Empires, strong and weak

The communities discussed so far were relatively small. In the case of tribes without rulers this was due to a combination of factors, including an economy based on hunting, gathering, fishing, cattle-herding, and slash-and-burn agriculture; the nomadic or semi-nomadic lifestyle that these activities made necessary; the wide open spaces that they entailed; and the weakness of government itself.

Though chiefdoms tended to grow larger than tribes without rulers, in their case the number of people who could be controlled from a single center was limited by the absence of an administration making use of the written word. As already noted, the resulting instability was reinforced by the prevalent system of polygyny which very often caused the rulers to beget numerous offspring. Unless great care was taken and proper arran­gements made in advance, each time a chief died the result would be a succession crisis which threw the people into a turmoil and provided sub-chiefs with an opportunity to break away.

As for city-states, they were small by definition. The citizens of each one regarded themselves as a separate people descended from a single race and worshipping the same gods;⁶² though Greek cities did acknow­ledge their common cultural identity, until the advent of the Hellenistic age (which, by subordinating most of them to larger political entities, reduced the differences among them), they were extremely reluctant to admit foreigners. Furthermore, and however small or large the number of those who possessed political rights, the essence of the city-state was a

⁶¹ Thucydides, The Peloponnesian War (London: Heinemann, Loeb Classical Library, 1921), II, 36.

⁶² For the Greek polis as a community of ancestor-worshippers, see above all N. D. Fustel de Coulanges, The Ancient City (Garden City, NY: Doubleday, 1956 edn.). For ar­cheological evidence on how the community of the few was probably transformed into that of the many, see I. Morris, Burial and Ancient Society: The Rise of the Greek City-State (London: Cambridge University Press, 1987). direct system of government in which all participated to one extent or another. Such a system required that nobody live too far away from the civic center where the assembly held its meetings and where the city's public buildings - such as temples, courts, theaters, etc. - also tended to cluster: say, the distance that could be conveniently covered by a man traveling on foot within a single day. Thus Athens, next to Syracuse the largest by far of all Greek city-states, probably had at its zenith a popula­tion of 250,000. Among them perhaps 30,000-40,000 were citizens; the balance was made up by their related and unrelated family members (the latter being slaves). All in all they lived in an area of no more than 600 square miles. Other city-states were much smaller. Often they counted their citizens in the low thousands and often only in the hundreds, as is illustrated by the fact that the Island of Crete alone was divided between no fewer than fifty different cities.

By contrast, empires - even the earliest ones - were often mighty organizations. Some were able to last for centuries and even millennia; this was particularly true of those which, like the ancient Egyptian and Chinese ones, were ethnically homogeneous and developed a political system that was all but identical with the culture in question. Homogene­ous or not, empires often covered hundreds of thousands, if not millions, of square miles and counted their subject peoples - to the extent that they could be counted at all - in the millions and tens of millions. For example, the Inca empire measured over 2,000 miles from north to south and may have had 6 to 8 million inhabitants.

The origins of some of the oldest empires, e.g., the Chinese and Egyptian ones, is unknown. Most of the remainder originated when one chiefdom conquered its neighbors - some of the most primitive of them, indeed, are perhaps best understood as chiefdoms grown large. This was the case with the Aztec and Inca empires, both of which owed their genesis to a series of exceptionally able warrior-chiefs: taking over from previous rulers, they went on to extend their rule in all directions, using the members of their own tribes to create a new ruling class. The Assyr­ian, Babylonian, Persian, Arab, Mongol, Ottoman, and Mogul empires also originated in the conquest by one chief and one tribe of many others.

By contrast, Rome grew directly out of a city-state and indeed as late as the beginning of the fourth century BC it was still possible for Constan­tine to claim that he had ‘‘avenged the res publica on the tyrant” (the pretender Maxentius).^[43] By the last decades of the second century BC, Rome, though still maintaining its old republican system of government, had expanded until it counted its citizens in the hundreds of thousands.^[44] With the advent of those two would-be social reformers, the brothers Gracchus, the system began to break down; once the so-called social war of 90-89 BC emancipated Rome's Italian allies and turned them into citizens it became entirely impractical.

At the head of each empire there stood - unsurprisingly - a single emperor. Early in the fourth century Diocletian attempted to divide Rome between two emperors, known as augusti, each with a designated successor to take his place when the time came; however, the attempt broke down as soon as he himself resigned the reins of government and does not seem to have found imitators since. In both Rome and China some emperors tried to regulate the succession by appointing their own sons - whether real or adopted - as co-regents during their own lifetime. In the Ottoman empire, to prevent them from becoming the center of intrigue, the emperors' often very numerous sons were brought up in a part of the palace known as ‘‘the cage'';^[45] once a new sultan gained power his first step was usually to order all his brothers strangled. Provided they were sufficiently prominent, female members of the imperial family could also play a role in the succession. Either they did so by engaging in intrigue in favor of their sons, or else by outliving one imperial husband and marrying his successor by way of an additional confirmation of his status; one tenth-century Byzantine princess even married three em­perors in succession.

