. As we shall see, the Carolingians were perfectly aware of tlle potential for fission provided by tlle nobility, and Charlemagne used some of tllem to his own ends as missi (McKitterick 1983: 93). However, like the counts, tlle missi were "aristocrats first, royal agents second" (Nelson 1995: 412). Moving from the agents to the instruments of administration, we must consider the importance of capitularies, or administrative decrees divided into chapters (capitula). These were among tlle written instructions tllat tlle missi carried tlU"oughout the empire (McKitterick 1983: 61-9; 1989: 32-5). Yet tlle capacity of this "capitulary legislation" to tell us about the cohesiveness of the polity is limited. What they tell us is "what should have happened. Ifwe take it to describe what did happen ... we can construct a picture of a very elaborate state" (Reuter 1991: 27; origillal emphasis). The point is tllat tlle counts could, and did, choose to ignore these imperial instructions: Neverth~less, it is noteworthy that from the late eightll century tllere was a vast increase in the use oftlle written word (McKitterick 1989: 33; Reuter 1991: 27; Wickham 1995a: 511). Even if the ideal frequently departed from the real, texts such as the capitularies demonstrate the Carolingian desire for order and control, and tlleir appreciation of the use of writing in achieving that end (Giddens 1981: 95; Goody 1986). The Carolingian infatuation with writing, and its connection with order and control, is exemplified in their obsession with lists: "lists of peasant labourers, lists of those for whom to pray, lists of estates, lists of possessions, lists of books, lists of the dead" (Innes and McKitterick 1994: 200; see Goody 1977: 74-111). The strength, and the weakness, of the Carolingian administrative system was that it was mediated through the aristocracy. The empire would last as long as the benefits of "belonging" outweighed those of operating outside the system. The problem for the Carolingians, as for many imperial peoples, was the need perpetually to renew these benefits.

Warfare At its zenith, the Carolingian empire comprised a reasonably coherent entity, with well-defined and defended borders (Reuter 1990: 393-4; J. Smith 1995: 177). At its core were tlle lands oftlle Franks; tllese were surrounded by a ring of conquered non-Frankish regna (kingdoms) such as those of Alemannia, Aquitaine, Bavaria, and Thuringia (Fig. 16.1). On the periphery lay tlle lands of peoples intermittently subject to the core (T. Noble 1990b: 335; J. Smith 1995: 171). It is significant that the empire did not expand further. Considering tlle

The Carolingian empire: Rome reborn? process of expansion, the reasons for its cessation, and the consequences of consolidation provides insight into Carolingian imperial strategy. While the warfare waged by tlle early Carolingians was on a scale not previ0usly seen in tlle early middle ages (Reuter 1985: 89), it remains the case that the expansion of tlle Frankish regnum, and tlle creation of Carolingian imperium, was tlle work not of a standing army, bu t of war bands composed largely of an elite whose peripatetic existence was dominated by fighting, feasting, and conspicuous consumption (Reuter 1985: 91). AltllOugh it is clear that the integration oftlle Carolingian empire could not have depended on tlle military alone (Nitz 1983a: 173), tlle same is true of many otller imperial systems. As in the administrative sphere, the Carolingians were well aware of the potential for fission from local or etlmic powers and took steps to prevent tllis happening. In tlle territory oftlle Frisians tlley confiscated Roman fortifications which were still in use (Heidinga 1990: 27) while in other areas (Aquitaine, Saxony) fortified sites predating tlle Carolingian conquest were eitller not rebuilt or destroyed (McKitterick 1983: 51). Expansionist warfare contributed to the reproduction of tlle empire in other ways. Apart from constituting another venue for tlle kind of collective action seen in the assemblies, and adding to the glory of the Frankish people tlll"ough the submission of otller gentes, warfare produced plunder, which could be given as gifts to the Frankish nobility, and land, with which the high aristocracy could be rewarded (Reuter 1985: 80-1, 88). Both served (at least in theory) to bind the elites to tlle imperial core (Nelson 1995: 387). The extraction of tribute from subject peoples could be used to the same ends. In 795 and 796 Frankish forces attacked the heart of the Avar khaganate (tlle so-called "Avar ring"), seized the vast treasure stored there, and redistributed it to Frankish churches and nobles (Fig. 16.1; Fouracre 1995: 104). It has been noted that this was "the last really large aggressive military operation conducted by the Carolingians" (Reuter 1990: 391). In the years after 800, despite occasional warfare along the imperial peripheries (T. Noble 1990b: 339-40; J. Smith 1995: 172-5), the sources suggest a new emphasis on consolidation and defense (Reuter 1990: 393). This had profound implications for imperial integration and the nobility's sense of "belonging," not least because consolidation meant tllat there were no longer quantities of treasure and land with which to reward the elites. In these circumstances, the only way elite demand could be satisfied was by "internal expansion ... by increasing one's share of political power and tlle rewards which went with it" (Reuter 1990: 405). One of the consequences was increased exploitation of the pauperes (the poor), who "were on the defensive ... against aggressions by all kinds of power nIl people ... even (or especially) counts, the nominal guardians of the weak" (Wickham 1995a: 534). The threat to Carolingian integration posed by their inability to reward the elite would have been nlrtller exacerbated by the increased oppression of tlle peasantry, leading to tlle removal of any sense of "belonging" which either may have felt for the imperial system.

