ALL THE KING’S HORSEMEN: WHAT AFTER ALL ARE FEUDALISM, SEMI-FEUDALISM, AND FEUDALIZATION?

Azar Gat

The nature, scope, and applicability of the concept of feudalism are all notoriously elusive. Historians have traced, and debated, the development of feudalism in specific historical circumstances, predominantly medieval Europe. But by professional inclination, historians of this region and period have only a cursory interest in feudalism elsewhere in time and place, if they do not regard the generalized framing of the question as wholly suspect. Indeed, more recently, the claim that major features traditionally associated with European feudalism crystallized and formalized later than earlier believed has had considerable impact on scholarly attitude towards the concept of feudalism even with respect to the European case itself.1 ‘Nominalists’ and ‘splitters’ have found support for their long-standing view that the concept obscured more than it clarified.2 In this mood it has almost become the vogue to hold that the concept of feudalism was untenable, indeed, that there never was ‘feudalism’ in reality. Whether or not they subscribed to such an extreme statement, just to be on the safe side, many scholars nowadays simply tend to avoid using the concept.

Acknowledging that every general interpretative concept functions as shorthand that encompasses diversity and flux, this article claims that ‘feudalism’ – no less than ‘capitalism’, ‘industrial society’, or ‘liberal-democracy’ – is a historically meaningful concept not only with respect to medieval Europe but as a much wider phenomenon. Herein lies that other unresolved ambiguity concerning feudalism: assuming that there was such a thing, was it a one-time European idiosyncrasy or was it a broader social category, also identifiable in other societies? If the former is true, why was feudalism unique to medieval Europe? If it was not, what was the feudal phenomenon that we ought to seek? Giants of social theory have differed on this since the eighteenth century. Among historians, Bloch, holding a centrist position, has expressed the lingering question marks as follows in his most theoretical chapter, ‘Feudalism as a Type of Society’: ‘Feudalism was not “an event which happened once in the world”. Like Europe – though with inevitable and deep-seated differences – Japan went through this phase. Have other societies also passed through it? And if so, what were the causes, and were they perhaps common to all such societies? It is for future works to provide the answers.’3

Classical ‘definitions’ of feudalism have mainly confined themselves to listings of the feudal traits - social, political, economic, juristic, technological, and military.4 Of these traits, mostly abstracted from the European model, some scholars have regarded the feudal-vassal personal relationship as feudalism’s main feature; others believed that it was the fief and the seigniorial-manorial system; others underlined the devolution of power and authority; others the military aspect; yet others insisted on all these elements.5 However, a deeper, generalized understanding of the conditions that brought feudalism into being, and of how feudalism stood in relation to other historical social and military regimes, has rarely been attempted. In the absence of such an understanding, it remains unclear how the various listed traits

2

related to one another and how fundamental one trait or the other was for judging whether or not a particular society outside north-western medieval Europe should be regarded as feudal. It is such a deeper explanation that this article ventures to propose. Large States, Natural Economy, Horsemen, and Feudalism We begin by noting that feudalism was invariably a product of state structure, and, provided that it did not lead to the state’s complete disintegration, it remained a form of state structure, albeit a segmentary one. Feudalism is not to be conflated with non-state, wholly localized, kin-based, chiefdoms, as some leading anthropologists of social evolution have suggested.6 Quite to the contrary, feudalism characteristically evolved in large states.7 To understand feudalism, then, we must first address the most cardinal question of politics in such states: How were they to be governed and financed? By what methods were they to extract resources and raise armed forces?

State rulers were normally themselves by far the largest property holders in the form of royal domain: vast estates which they owned, managed, and from which they extracted revenues, directly. As Max Weber has shown in his discussion of the ‘patrimonial state’, to maximize central authority state rulers would have ideally treated their whole realm in much the same manner, directly administering taxation and conscript labour.8 Armed forces would similarly be centrally and directly either conscripted or paid for from the revenues of taxation. To accomplish such centralization of power, however, two preconditions would have had to be met. First and foremost, advanced economic, transportation, and bureaucratic infrastructures would have had to be in place. Revenues in kind and money would have had to be assessed, collected, transported to the centers of power, stored, and reallocated, all managed by paid (or rationed) bureaucrats. Conscript and professional manpower – for both civil and military purposes - would have had to be similarly administered. The second precondition, related to the first, is that regional power holders would have had to be curbed.