Like the chiefs that many of them originally were, most emperors claimed to owe their position to some kind of divine connection. This was true even in Rome, possibly the most secular-minded empire of all; already Caesar took the position of pontifex maximus, whereas his succes­sor Augustus allowed temples to be built for him in the provinces, if not in Rome itself.

If Roman emperors took time before they developed into gods, else­where the connection was obvious from the beginning. Some were them­selves god incarnate: so with the ancient Egyptians and the Inca, both of whom worshipped the sun, which proved to be a popular choice for imperial ancestor; so also in China, where the emperor was the son of Tien or heaven. Elsewhere the various arrangements that existed invariably implied a strong measure of supernatural support. Thus Mesopotamian emperors, though apparently not claiming to be divine themselves, were often represented in art as confronting the deity face to face and receiving its commands. Arab khaliffs claimed to be descended from Mohammed. As ‘‘heads of the faithful,'' they used their positions to exercise not just secular but religious government as well; and the same was true of the Ottoman sultans. Even when the emperor was neither a god nor a descendant of the prophet, as in Byzantium, he acted as head of the church and, indeed, it would be hard to say which one of his functions, the secular or the religious, was the more important. In fact perhaps the only emperor who did not unite secular and religious power in his own person was the Western Christian one; but even here the first among them, i.e., Charlemagne, regarded himself as head of the church as well.

⁶⁶ Potter, Prophets and Emperors, pp. 122, 128-9.

⁶⁷ L. Halperin, Charlemagne and the Carolingian Empire (Amsterdam: North Holland, 1977), pp. 148-9.

Whatever their precise relationship with the gods, all emperors were absolute rulers who combined legislative, executive, and juridical func­tions in their own persons. There could be no question here of con­stitutional restraints of a separation of powers; as the Latin saying had it, salus principis lex est (law is what is good for the emperor) and princeps legibus solutus est (the emperor stands above the law). Similarly, Hellenis­tic rulers were nomos empsychos, law incarnate, ruling over both the bodies of men and their religious beliefs. To show what absolute power really meant Antiochus III on one occasion gave his own wife to his son, Antiochus IV, to marry. To the army, which had been assembled to witness the occasion, he explained that ‘‘this is not done according to the laws of god or man but because it is my will; and since they are both young, they will surely have children.”^[46]

Ideologically speaking, most empires developed doctrines whose pur­pose was to confirm the subject population in its obedience to the powers that be. Thus China had Confucianism in its two forms, ‘‘paternal” and ‘‘legal.”^[47] The former presented the empire as a vast family in which the young and the subordinate owed piety to their elders and betters, whereas the latter emphasized the role of discipline and prescribed drastic punish­ment for those who disturbed the heaven-mandated social structure. In the Arab, Ottoman, and Persian (from the middle of the seventh century on) empires, a similar role was played by Islam (the word itself means ‘‘submission’’) which in at least some of its versions emphasized fatalism, resignation, and obedience. Finally, the ancient philosophies such as cynicism, epicureanism, and stoicism all grew out of the ruin of the independent city-state and are best understood as reactions to despotism, whether Hellenistic or Roman. Thus the cynics taught that, to compen­sate for the loss of freedom, man should give up his possessions and withdraw from the world. Epicureanism suggested that men, likewise withdrawing into private life, should focus on enjoying themselves, whereas stoicism on the contrary put the emphasis on endurance, service to one’s fellow men, and - if things became too hard to bear - suicide as a way of escaping to a world in which even the emperor’s long arm could not reach.^[48] In time all these ideologies were overtaken by early Christianity, which, in the words of its founder, gave Caesar what was Caesar's due while enabling the believer to focus on saving his own soul.⁷¹

So long as emperors behaved themselves, their rule might be benefi­cent. However, there always existed the danger that, motivated by need or greed or sheer madness, they would cease to do so, in which case the results would be unfortunate for the members of their immediate entour­age in particular. Already in ancient Egypt we come across the story of an official who breathed a sigh of relief for not being punished after acciden­tally touching Pharaoh,⁷² and the biblical Esther, having approached Xerxes without first asking permission, was fortunate to escape with her life. In China officials often wore heavy padding to cope with the floggings to which they were likely to be subjected and which could incapacitate a person for weeks on end; from the Middle East and pre-Columbian Latin America we have accounts of the spectacular punishments often inflicted by emperors on subordinates who displeased them. In Rome, according to the historian Seutonius, fear of the emperor sometimes drove people to commit suicide in self-defense and to leave everything they had to him.⁷³ In brief, an emperor could do anything to any of his subjects, whereas conversely any cruelties that he did not choose to visit on them counted as pure indulgentia on his part.⁷⁴