399

.J

b U

I1

..:r

400

John More/and

The economy Nelson has commented that plunder and tribute constitute a very fragile basis for social reproduction (1995: 393) and, if we accept Reuter's arguments that the end of empire can be linked to the non-availability of plunder and tribute (1985; 1990), she is certainly right. Here, however, I want briefly to describe some developments in the economy that might be seen as enhancing the integration of the Carolingian polity. Recent archaeological research shows that long-distance trade complemented plunder as a source of gifts for elite distribution. This trade was mediated through coastal sites such as Dorestad in Holland (Besteman 1990: 107-10; Hodges and Whitehouse 1983; Lebecq 1997). These emporia frequently had tlleir origins in tile Merovingian period (Wood 1994: 293-9), but archaeology reveals substantive changes in tile late eightll and early nintll centuries, demonstt·ating tllat tlley were centers for craft production as well as exchange (Hodges 1989; van Es 1990: 173-5). The importance of control over both the production and tile exchange of gifts for tile construction of elite power has recently been highlighted (Gosden 1989). Charlemagne may have attempted to impose standardization and anotller mode of economic integration through his coin reforms of 793/4, which imposed rigorous controls on minting (Grierson and Blackburn 1986: 206-10). This need not, of course, mean that the Carolingian empire was integrated through tile operation of a monetary economy. The importance of gifts, suggested equally by tile emporia and by the plundering of the Avar ring, means that tile reformed coinage operated within an economy still structured on "traditional" principles (see Mm-eland 1999). The practice of establishing a loyal elite in conquered areas seems to have been a strategy widely used in imperial systems and the Carolingians were no different_ Steuer (1989: 101) notes that with the conquest ofItaly in 774, 360 Franks, 160 Alemans, 15 Bavarians, and two Burgundians were appointed to important posts within the Italian kingdom (also Delogu 1995: 306). But we should be careful not to overestimate the extent to which the Carolingians rewarded members of their own people in this way. Another familiar imperial strategy is the incorporation of native e1ites (Woolf, this volume), and in this context we should note that, in areas conquered by the Carolingians, Werner (1979) has been able to demonstrate a continuity in positions of power of both Merovingian and the descendants of Gallo- Roman elites. Given the close link between church and state (see below, p. 402), and the previous importance of Christian missionaries in the eastern part of the empire (Wood 1994: 304-21), it is not surprising that the church was also involved in the process of colonization tile re (McKitterick 1983: 60). After the conquest of Saxony and Bavaria, the monasteries of Lorsch, Fulda, and Amorbach were granted lands. As well as providing a Carolingian "presence" in disputed territory

The Carolingian empire: Rome reborn?