These formidable preconditions for bureaucratic centralized states rarely materialized in the pre-modern world. Larger states typically emerged as overlordships, as one local ruler gained the ascendancy over the other. They variably continued to rely on the regional aristocratic power holders to govern the realm, both because the central authority lacked a developed bureaucratic apparatus and because the local aristocracy was powerful enough to maintain its social and political position vis-à-vis the center. Indeed, at the other extreme from central bureaucratization, power delegation to, and appropriation by, the regional leadership could result in power fragmentation and even in the virtual breakdown of central authority. Regional forces or provincial state officials occasionally usurped political power in larger states and empires, establishing effective autonomy or even formal independence from the central authority. Such political fragmentation and disintegration, when chunks of the state’s civilian and military bureaucratic apparatus broke loose and reformed on a smaller scale under provincial governors or generals turned autonomous rulers, are not to be equated, however, as it often has been, with that particular type of regime which belongs within the fragmentation and disintegration range: feudalism. Nor, it will be noted, is feudalism to be equated with aristocratic dominance in the countryside, which was pretty much the rule in pre-modern state-societies.

The more stratified these societies were, the more weight the landed aristocracy carried – socially, economically, politically, and militarily. It carried great weight because it was rich and powerful, and possessed a host of retainers and slaves in a social environment characterized by a graduated hierarchy of statuses and classes

3

among the mass of judicially free population. Merovingian Francia and Anglo-Saxon England are instances of such societies in early medieval Europe, but examples from across time and space abound. These societies may have had the potential of developing feudalism without being the same. Similarly, while scholars have rightly emphasized the backward agrarian character of the seigniorial-manorial economy, most pre-modern natural economies did not develop that peculiar form of economic, political, and juristic subjugation. What made feudalism special in the family of fragmented states and among societies where the landed aristocracy was dominant was that feudalism emerged as a non-bureaucratic landed-equestrian system for military purposes; the landed-equestrian military elite was then able to usurp political power from the center and effectively rule their localities, reducing the countryside’s population not merely to subservience but to servitude, thus turning feudalism into a full-fledged non-bureaucratic socio-political regime.

The role of the warhorse in this process has remained curiously unclear. All classical definitions of feudalism specify that feudalism involved the supremacy of a specialized class of warriors, predominantly sustained by land allocation. Surprisingly, however, all these definitions ignore a crucial trait which is otherwise practically synonymous with feudalism; that is, that these warriors were invariably horsemen.9 Scholars have understandably recoiled from pinning an entire, multi-faceted social regime on an animal, important as this animal may have been. They have avoided this even though European feudalism was wholly identified with military horsemanship, and has even been famously explained as such by Bruner and White (leaving aside for the moment the criticism regarding White’s particular development of Bruner’s thesis, criticism which the present author shares).10 To be sure, central authorities in many state societies - from early ancient Mesopotamia and ancient Egypt on – also allocated land to sustain foot warriors. Some scholars have even designated such cases as feudal. All the same, there was no feudalism except for horse feudalism, with only landed horsemen having the potential of generating feudalism in underdeveloped agricultural societies. The reason for this was not zoological but social-economic: sustaining a horseman – and a heavily armoured one in particular - was far more costly than sustaining a foot soldier, thus necessitating possession of a much more substantial landed property, or ‘estate’. In societies for which records exist, horsemen normally possessed, or were allocated, at least twice and up to fifteen times more land than infantrymen, with heavy, fully armoured, cavalry occupying the top on this scale.11