As they appeared to their subjects as awesome, indeed almost divine creatures, another consequence of the emperors' position was their claim to be universal rulers. The modern state regards itself as one sovereign entity among others, but empires by definition could not accept equals. Looking beyond their borders they saw not other political communities with a right to an independent existence, but barbarians who at worst caused trouble and at best were not worth conquering. Already in Me­sopotamia the earliest Accadian emperors claimed to reign over the ‘‘four quarters of heaven,'' a tradition later followed by their Assyrian and Babylonian successors right down to the Persian ‘‘kings of kings.'' The emperor of China carried the title of ruler of ‘‘all under heaven,'' whereas Rome identified itself with the oikoumene, Greek for ‘‘the inhabited world,'' which was said to stretch from the British ocean to the Tigris.⁷⁵ This notion was later taken up by Charlemagne who, like his successors, carried the orb to symbolize his position. All of them claimed to be the

⁷¹ A good account of early Christian political attitudes is A. Cunningham, The Early Church and the State (Philadelphia: Fortress Press, 1982).

⁷² K. Sethe, Urkunden des agyptischen Altertums (Leipzig: Heinrichs, 1921), vol. IV, pp. 608-10.

⁷³ Suetonius, Lives of the Emperors (London: Heinemann, Loeb Classical Library, 1965), Tiberius, 49, 1; Caligula, 38. See also P. Plass, The Game of Death in Ancient Rome: Arena Sport and Political Suicide (Madison: University of Wisconsin Press, 1995), pp. 89ff.

⁷⁴ Pliny, Jr., Letters and Panegyric (London: Heinemann, Loeb Classical Library, 1975); H. Cotton, ‘‘The Concept of Indulgentia Under Trajan,” Chiron, 14, 1984, pp. 245-66.

⁷⁵ E.g. Dessau, Inscriptiones, vol. I, p. 168, no. 754. rightful rulers of the universe as a whole,^[49] as did the emperors of pre-Columbian Mexico and Peru.^[50]

To the extent that these and similar claims did not correspond to reality they could sometimes lead to comic results. Thus Suleiman the Mag­nificent once condescendingly wrote to Francis I of France that ‘‘your pleas for aid [against Charles V] have been heard at the steps of our throne”; refusing to accept Queen Elizabeth of England as his equal, Ivan IV the Terrible of Russia spoke of her as ‘‘a mere maiden.” As late as the first decades of the eighteenth century, diplomats from European states seeking an audience with the Porte had to keep up the pretense of addressing a superior by donning Turkish dress over their own clothes. In the Far East, whenever a Japanese delegation visited the Chinese capital of the day they would demonstrate their independence by using language not appropriate for subjects, which offense the Chinese, provided of course they wanted to maintain correct relations, would magnanimously ‘‘forgive’’ by ascribing it to their guests' supposed ignorance of the correct forms.

Religion apart, the two pillars supporting imperial rule were the army on the one hand and the bureaucracy on the other. Few empires seem to have gone as far as Rome where the title ‘‘Imperator,’’ or victorious commander, invariably headed the list of imperial offices, but the connec­tion between political domination and military might was always clear. Compared to the polities previously discussed, the armed forces raised by empires were enormous. Nor did these forces consist simply of bands of warriors or personal retainers or popular militias. Instead, they were regular troops who made soldiering into their chosen career, served for long periods, were commanded by professional officers, and were cen­trally paid from the imperial treasury. Their numbers could grow into the hundreds of thousands - but here it is important to note that the ex­tremely low degree of social and economic development of most empires (only a few ever carried their existence into the period of the industrial revolution, and those that did, such as the Ottoman and Chinese, expired within decades of encountering more advanced civilizations) prevented them from putting more than 1 or 2 percent of their populations under arms. Thus Rome, by no means the least developed among them, at its zenith counted only some 300,000 troops - and when that figure was doubled under the late empire its economy broke down under the strain. Elsewhere the figures were probably even lower.

Furthermore, the cost of standing armies was such that most empires kept only a comparatively small number of regular troops; even in Rome, which probably went further in this regard, the first thing Vespasian, for example, did when setting out to subdue Judea in 66 BC was to gather auxiliaries.⁷⁸ Elsewhere the Persian Immortals, the Chinese Imperial Guard, and in the Ottoman empire the Janissaries hardly exceeded the (usually) lower tens of thousands. These units would serve the triple function of standing army, garrison for the capital, and police force responsible for dealing with internal uprisings. Meanwhile the bulk of the army did not consist of regulars but acted under some kind of feudal arrangement. In return for being granted land and exempted from tax­ation they would make themselves available on a temporary basis, as was the case of the Ottoman Sipahis. Alternatively they consisted of short­term conscripts who were normally available only for the defense of their home provinces, as was the practice in certain periods of Chinese history and among the Inca. As to navies, they were a much more expensive proposition still. Few empires in history were able to construct fleets of warships and maintain them for very long. The normal method was to rely on maritime cities which in time of war provided both the ships, suitably modified, and the crews.