- just as the Spanish missions provided botll a religious and political presence on the frontiers oftlleir American empire (see Deagan, this volume, pp. 188-9) - tile monasteries developed sophisticated patterns ofland use and settlement allowing increased exploitation of tile se areas (Nitz 1983b). It is noteworthy that surveying was one of tile "minor arts" studied by the monks at Fulda, providing a rather direct example of tile link between learning, the creation of order, and an intensification of production (Contreni 1995: 745). Transformations in tile rural productive system can be seen in otller regions. In tile Veluwe (in centt·al Holland), Carolingian-period expansion into formerly unoccupied areas was probably connected to the massive iron resources of the region (Heidinga 1988), forming part of a more general pattern of eighth- and early-nintll-century movement into formerly marginal zones such as peat bogs and salt marshes (see, for example, Besteman 1990: 106). It is difficult to tell to what extent this expansion was centrally directed or was tile result oflocal initiative, but tile overall trend was toward economic expansion. We should not underestimate the significance of writing in generating and monitoring this expansion (Verhulst 1995: 490). The managers of royal estates were required to present accounts to the king three times a year and "if they were not numerate tllemselves, they would need to find help" (Nelson 1995: 410; for examples of such accounts, see Dutton 1993: 75-6). The resources thus generated, along with those which flowed from the long-distance trade system, no doubt funded parallel changes in the fabric of what we might call the "superstructure" of Carolingian society.

REWRITING THE PAST AND THE PRESENT Although de facto rulers of the Frankish realm, an impediment to the legitimacy ofCarolingian power lay in their having broken the natural line of succession (G. Brown 1994: 45). To overcome the comparative "shallowness" of their claims to power, they initiated a series of short- and long-term strategies to embed themselves more firmly as legitimate rulers. These strategies essentially involved (re )constructing the past, reworking the relationship between the past and the present, and seeking divine sanction for their rule. In 750 Pippin Ill, "with the consent of the Franks," sent messengers to Rome to seek Pope Zacharias' approval for his replacement of the sacred kingship of the Merovingians (Nelson 1994: 54). Zacharias was being asked to "provide the divine authority which could overrule human tradition" (Fouracre 1995: 96). The pope gave his consent and, at Soissons in 751 - "by the election of all the Franks ... by the consecration of bishops and by the acknowledgement of lay magnates, by the Queen Bertrada as the rules of ancient tradition require" Pippin was "elevated to the kingdom" (cited in Nelson 1995: 424). The Carolingian dynasty had been inaugurated; the pope's consent and the anointing of Pippin by the Frankish bishops signaled the creation of a pact between the

401

402

The Cal'o/i1lgiall empire: Romc rcborn?

Morc/and Franks and the church. This connection was reint()l'ced in 754 when Pope Stephen anointed Pippin's t~lIl1ily (Fouracre 1995: 98). The "spiritual brotherhood" thus created between the papacy and the Carolingians prot()Undly influenced the latter's perceptions of kingship and society, and meant that they could be called upon to protect papal interests. The culmination of this Franco/papal relationship was reached on Christmas Day 800, when the Frankish king Charles (Charlemagne) was crowned emperor by Pope Leo II I in Rome. Apart from Carolingian power, the church was the only institution that transcended the ethnic divisions of the empire. As such the link with the church was to be of great significance in Carolingian attempts at "ideological incorporation." With its heart in Rome, and as the inheritor of many of the traditions of Roman antiquity, the church provided the Carolingians with a connection to that f(ml1er great imperial system in Europe. In addition, the church provided models of unity and leadership with its belief in the commll1titas Jidelitttrl (the community of the f~lithful), which transcended the sense of belonging provided by thegcns, and with its understanding of the proper ordering of society. Finally, through its presence on the ground, its organizational structure, and its control of the written word, the church provided the Carolingians with logistical support in the dissemination of their message. Wc sec aspects of this message proclaimed in the products of what has been called the "Carolingian Renaissance." The term Carolingian Renaissance was probably first used in 1838 (McKitterick 1977: xvi), and since then there has been debate about its relationship with the Merovingian past (Nelson 1977). Some sce it as a radical departure, mirroring the rupture created by the Carolingian seizure of power (Ullll1ann 1969: 27-8; G. Brown 1994: 4); others root it firmly in the Merovingian era, considering it the culmination of centuries of cultural development (R. Sullivan 1989; Wood 1994: 322-4). Yet perhaps these polar extremes miss the point. In common with many societies, the Carolingians engaged with the past chiefly in the interests of their present (sce Alcock, this volume; Smyth 1998). This is seen most graphically in their political rewriting of Merovingian history: in their construction of narratives that emphasized their significance in world history, and in the recounting of fictional genealogies that linked them not only to the Mel'Ovingians and Romans, but to the Trojans as well (Fouracre 1984; Asher 1993; Innes and McKitterick 1994; McKitterick 1995b; Wasno 1997: 44-50; E. Brown 1998). Just as they constructed an appropriate history f1'om li'agments existing in their present, so in terms of'material cultlll'e - it can only be expected that their reading of past texts and artif:lCts would result in the creation of' new l(ll'Il1s and concepts. As Chartier has argued: cultural or intellcctual "consumption" [must 1 itself' be taken as a t()['In of' production which, to be sure, manut:\cturcs no objcct [though it may contribute to such I, but which constitutes representations which are never identical to those that the producers (the authors or artists) have introduced into their work.