It should be remembered that in sedentary societies the horse possessed little economic-utility value. Carts and ploughs were tracked by oxen until the breast or shoulder horse harnesses spread through Eurasia during the first millennium AD, replacing the inefficient throat and girth harness of antiquity. On the other hand, the horse required specialized and expensive feeding. The main utilitarian function of this luxurious possession was its use as an instrument of power, strengthening the aristocracy both in war and domestically. The advent of the horse added a new element to the centralization-fragmentation tension. To the extent that it increased the military strength of the aristocracy in relation to the rest of the population, turning the aristocracy into a mounted elite force, it ipso facto also strengthened this elite in relation to the central state authority. At the same time, provided with longer and easier reach, mounted elites became that much more able to dominate dispersed peasantry in the countryside. Consequently, whether originally the local rich and powerful of an agrarian society developed into an equestrian military class, or warriors were allocated equestrian estates by the state in order to sustain them as

4

horsemen (both processes variably intermixed), the landed horsemen constituted ipso facto the local rich and powerful, whereas landed foot warriors did not.

As Max Weber, who stretches the concept of feudalism to cover all sorts of land-holding warriors, admits: ‘The last-mentioned cases are functionally and also legally similar to the fief proper without being the same, because even privileged peasants remain, socially speaking, peasants or, at any rate, “common people”.’12 Coulborn and the other contributors to Feudalism and History have followed the same route, discussing – but ruling out – feudalism in cases of infantry ‘fiefs’, such as in ancient Egypt and Hammurabi’s Babylonia. Only horsemen were capable of becoming – in effect became - agents of processes of feudalization in all recognized historical cases.

Of course, as many, probably most, societies of the pre-modern world demonstrate, the mere presence of the warhorse was insufficient for feudalism to evolve. Specific social, economic, political, and military conditions had to be in place as well. This article claims the following with respect to the conditions that were conducive to the growth of feudalism: feudalism could only emerge a) in societies that possessed the horse (and hence only in Eurasia from around the middle of the second millennium BC); b) in circumstances that granted the horse preference as an instrument of war; and c) in large states with the most rudimentary small-scale agrarian economy, states which lacked the economic and bureaucratic infrastructure to support and administer the desired but expensive mounted troops by means other than land holding - ‘rent’ was substituted for ‘tax’.13 All of these three prerequisites had to be present for feudalism to evolve.

The crucial but insufficiently recognized factor here is the following: in all pre-modern states military expenditure constituted by far the largest item on, often the large majority of, the state budget; and horsemen were the most expensive military arm. Indeed, where the mounted arm was paramount, running a state was pretty much tantamount to the ability to raise and sustain that arm. It was this task that generated feudalism. Poor administrative apparatus and rudimentary, small-scale, economy meant that de-centralized out-sourcing on an individual basis was the governing principle of the system all the way down the command-administrative structure. Subordinate commanders, regional governors, and local strongmen became territorial lords, responsible for raising and leading the horsemen in their respective territories. They, in turn, had no other option but to repeat the process downward through land allocation to subordinates. Only in the lowest tiers of this hierarchic structure did the network become sufficiently small-scale for the regional lords in some feudal systems (e.g. Japan) to be able to keep their warriors with them on rations and other payments in kind rather than further allocate land to them.

For the problem with land tenure was that it placed the means of payment in the hands of the service providers rather than reserve them under central control. Thus, benefits could not be held back at will when the state wished to terminate the relationship or if the benefactor failed to meet his obligations satisfactorily. It is this which explains the prominence of the personal oath of allegiance in feudal systems, most notably the European, which except for that truly significant factor, the balance of power between the lord and the benefactor and the ultimate threat of the fief being confiscated, was the only means for ensuring that the latter kept his side of the bargain.14 In Japan, where payments to the mounted warriors by the feudal lords (daimyo) were more common than in Europe, the warriors controlled a much smaller portion of the land, the oath of fealty played a less significant role, the gap between lord and knight was wider, and warrior mobility from one feudal master to another

5

was greater.15 All these long-noted but little explained differences between Japanese and European feudalism were closely interrelated. Agrarian possession and political and judicial authority thus devolved lower down the feudal hierarchy in Europe than in Japan, but the principle was the same. Sitting on the means of payment and holding the monopoly over armed force, the landed-equestrian warriors over time were invariably able to extract hereditary rights of possession over their estates, and indeed, appropriate for themselves political and judicial authority over the peasants in the surrounding countryside, whom they ultimately reduced to servitude.