Sharply differentiated from the army, the civilian bureaucracy consis­ted of literati. Originally they were probably priests who had mastered the sacred art of writing - such was the case in Egypt, where the script that they used is still known as hieratic (priestly), Mesopotamia (the Accadian and Babylonian empires), and among the Aztecs and the Inca. Later they came to be selected from the upper social classes. Once they had a foot on the ladder, promotion often proceeded on a more or less regular system, tempered, needless to say, by family connections and also imperial ira et studio; only in China since the days of the T'ang dynasty did there exist a system of examinations which was rigidly enforced as a vehicle of selec- tion.⁷⁹ In theory everybody (except certain classes of convicted criminals) could present himself. In practice the time and expense needed to prepare for the examinations were such that only the sons of officials or of rich merchants could apply. Still, the system did prevent the creation of a hereditary aristocracy, as indeed it was designed to do.

Insofar as a collusion between bureaucrats and military commanders might prove fatal to the regime, it was the first concern of every emperor to keep the two apart. This achieved, the number of imperially appointed administrators was often surprisingly small. Both in Rome and in Ming China the bureaucratic positions directly under the emperor's control

⁷⁸ Josephus Flavius, The Jewish War (London: Heinemann, Loeb Classical Library, 1965), III, I, 3.

⁷⁹ See I. Miyazaki, China's Examination Hell (New York: Weatherhill, 1976), for the details. probably numbered fewer than 10,000 - and in the latter case they ruled over a population estimated at 150 million.^[51] It is true that each Roman official and each mandarin had his assistants in the form of ambitious family members, retainers, and slaves (or often freedmen) who were fed from their tables, served on their staffs, and could be made to carry out routine duties, go on errands, and the like. Even so, however, their overall numbers remained limited and certainly bore no relationship to modern administrative systems.

Just as the emperor himself combined various functions in his own person, imperial officials were administrators and judges rolled into one; though Roman senators during the early empire still had the right to be judged by members of their own class, the idea that the two ought to be separate was essentially a modern European development that does not date further back than the seventeenth or even eighteenth century.

These tasks apart, the most important responsibility of any official was to gather taxes on the emperor's behalf. Everywhere the two most impor­tant taxes were the land tax, consisting of a share of the crop, and a poll tax. In some of the more primitive empires, e.g., the Aztec,^[52] Inca,^[53] and early Chinese ones,^[54] these had to be paid in kind; elsewhere they were usually commuted into money at rates that were set, needless to say, by the authorities themselves. To these sources of revenue would be added the imperial monopolies. From pre-Columbian America through Rome to China, the latter often consisted of the most valuable products in­cluding salt (a certain quantity of which had to be purchased by each family each year), metalware, precious stones, certain kinds of skins, furs, and feathers, and, in the case of Rome, the famous balsam that was grown on the shores of the Dead Sea. In most places mines, forests, rivers, and lakes were considered imperial property; in some empires the same applied to certain species of animals considered uniquely precious either because of the price that their products commanded or simply because they were large and hunting them was correspondingly prestigious. All these different kinds of resources would be exploited either directly by the emperor's own appointees such as office-holders, retainers, and slaves, or, which was perhaps more likely, leased to the highest bidder and exploited in return for a share in the profits.

A tribute-list drawn up by the Assyrian king Tiglat Pileser III (reigned 745-727 BC) included ‘‘gold, silver, tin, iron, elephant hides, ivory, multicolored garments, linen garments, blue-purple and red-purple wood, maplewood, boxwood, all kinds of precious things... flying birds of the sky whose wings are dyed blue-purple, horses, mules, cattle and sheep, camels, and she-camels together with their young.”⁸⁴ With these and other sources of revenue at their disposal, many emperors were able to build spectacular fortunes; Augustus, for example, declared the whole of Egypt an imperial estate and closed it to members of the governing, that is, senatorial class. The forbidden city in Beijing, like Nero's domus aurea, contained untold treasures; those accumulated by the Aztec and Inca rulers have become legendary. When Alexander the Great entered the Persian capital in 330 BC he found Darius' treasury filled with 50,000 gold talents (whose late twentieth-century value would be $1,800,000,000-3,000,000,000) - this, on top of uncounted quantities of silver and other precious objects that had been hoarded over a period of perhaps two and a half centuries.