(1988: 40)

403

16.3 T7JC llatC/JOll.l'c (Torhalle) (It L01'SCh.

Just as the translation of the "relics of a Roman saint could acquire a wholly new meaning in a new cultural and historical setting in Carolingian Francia" (Innes and McKitterick 1994: 216), so the Carolingian appropriation (or consumption) of the material cultlll'e of the Roman and Merovingian past resulted in the creation (or production) of /(>1'111$ which owed much to the past but which were fundamentally the product of the present. The Carolingian reading of the Arch of Cons tanti ne, which resulted in the creation of'the gatehouse to the monastery at I,orsch (Fig. 16.3), 1:11' fi'om being a direct copy Ij'om late antiquity (sce Krautheimer 1971 ), was an active creation of'the early middle ages in which the l(ll'Il1 of the Roman monument was evoked and rc-created. As we shall sce, the "feeling" of late Antiquity, particularly connections with the emperor Constantine, was of'special significance in the creation of Carolingian art and architectlll'e. In their appropriation and reworking of the material world of the Romans and Mel'Ovingians, the Carolingians were both rooting themselves in those pasts and creating a new world. As in the eastern Roman empire, they were "not so much concerned to preserve the past as to redeploy it even reconstruct it to negotiate present contingencies" (Alcock, this volume, pp. 343-4). Exercising an intellectual hegemony which transcended the limits of their political power, the Carolingians attracted some of the greatest thinkers in Europe to their COlll't (Ullmann 1969: 3; G. Brown 1994: 28-31; Nelson 1995: 423). Working with a common purpose in the palaces, cathedrals, and monasteries, men like Alcuin of' York and Theodulf of Orleans constructed Carolingian political philosophy, and patronized and produced art and architecture

404

k/orcln 11 ti

Romc rcborn?

405

(McKitterick 1983: 145). The concepts of wholly disinterested learning and of the neutrality of art would have been alien to Alcuin or Theodulf. In fact Theodulf was almost certainly the main author of the Libri Cl11"Olilli, a late-eighth-century treatise on the role of images (Freeman 1957, 1965), in which it was argued that they enhanced beauty and served a didactic purpose as "books for the illiterate" (Contreni 1995: 818). Theodulf argued that education and learning provided the "soldiers of Christ" with weapons which, "along with inward devotion would enable the clergy to penetrate the mysteries of Scripture and lead God's people to paradise" (cited in Contreni 1995: 756). Learning and scholarship were integrally connected with the "construction of society. "Two themes the rebirth of the Prankish nation as the people of God, and the rebirth of the Roman empire permeate the first and were to lead to the realization of the second.

PoputttS Dei Founded on the writings of early Christian thinkers, Carolingian scholars produced a fundamentally Christian political philosophy (Krauthcimer 1971: 231 ). The church was to unite the different ethnic, linguistic, and social groups of the empire into the jiO/llltllS Dci (Ullmann 1969: 21 ); the Franks were to replace the Jews as the chosen "people of God." The apse mosaic of Theodulf of Orleans' church at Germigny-des- Prt's portrays the Ark of the Covenant in the Temple of Solomon, the symbol of salvation promised to the Jews, surrounded by f()llr cherubim (Fig. 16.4). The two smaller cherubim represent the Jews; the larger ones the extension of that salvation to the Gentiles and, by implication, the Franks (Freeman 1957, 1965; Vicillard-TroYckourolr 1962). The theme of the mosaic in the dOl1le of the Aachen palace chapel was the Apocalypse: "with the Elders gathered about the throne of God in adoration, representing the chosen people of God, with whom the Franks identified" (Nees 1995: 837). The depiction of this image, in the room which contained the Frankish royal scat, relayed a message which would not have been lost on those who gathered around the throne of Charlemagne (sce also Illl1eS and McKitterick 1994: 215). The church and the state were to be essent ially onc ('/'. Noble 1976). The king had secular jlo/JIcr; the clergy had divine n II!/Joril)' (Krautheimer 1980: 115-16). The anointed kings were now linked directly with the Divine and "rebellion against their authorit y ... became rebellion against God" (G. Brown 1994: 45). The king's power came li'om God and no longer fi'om his people. As in Pippin's coronation, blood and tradition had been supplanted by divine legitimization. The earthly order was to rdlect that in heaven (G. Brown 1994: 24-5) a notion which would have naturalized and fixed the hierarchy of the fl>rIl1er.