This was the feudal vicious circle. With the state attempting to shortcut the need to collect revenues and reallocate them again through a central bureaucratic-administrative machinery which it possessed, if at all, only inadequately, the horsemen and their command system were directly linked to the sources of revenue in the countryside, only to take control over these sources and countryside, further drying up the state’s income and ability to sustain a central administrative system. And there was another, final, twist that completely overturned the logic of a system which had been intended to secure a standing force of professional warriors for the state: in due course, the increasingly empowered landed-aristocratic mounted warriors were often able to impose a time restriction on their obligatory period of military service for the suzerain, which in feudal Europe, for example, was limited to no more than sixty or even forty days.

In view of these processes, Ganshof has implied that feudalism represented a miscalculation by the state’s central rulers, who failed to foresee the ultimate outcome of the initial steps of land allocation.16 Other scholars have held, on the contrary, that the feudal structure was the only one that kept together and sustained otherwise tenuous realms.17 In their conflicting ways both views in effect highlight the dilemma of poorly developed large states. Indeed, both sides of the coin are reflected in the ‘deviant’ case of Norman England, as compared to the ‘classical’ French model. Inevitably relying on a feudal-knightly power structure for establishing their rule, yet empowered by their unique position as conquerors and benefiting from hindsight on the French experience, William I and his successors imposed an oath of fealty to the crown on all fief holders, whether or not they were the crown’s tenants-in-chief. By this and other measures, including the imposition of a system of royal justice, the Norman kings took care to prevent total fragmentation of the state, creating a more centralized form of feudalism.

The process of feudalization was self-reinforcing in yet another way. The greater the social dominance of ‘big men’ and their retinues in traditional, agricultural societies, the more mass popular militias declined in significance and military effectiveness. This was a two-way process: the more subservient the populace, the more reluctant the elite was to see it possess and become accustomed to the use of arms, which might be used domestically against the socially superior; at the same time, impoverished, disenfranchised, servile, and dispirited mass peasantries, with very little stake in the society and in the fruits of war, were poorly motivated for war and exhibited scant martial qualities, as compared to self-reliant freeholders. Consequently, the mounted warriors’ military and social preponderance would ‘ossify’. Even when the particular circumstances that had favored horsemen and triggered feudalism had changed, the state found it difficult to raise national infantry armies, for power relations in society had been transformed by the process of feudalization in such a way that significantly reduced the number and standing of free men.

6

Thus, in Western Europe, for example, infantry armies only reemerged - regularly and soundly defeating the knightly cavalry - with the revival of urban-civic communities and of central state power from the twelfth century on. Well known examples include northern Italy, Flanders, the Swiss cantons, Scotland, England, and Bohemia. It should be remembered that this ‘infantry renaissance’ took place well before the advent of field firearms, which in popular view are associated with the demise of the mounted knight. Nor is the defeat of the equestrian elite to be explained by any other innovation in weapon technology, such as the longbow. In most cases, above all the Swiss, the defeat of the knightly cavalry was achieved by the revived dense infantry formation, that employed pretty much the same shock weapons and old and simple mass tactics used in antiquity. Neither the rise nor the demise of feudal chivalry were primarily related to technological change.

Indeed, perhaps the most misleading commonly held view with respect to horsemen is that they were invariably superior to infantry militarily, or at least became so sometime in history, e.g. (according to White) after the introduction of the stirrups throughout Eurasia in the middle of the first millennium AD. As anybody familiar with the subject is aware (and as Machiavelli points out in Book II of his On the Art of War), the horse is a sensitive and highly vulnerable animal. Consequently, horsemen were hardly able to withstand a head-on clash with infantry, provided – and these are major conditions – that the latter were massed in close order, kept their cohesion and morale, and were equipped with the necessary though simple type of weapons (mostly spears or pikes). On the other hand, the horsemen’s chief advantage was their mobility, particularly on open ground. The power equations between infantry and horse were largely modulated by the different balances between these variables in a diversity of specific settings. To put it as succinctly as possible, the horsemen’s effectiveness gained vis-à-vis infantry in the following circumstances: a) the flatter the terrain, where horsemen were able to operate swiftly, both tactically and strategically, unhindered by obstacles of a rugged - e.g. mountainous, wooded, and swampy - landscape; b) the greater the distances of military action within larger theatres of operations, where the horsemen’s much greater strategic mobility would come into play.