In the absence of the abstract state, it should be noted about all these riches that they belonged to, or at any rate were at the disposal of, the emperor in person. As in all pre-state polities except the city-state, he had to be the richest person by far in his domain; any competitors in this respect were ipso facto dangerous and had to be eliminated. In Byzan­tium, China, and elsewhere, attempts were sometimes made to keep separate receiving chambers for the two kinds of revenue, that originating in taxation and that deriving from the emperor's ‘‘private'' property; likewise, two chambers were maintained to separate the expense of the palaces from those of the army and the administration. In practice, though, these distinctions could seldom be maintained. Thus Roman and Chinese emperors regularly forgave the payment of taxes by regions that had been struck by natural disasters; the praise for generosity that they earned as a result would have been wholly out of place if the sums in question had not been perceived as earmarked for their private pockets. Conversely, when in need of money for war or other purposes emperors very often resorted to their ‘‘private'' resources. They sold plate, pawned palaces and estates, and even married their offspring to the highest bidder; when Charles V stood in need of money to fight the Protestants in Germany, he used the dowry brought in by his son Philip as a result of the latter's marriage to a Portuguese princess.⁸⁵ The prevailing system was curtly summed up by the Roman historian Tacitus when, on one occasion during the reign of Tiberius, the Senate voted to transfer money

⁸⁴ H. Tadmor, The Inscriptions of Tiglat Pileser III, King of Assyria (Jerusalem: Israel Academy, 1994), pp. 69-70.

⁸⁵ K. Brandi, The Emperor Charles V (London: Cape, 1967 edn.), p. 495. from the old Republican aerarium to the imperial fiscus - ‘‘as if it mat­tered.”⁸⁶

What was true of the imperial possessions - one is at a loss to find a more accurate term - also applied to the army and bureaucracy. Both consisted of the emperor's men and served him rather than the state; some did so in his private capacity, others in his public one. In practice the distinction between the two was likely to break down. To add a certain luster to the court, some of the empire's highest dignitaries were often used to look after the emperor's personal needs such as bearing his cup, running his wardrobe, looking after the stables, and the like. Conversely palace servants, whether free or servile in the sense that they could be bought and sold, were often used on ‘‘public'' tasks including command of the imperial bodyguard and filling vitally important administrative posts such as that of secretary to the emperor. Sometimes the personnel in question originated in the imperial harem (which itself meant not just the place where the emperor's women were kept but also included such institutions as the mint and the armory) and consisted of eunuchs. This was the case in the Persian, Byzantine, Arab, Ottoman, and Chinese empires, among others.

As the campaigns of the Byzantine general Narses and the Chinese admiral Zheng He - both of them eunuchs - demonstrate, government by household was not necessarily less competent or more corrupt than that which was exercised by the regular administration. It did, however, tend to circumvent the elite of high-born literati who saw their positions usurped, and their access to the emperor controlled, by persons whom they held in a mixture of fear and contempt. Reduced to political im­potence, they would vent their spleen in writing. This probably explains the bad name that government by household has frequently acquired both at the time and in the eyes of many subsequent historians.⁸⁷

Whether their personnel was free or not, both the members of the administration and those of the army owed their allegiance to the em­peror as a matter of course. The oaths that they swore were addressed to him personally and had to be renewed each time a successor ascended to the throne; in return, they could normally expect to receive some kind of largesse. To ensure the continuing loyalty of senior administrators and officers, many emperors engaged in the practice of distributing presents to them at regular intervals, the value of each gift being carefully graded

⁸⁶ Tacitus, Annales (London: Heinemann, Loeb Classical Library, 1937), vi, 2. The same was true in Byzantium: W. Ennslin, ‘‘The Emperor and the Imperial Administration,” in N. H. Baines and H. L. St. Moss, eds., Byzantium: An Introduction to East Roman Civilization (Oxford: Clarendon Press, 1948), p. 283.

⁸⁷ For an example of a mandarin inveighing against government by eunuchs, see de Bary, ‘‘Chinese Despotism and the Confucian Ideal,” pp. 176-7. according to the recipient’s rank so as not to offend the others. The confusion between the private and the public was further accentuated by the fact that many members of the propertied classes, merchants in particular, were also in charge of collecting taxes - especially those on commodities - and were held personally liable for delivering them to the imperial treasury. If the latter faced a deficit, both merchants and officials could be called upon to make forced loans. Thus imperial service and private enterprise became fused. The abstract state being absent, the entire structure amounted to little more than a gigantic racket, one in which the emperor combined with his servitors, whatever their precise status, to fleece the remainder of the population.