Rornft renascens A revival of classical antiquity permeated all aspects of lill- at the Carolingian court, the monasteries, and the royal estates. Classical authors were read, COI11mented upon, and copied (Bullough 1977; I nnes and Mc Kitterick 1994). I t has

even been suggested that in its decoration the Utrecht Psalter (produced between 820 and 830) "resembles the stateliest surviving late Antique copies of the Aeneid" (G. Hendcrson 1994: 264). The Carolingian appropriation of Roman rhetoric in art and literature can be seen as an attempt to ground their new-f(HlIld imperial status in the general past glories of' Rome. Onc particular set of' late antique symbols - those connected with the emperors Constantine and Theodosius - were more specifically appropriated, however. These rulers, who first linked Christianity with authority over the Roman world, were the two "good" emperors portrayed in the fi'escoes at Ingelheim (Krautheimer 1971: 237). In 788 Pope l-Iadrian 1 acclaimed (:harlemagne as the new (:onstantine (Herrin 1987: 386); (:harlemagne 's image on his coinage of 805 echocs depictions of that same emperor (Mc Kitterick 1988: 31). The illscription accompanyillg an illustration of Charles the Bald likens him to "Josiah and Theodosius: /osiah the holy king, I'dl)rmer of Judah, Theodosius the imperiallawmakcr" (Porcher 1970: l43; Mc Kitterick 1988: 32; Nelson 1989; 1992: l7). The Carolingian palace at Aachen (Fig. 16.5) - Charlemagne's NOI'I1 ROIIltT shows parallel developments ill architecture (Krauthcimcr 1971: 236). Krautheimer lists connections bctween the palace in the north and Constantine's

16.4 '/l)e apse lIIosaic I1t GITlllilllly-nes-Prls.

406

The CI7l'olingiml cmpire: Rome I'c/701'n?

John JYIoreianrl

:.:::::::::::J:::::::.... 11

:: 11

::

;; 11

:1

"

50

:.... ······::f:~:::~~·:··· ....... 16.5 Sche/llatic pla1/ of the palace at Aachell, showing the royal c/)(/pcl to the sollth alld the palllet: to the 1I01'th.

meters

--N-·->--

construction at the Lateran in Rome: the Aachen palace was often called "the Lateran"; a bronze figure of a she-wolf stood in the vestibule of the Aachen chapel, matching the Roman !tt/m at the Lateran; an equestrian statue was installed at Aachen parallelling the Lateran statue of Marcus Aurelius. As Krautheimer concludes: "the palace at Aix [Aachen] attempted to emulate not just any Roman monument, but a monument of Christian antiquity in Rome" (1971: 235). The same can be said «)t' many of dle churches of the Carolingian empire modeled on churches specifically associated with Constantine: St. John's Lateran, St. Peter's and SI. Paul's in Rome (though sce Nces 1995: 824). Within a general consumption of classicism, the Carolingians appropriated particular symbols to produce specific messages: of Romanlless, of power on a grand scale, of Christian illljlcri1tlll. Thc Carolingian present was being grounded in the past of the Constantinian empire, providing existing power structures and social relations with a Christian imperial ancestry. The questions remain: How was the message disseminaledr \"'ho was intended to consume itr How did it relate to the reality of Carolingian illljlcrilllllr

']}ansrnittinq the dominant ideO/OilY '-

L.

A certain "court culture" can be postulated as onc means of instilling the desired ethos and disscminating it abroad. Like the other major courts of Europe, the Carolingian palaces were populated by sons of the nobility sent there to acquire the appropriate social and political skills (Nelson 1995: 4(4). While some of this "education" may have been 1I'Inalized in the palace school (Contreni 1995: 713), lessons could also be read from the /1'111 of the palace and churches and from the images with which they were decorated. The Arl1/umitio C;cIICl'I1tis of 789 decreed that schools were to be set up 1