The notion that enhanced cavalry power inaugurated not only feudalism but the Middle Ages in general is well entrenched in many minds. In actuality, however, horsepower did not have that much to do with the fall of the Roman Empire, and the stirrups had even less. In the late Roman Empire – and well before the stirrups - cavalry was growing in significance, especially in the eastern provinces. The long Danube frontier required more mobile forces to forestall Gothic and Hunic horse raids from the Ukrainian steppe; the long Near Eastern limes even more badly demanded mobile mounted troops to counter the large cavalry forces of the Sasanid Persian Empire in the vast open spaces of the Levant. The Battle of Adrianople (AD 378), where the pre-stirrups Gothic cavalry suddenly arrived to take the Roman army from the flank and by surprise as the legions were attacking the Gothic infantry and wagon camp, leading to the Romans’ annihilation, is largely responsible for the misconception, as if the battle inaugurated the ‘Age of Cavalry’.18 In fact, quite a number of battles in antiquity had been decided by a similar cavalry ‘hammer and anvil’ action (Alexander’s, and - involving Rome – Cannae and Zama, to mention but a few well-known instances). Furthermore, Adrianople was a traumatic but isolated episode in Rome’s fall. The Goths from the Ukrainian steppe indeed possessed a strong cavalry arm, but most of the Germanic peoples that invaded and dismantled the Western Empire in the fifth century primarily consisted of tribal foot warriors.

7

The horse was instrumental in the fall of the Western Roman Empire only indirectly, in the sense that Hunic raiding into Eastern and Central Europe drove the terrified Germanic peoples wholesale into the Empire. Cavalry did not suddenly prove superior to Roman infantry; rather, the advent of the steppe horsemen triggered a sweeping chain reaction among the Germanic peoples all across Rome’s barbaric march, whose migration, settlement, and predominantly foot armies ultimately overwhelmed the empire. Indeed, the Goths’ reliance on cavalry was one reason for their ultimate defeat by the Frankish infantry.19 As is well recognized, it was only the particular military threats facing Francia much later on, during the eighth to tenth centuries, that called for highly mobile and readily available mounted troops. And similar strategic requirements that favored professional horsemen were conducive to the growth of feudalism in some other poorly developed large states as well. Scholarly opinion is more or less in agreement on the designation feudalism only with respect to three historical cases. The earliest of these is China, following the introduction of the horse and war chariot from the Eurasian steppe in about 1200 BC. The war chariot increased the power of the regional lords and their retinues, residing in fortified lodgings or ‘castle towns’, vis-à-vis both the overlord and peasantry. At first, the number of chariots was small. Archaeological evidence of the late Shang shows that they mainly served for the purpose of the king’s transport and ceremonial activity. Even the Chou Dynasty from the west, whose overthrow and replacement of the Shang as overlords in about 1050 BC may be at least partly attributed to their superiority in chariots, reportedly possessed only some 300 chariots during the conquest. However, the new overlords increasingly relied on the mounted arm to serve as a readily available force, rapidly deployable through their vast realm. The old provincial aristocracy transferred to chariots, and the Chou extensively allocated new estates and created vassal states as a means for governing new territories. The local conscript infantry militias declined in proportion to the military and social rise of the mounted aristocracy. The feudal snowball was gathering momentum. By the time of the later, Eastern, Chou (from 842 BC), the monarch’s effective power was confined to the royal domain. By the so-called Spring and Autumn Period (722-481 BC), the realm disintegrated into hundreds of practically autonomous polities, whose rulers or ‘dukes’ maintained only the semblance of vassal subservience to the Chou overlord to whom they swore allegiance. In the resulting anarchy, the regional aristocratic chariot warriors (shi) were engaged in endemic warfare among themselves, cultivating typical knightly warrior ethics.20 The two other, better known, cases of feudalism are the Japanese and the European, in both of which the circumstances in which feudalism emerged and its trajectory were remarkably similar. In both the Japanese and Carolingian realms, society was small-scale agricultural, practically non-urban and non-mercantile, and overwhelmingly illiterate. Communications were also poor. In Europe, the legacy of Roman civilization in all these fields had sharply declined after the Empire’s fall. In Japan, the cultural imports from Chinese civilization, though highly significant, had nevertheless been confined to the center and were superficial in their penetration of the country and society. Growing into large-scale state form only shortly earlier, both the Japanese and Carolingian states possessed underdeveloped infrastructures to contend with the administrative and military organization of their vast realms. Furthermore, both of them had until then waged their wars mainly using short-term peasant militia armies, and both were finding this instrument less suitable for their new, remote, frontier wars.