The emperor’s absolute power on the one hand and the absence of any clear distinction between the private and the public on the other meant that the one institution more or less safe from arbitrary interference was the established religion or church. Often it had a system of taxation running parallel to that of the emperor himself, as among the Inca; in other places it owned extensive estates second only to his, as in early medieval Europe. Even where the emperor also acted as the head of religion, as was the case in most empires, these factors gave it a certain autonomy. To be sure, the resources made possible by this autonomy often roused the cupidity of emperors who tried to appropriate them by various means. On the other hand, the fact that they owed their position to religion usually dictated a certain circumspection in dealing with it. To clash with the church openly might be to court disaster; many emperors who did so came to a bad end as did the Egyptian Tutankamun (Achenaton) after trying to substitute new gods for the old. The Secleucid Antiochus III was killed in 187 BC after robbing a temple of Ba’al; apparently one reason behind the easy Spanish success in Peru was the fact that, just before Pizarro and his men arrived, the Inca Atahualpa had quarreled with the priesthood in his attempts to reduce the expense of the imperial mummy-worship of which they were the principal benefactors.^[55] Conversely, the relative security offered by the temple often turned it into a place where ordinary people flocked to deposit their belongings. This enabled it to engage in banking operations and turn itself into a center of commerce; the money-changers whom Jesus drove from the temple in Jerusalem would have been a common sight.^[56]

Another factor capable of putting limits on the emperor’s power was associated with time and distance. Given the extreme centralization of their political systems, once the armed forces of empires had been defeat­ed in battle, conquering vast spaces was sometimes relatively easy and could be accomplished in a fairly short time; as both Alexander’s con­quests and those of the Mongols show, however, administering them was quite another matter and, indeed, the less homogeneous, ethnically speaking, any empire, the greater the difficulty. To overcome these han­dicaps the more long-lived empires left their physical mark upon the landscape in the form of massive ‘‘public,’’ for which read imperially instigated (and sometimes paid for) works. The Chinese and Romans are famous for the fortifications with which they surrounded their borders. The Persians, Romans, and Inca all excelled as builders of roads, aque­ducts, and bridges. Neither pre-Columbian Mexico, nor the Egyptian empire, nor the various Mesopotamian ones would have been conceiv­able but for the systems of canals that they built and which served for irrigation and transportation; as Tiglat Pileser boasted, ‘‘I dug out the Patti Canal... and made it bubble with abundant water.’’^[57] Less durable, but equally important for holding the empire together, were the messenger systems that connected the provinces with the capital and which, in the case of Rome, were capable of distributing the imperial edicts to the remotest provinces in a matter of one to four months,^[58] as well as carrying out the censuses held from time to time.

In Hellenistic Egypt, according to one surviving document, so tight was the system of imperial control made possible by these means that even a prostitute wishing to ply her trade for a single day in a given city had to apply for a license and, presumably, pay for it.^[59] Byzantine emperors made repeated attempts to regulate the economy by prescribing every­thing from the opening hours of shops to the prices that could be charged for various commodities.^[60] The Inca even developed a special system of recording - consisting of colored knots or quippu - for tax purposes. According to Garcilaso de la Vega, himself the son of a Spanish conquista­dor and an Inca princess and thus familiar with the culture, so thorough was the imperial information system that it proved capable of recording every last pound of maize and every last pair of sandals produced throughout the empire.^[61]

On the other hand, there exist indications that the imperial records were often full of holes and that the results of the attempts to gain information were mediocre at best. For example, when Xerxes set out to conquer Greece in 490 BC he was entirely unaware of the existence of the second wealthiest person in the realm - so wealthy, in fact, that he could support the imperial army (according to one source, a million and a half strong) from his own private resources as they passed through his es­tates.⁹⁵ Though Hellenistic Egypt as the heir to millennia of imperial rule was probably among the most tightly governed of all empires, Ptolemy IV Philopater on one occasion found himself unable to determine whether in fact he had granted the city of Soli certain privileges (immunity from having troops quartered on them) as its citizens claimed.⁹⁶ Concerning both Rome and Ming China it has even been argued that the absence of good maps, good ‘‘data bases,” and good communications reduced their emperors to passivity so that they dealt only with those cases that were brought in front of them or else were confined largely to ritual func­tions.⁹⁷ To be sure, it is hard to imagine how a person as enterprising as Septimus Severus, who had started his career as a junior officer, could have given up his active lifestyle after ascending the imperial throne in AD 193. On the other hand it is clear that no emperor could know everything and that these limitations often represented an important constraint on their ability to rule.

In fact, given the problems of time and distance as well as the limits of the information at their disposal, many emperors preferred to deal with entire communities - tribes, chiefdoms, villages, cities, even client-kings - rather than with individuals. In both Tenochtitlan and Quizco the Aztecs and the Inca used as their basic social building block not families, but wards known as calpullin and alyu respectively;⁹⁸ outside the capitals, too, their rule was indirect, being exercised by means of subordinate tribal chiefs whom they had conquered. China originally had the so-called eight-well system which counted peasant families by ‘‘wells,’’⁹⁹ whereas Rome for a long time left administration to hundreds upon hundreds of autonomous city-states as well as numerous client-kings. Instead of coun­ting their entire populations, empires preferred to calculate censuses by households or hearths. Both of these institutions were likely to contain unrelated as well as related ‘‘family’’ members, with the result that

⁹⁵ Herodotus, The Histories (London: Heinemann, Loeb Classical Library, 1943), 7, 27-9.