8

For the endemic raiding, counter-raiding, and manhunt operations against the tribal barbarians on its northeastern frontier, who relied heavily on horse archery, the Japanese state found standing horsemen retinues of the provincial notables and large estate owners (shoen) far more effective than the cumbersome and poorly motivated peasant conscripts. The Carolingian rulers - finding as their main military challenges warfare in remote frontier zones, mounted raiding by the Moslems from Spain and by the Hungarians, and maritime and river raiding by the Norsemen - similarly inclined towards readily available and fast moving cavalry as their most effective military force. Thus, both Japan and Francia gradually but increasingly relied on mounted troops sustained by land holdings, leaving their peasant infantry militias and conscript forces to decline. Conscription was officially abolished in Japan in AD 792, while the Frankish kings by the mid-ninth century openly preferred the mobilization of horsemen who would come under the command of regional lords. Here again the process of feudalization was to run its course during the following centuries. The regional lords and mounted warriors extracted hereditary rights over their land, extended social, political, economic, and judicial domination over the countryside, reduced the free peasants to servitude, and turned themselves into a closed chivalrous aristocracy (samurai; knights).21 Central power was substantially diminished, in some places to near insignificance. In both Japan and Europe, feudalism evolved not only along largely similar lines (also exhibited by the Chou), but more or less during the same period of time. The mounted warriors sustained by land holdings in the provinces began to rise in significance in the eighth century AD, with the feudalization of the system peaking in the eleventh and twelfth centuries in north-western Europe and in the fourteenth to sixteenth centuries in Japan. Adherents of Lynn White’s thesis might regard this parallelism as a vindication of it. After all, even though White confined himself to European feudalism, he explained it by the invention and diffusion of the stirrups throughout Eurasia in the middle of the first millennium, which should be expected to have affected other regions of the landmass as well. However, not only were the rising horse warriors in Japan mounted archers rather than lancers (whereas the latter was the type of cavalry which is supposed by White to have grown decisively more effective because of the stirrups); it should be noted that across the Eurasian landmass it was only in Western Europe and Japan that feudalism emerged as a full-blown system. This exclusiveness indicates a particular similarity in conditions between the two regions, which also warrants the extension of the designation medieval to Japan, alone of all the other regions of the world to which this European periodization is arbitrarily applied, for their developed and urban civilizations were proceeding pretty much as before. Indeed, why did feudal regimes par excellence materialize only in these two backward regions, along the two edges of Eurasia, rather than throughout the landmass, where the stirrups had diffused? Moreover, why, as most scholars more or less agree, did feudal regimes par excellence only materialize in very few historical cases, either in ‘medieval’ times or earlier? According to the argument advanced here, it was the scarce combination of all of the above-mentioned necessary preconditions for the emergence of feudalism that explains its distribution and relative rarity. In all other large state-societies that possessed the horse, full-blown feudalism did not evolve because: a) strategic conditions did not favor mounted troops; and/or b) the populace – rural or civic - was able to uphold its social and military position vis-à-vis the mounted elite; and/or c) society was sufficiently developed in terms of the economy, communications, urbanism, and literacy for the central state authorities to

9

produce the infrastructure needed to sustain and administer the armed forces directly, relying on the state’s own revenues and bureaucratic system. They were thus able to avoid the drift down the slippery slope of military and economic power outsourcing and political fragmentation, so detrimental to central authority. The fundamental significance of these preconditions for feudalism is demonstrated by the so many more cases of state societies that incorporated feudal traits or were even ‘semi-feudal’, without reaching the feudal ‘pure model’ exhibited by Chou China, medieval north-western Europe, and medieval Japan. These ‘semi-feudal’ societies have regularly confounded scholarly discussion regarding feudalism’s wider applicability. Consequently, narrow and broad concepts of feudalism emerged. The narrowest, originating with Montesquieu, confined feudalism to Europe.22 To this Marx and most modern scholars added the one or two other cases of the ‘pure model’ and, reservedly, possibly a few more proximate cases.23 Yet broader concepts of feudalism, originating with Voltaire and developed by Weber and many Marxists, employed a looser frame to incorporate wider categories of landed-military aristocratic dominance.24

How does this feudal ‘gradualism’ impinge on our understanding of the deeper causes of the feudal phenomenon? According to the line of explanation pursued here, feudal traits and even ‘semi-feudal’ systems existed where the preconditions for feudalism only partially materialized. Most typically, partially feudal states relied on the landed aristocracy - or, indeed, initiated a systematic policy of land allocation - for the maintenance of their mounted troops, either because of the system’s economic simplicity and/or on account of the central authority’s need to compromise with and accommodate regional aristocratic power. However, partially feudal states presided over a more developed commercial, urban, and literate society than that characteristic of the ‘pure model’. Consequently, they possessed central bureaucratic and tax collecting systems to a degree that also made it possible for the central authority to rely on its own revenues and raise troops from sources other than the feudal. In addition, the central authority was more capable of administering the landed horsemen directly, through its own command and administrative structure, rather than being obliged to rely on a landed-feudal hierarchic pyramid. Consequently, the economic-administrative principle of land tenure for raising horsemen, with the usurpation and fragmentation of political power which is feudalism that it brought in its train, was not allowed to take over entirely. Instead, this principle and, hence, feudal tendencies remained balanced and restricted by other methods of financing and raising troops. A more mixed social, political, and military equilibrium, and, thus, a more centralized state, was the result. The landed horsemen, while certainly powerful both militarily and socially, and often enough also able to secure hereditary rights over their land and a degree of domination over the countryside, were less successful in appropriating political and judicial authority to the virtual fragmentation of the central state.25 It thus turns out that feudalism stemmed from one of the options open to large states for raising and sustaining the costly mounted arm - their most significant military, economic, and administrative task where this arm predominated militarily; indeed, that it was a consequence of the ‘primitive’ option, which, in the absence of a developed central economic-bureaucratic infrastructure, linked the upkeep of horsemen directly to the sources of revenue in the countryside. Contrary to the implication of the simple Marxist model of the Communist Manifesto, feudalism does not represent a ‘higher’ evolutionary stage in relation to ancient society in world history. True, in medieval Europe and Japan feudalism emerged after ‘antiquity’ (and may or may not have been more conducive to the subsequent growth of capitalism);

10

but in Europe it evolved in the large but relatively backward north-western Frankish realm after the collapse of the advanced literate, urban, moneyed, and bureaucratic societies of classical Mediterranean antiquity centuries earlier; while in Japan it gained power at the expense of a newly created large centralized state that was economically and socially more or less as undeveloped as its feudal successor - its rudimentary, imported Chinese civilization notwithstanding.