⁹⁶ C. B. Welles, Royal Correspondence of the Hellenistic Period: A Study of Greek Epigraphy (New Haven, CT: Yale University Press, 1934), no. 30, p. 136.

⁹⁷ F. Millar, The Emperor in the Roman World, 31 BC-AD 337 (Ithaca, NY: Cornell University Press, 1977); Millar, ‘‘Emperors, Frontiers and Foreign Relations, 31 BC to AD 378,’’ Britannia, 13, 1982, pp. 1-23; R. Huang, 1587, aYear ofNo Significance (New Haven, CT: Yale University Press, 1981).

⁹⁸ Conrad and Demarest, Religion and Empire, pp. 52, 97.

⁹⁹ Ch’ienMu, Traditional Government in Imperial China: A Critical Analysis, Chun-tu Hsueh and G. T. Totten, trs. (Hong Kong: Chinese University Press, 1982), pp. 23-6. determining the correct multiplier must have proved as difficult for contemporary administrators as it has for subsequent historians.

The deficiencies in their administrative systems may also explain why many empires were reluctant to impose a single system of law upon all their inhabitants. Originating in conquest, many of them were heterogen­eous by definition; hence, so long as subjects followed orders and paid their taxes, most emperors were content to leave them alone. Apart from the fact that many, possibly most, of those involved in running the empire never appeared on the imperial payrolls, the system meant that, when officials came to perform their duties, they very often saw themselves confronted not by individuals but by organized communities of the most diverse kinds. This probably reduced their power and in many cases must have reduced government to a bargaining process.

Another factor which tended to work in the same direction and impose practical limits on the emperor's power was of a financial nature. No empire in history ever seems to have reached the point where it developed a single treasury (whether ‘‘public'' or ‘‘private'') into which all revenues went and which in turn was responsible for making all payments, the most important reason being that, given the expense of transporting bullion and the attendant risk, a large part of the sums collected by taxation always remained in the provinces and was used to meet local expenses. The same was even more true of deliveries in kind, most of which were likely to stay in provincial storehouses; to say nothing of the corvees or forced labor that the subject population was often made to perform in working the imperial estates, erecting and maintaining imperial works, and providing transport and other services to imperial officials. Each of these factors meant that only a small fraction of the sums raised, or the labor expended, ever reached the capital and was put at the free disposal of the emperor. The rest remained at or near the place of origin and could profit him, if at all, only indirectly and to the extent that he controlled the local officials.

To prevent local officials from going their own way and appropriating imperial resources for themselves, emperors employed a variety of devi­ces. Officials could be hindered from forming local ties by being rotated between one post and province and another. In Hellenistic Egypt there seems to have existed a dual administration, i.e., one set of bureaucrats charged with collecting taxes and another with supervising them;¹⁰⁰ elsewhere there were often roving inspectors such as the Roman quaestores (later replaced by the agentes rerum) and the Chinese visiting secretaries who appeared under a variety of titles but whose function was always the

100

M. Rostovstzeff, The Social and Economic History of the Hellenistic World (Oxford: Clarendon Press, 1941), vol. I, p. 251. same.¹⁰¹ The kings of Assyria at its zenith formed the habit of appointing eunuchs - ‘‘the son of nobody” - to govern newly conquered towns,¹⁰² a method which presumably had the additional benefit of humiliating the conquered.

Though each of these methods provided a practical answer to the problem, none was capable of solving it once and for all. Inspectors, whether traveling or permanent, could be misled or bought by the men on the spot. Transferring officials from one post to another merely meant that they did not have time to familiarize themselves with their districts and thus tended to increase the power of the heads of local communities at the expense of the center. The use of eunuchs obviously made it impossible for families to consolidate their hold on power but, for that very reason, probably endangered stability in other ways. In brief, an emperor might make himself absolute within his own capital. However, by and large the further away any particular province the harder it was to impose the imperial will.

Under such circumstances the danger always existed that subordinate authorities, whether religious or secular, would use any difficulty ex­perienced by the center - often a war or a succession crisis - in order to cease obeying the emperor and to break away. In particular, provincial governors might use the economic resources at their disposal to build up armed forces, whereas military commanders might use their troops to appropriate an economic foundation. The task of both was often made easier by the self-governing communities that most empires contained. Unless they were physically removed from their original abodes and sent into exile, which was a regular practice among both the Assyrians and the Inca,¹⁰³ subject tribes often provided feudalism with ready-made buil­ding blocks. The ultimate outcome would be the disintegration of the empire and its replacement by a much more decentralized, if still strongly hierarchical, system of government;¹⁰⁴ and indeed feudalism itself could be regarded simply as the political structure that emerged when empires fell on evil days. This was the case in Western Europe during the Middle Ages and in Egypt and China during their various interdynastic periods,

¹⁰¹ On the duties of second-century AD provincial inspectors, see the account of a contemporary Ts’ai Chin, printed in Wang Yu-ch’uan, ‘‘An Outline of the Central Government ofthe Former Han Dynasty,” Harvard Journal of Asiatic Studies, 12, 1949, pp. 159-60.

¹⁰² Tadmor, The Inscriptions of Tiglat Pileser III.

¹⁰³ For the Assyrians, see the biblical book of Kings; for the Inca, seeM. A. Malpass, Daily Life in the Inca Empire (Westport, CT: Greenwood Press, 1996), p. 72.

¹⁰⁴ M. Weber, Economy and Society: An Outline of Interpretative Sociology, G. Roth and C. Wittich, eds. (New York: Bedminster, 1968), vol. III, ch. 12; and O. Hintze, Staat und Verfassung: gesammelte Abahandlungen zur algemeinen Verfassungsgeschichte (Gottingen: Vanderhoek, 1961), vol. I, pp. 84-119. as well as in a large number of other societies including Persia, Byzan­tium, India, and Japan at various points in their histories.¹⁰⁵

Once the class of warrior-governors so characteristic of feudalism had come into being, it would invest heavily in military retinues as well as physical defenses that would enable them to defy the emperor. As each lord strove to make his domains as independent as possible the cen­tralized information-gathering, transportation, and defense system would be allowed to decay and disintegrate. Bureaucracies - and, to a large extent, the necessary literacy among the non-religious classes - postal services, censuses, even the most elementary means of transport, disap­peared; at no time were European roads as bad, and communications as difficult, as during the Middle Ages with its multitude of interlocking principalities. The regular armed forces too would melt away, perhaps to the point where they were reduced to a mere handful of retainers who were fed from the emperor's kitchen as he moved from one residence to another. Rights that had previously belonged to the emperor, such as the usufruct of economic resources (mines, forests, etc.), taxation, and coin­age would become dispersed and pass into the hands of numerous lords and barons.

Reflecting the emergence of feudalism, the imperial ideology collapsed. Its place was taken by a system that put much greater emphasis on the collective rights of the aristocracy on the one hand and the religious establishment on the other. According to Thomas Aquinas, probably the greatest and certainly the most systematic among medieval ‘‘political scientists,” government, far from being created by and for men, was an integral part of the divine order. As such it presented a seamless web in which each person and each class had his appointed place, which was not subject to arbitrary interference from above.¹⁰⁶ The stronger any em­peror the more any privileges that any person, body, or class possessed were regarded as within his gift and revocable at his will; but in feudal societies privileges became attached to their owners who in this way acquired certain ‘‘constitutional'' guarantees that were almost entirely absent from empires.¹⁰⁷

Once the warrior-governors ceased to be appointed by the emperor and succeeded in making their positions hereditary, the process would have reached its logical conclusion and the empire come to a de facto end, however alive it might still be in name. The resulting structure was based

¹⁰⁵ Forthehistoryoffeudalisminthesecountries,seeR.Coulborn,ed.₃ Feudalism in History (Hamden, CT: Archon Books, 1965).

¹⁰⁶ See D. Bigongiari, ed., The Political Ideas of St. Thomas Aquinas (New York: Hafner, 1981 edn.), particularly the introduction.

¹⁰⁷ The best modern accounts are M. Bloch, Feudal Society (London: Routledge, 1961), and F. L. Ganshof, Feudalism (New York: Harper, 1961). neither on family ties (though such ties were important in helping people form alliances) nor on bureaucratic fiat. Instead it rested on a network of fealty that bound each member of the aristocracy both to his superiors and to his inferiors. By taking an oath, vassals would commend themsel­ves to their lords and agree to serve them by offering counsel, bearing arms, and extending financial aid when necessary. In return they received protection, land for maintaining themselves and as many of their own vassals as were suitable to their station, and the rights that went with such land in the form of rent and various forms of labor services formerly due to the emperor and his representatives. Further down the social ladder, those who actually lived on the land and worked it tended to become glebi adscripti, tied to the soil. They were ruled over by the lord of the manor with hardly any interference on the emperor's part.

As the numerous transitions from centralized empire to decentralized feudal regime and back indicate, the two systems were not as far apart as would seem to be the case at first sight. Unlike tribes without rulers and city-states, empires whether strong or weak possessed a pronounced hierarchical character and a well-defined head even if, following feudaliz- ation, he was merely titular. Unlike the situation in chiefdoms these hierarchies were not simply based on ethnic identity and family ties but made use of relatively well-developed bureaucracies and standing armies on the one hand and of fealty on the other. Whereas emperors concen­trated all power in their own hands to the extent that practical consider­ations permitted, feudalism ensued when some sort of crisis caused the imperial prerogatives to pass into those of his soldiers and administrators, who became fused into a single hereditary class. The relationship be­tween the public and the private was the same, however, and differed notably both from the situation in city-states and from the one which obtains in the modern state. In the next section, the nature of that relationship will be explored more closely.