Indeed, China is an instructive case in point, for there feudalism evolved in the ‘right order’ during the Spring and Autumn period (722-481 BC), out of an archaic imperial overlordship (Shang/Chou), like in Japan and the Carolingian domain, but before the centralized-bureaucratic-urban-moneyed states of classical China that would replace it.26 Thereafter, in later Chinese history, even during periods of imperial disunity and political fragmentation and disintegration, feudalism was never to re-emerge as anything near the ‘pure model’. In contrast to north-western medieval Europe, sufficient levels of urbanism, commercialism, and literacy survived during China’s periods of disunity to support bureaucratic-administrative-moneyed systems in splinter states and at the service of regional warlords.27 Semi-Feudal and Centralized-Bureaucratic Regimes Thus, on the fragmentation-centralization continuum of state structure and military organization, feudalism in ‘purity’ occupied one end, coexisted with other forms in partially feudal systems, and tended to diminish in more bureaucratic regime types. This continuum is demonstrated by the civilizations of the ancient Near East, when the war chariot was introduced into them from the mid-seventeenth century BC on. While affecting as sweeping a revolution in military affairs as it would in Chou China, the chariot’s incorporation by the region’s various polities produced a diversity of political-administrative-military regimes, depending on the particular circumstances of each case.28

Information is uneven and patchy. For instance, we know little about the internal structure of the powerful Mitanni-Hurrian Empire in eastern Anatolia-northern Mesopotamia (late sixteenth to late fourteenth centuries BC). The evidence suggests that the kingdom’s elite chariot warriors (mariyannu) were sustained by landed estates, and that the monarch mainly functioned as a military overlord. Still, how the landed equestrian elite stood in relation to the rest of society and how feudal the system was remain in the dark.29

Information about the New Hittite Empire (c.1420- c.1200) is more abundant. The Hittite Empire had emerged as, and largely remained, an overlordship. Yet the Hittite Great Kings possessed a sufficiently developed bureaucratic machinery and vast treasures, derived from taxation as well as from tribute and booty. They were thus able to retain their predominance over the realm and keep regional aristocratic power holders in check. The mounted arm increasingly grew in strength and significance during the New Kingdom’s lifetime, expanding from hundreds of chariots to thousands, as campaigning became increasingly long-distance and directed into the Syrian and north Mesopotamian plains. According to Egyptian records, 3,500 Hittite and allied chariots participated in the Battle of Kadesh (1285/1274 BC) against Pharaoh Ramesses II.30 However, in addition to the mounted semi-feudal aristocratic vassals and their retinues, the king directly commanded regular armed troops in the form of a royal guard and mercenary forces, paid in money and in kind (rations). He also maintained frontier garrisons, supported by either or both systems of land allocation and rations. And he was in control of the labor and militia service of the peasants, who while being socially subservient and possibly declining in military

11

significance in view of all the above, were never reduced to servitude to the aristocracy. Semi-feudal, bureaucratic, moneyed, and conscript elements were intermixed and mutually balanced within the Hittite state structure, with the central authority retaining its ascendancy.31 State structure in Egypt of the New Kingdom, the third Near Eastern great power of the age of the chariot, was yet more centralized. With her relatively homogenous and isolated territory traversed by the Nile communication highway, her developed literate bureaucracy, and her powerful monarchy traditionally reinforcing one another to produce a highly centralized state, Egypt of the New Kingdom established a centralized chariot force. For Egypt, too, the great power struggle involved distant campaigning in the Levant, and the deployment of chariots, paid garrisons and mercenaries, and conscript peasants. Here, too, the chariot force was an elite corps, growing to thousands by the fifteenth to thirteenth centuries BC. However, in Egypt, the warriors who rode the chariots constituted a service elite, which served at state facilities and in the court. As in all agrarian societies, the granting of land was, of course, a principal means for remunerating this mounted warrior elite, but other means of payment, in cash and in kind, were also used. The mounted warrior held possession of his chariot, but much of the facilities were concentrated in a system of royal stables, from which central control was exercised.32

In the major ‘palace’-city petty-states of the Levant and the Aegean, the chariot forces were similarly or even more centralized. Mercantile, wealthy, and territorially small, these polities were highly centralized and bureaucratically run. In the most powerful of them, such as Ugarit and Hazor in the Levant and Mycenae and Knossos in the Aegean, the mounted warrior elite (the Hurrian term mariyannu was borrowed throughout the Levant) rode chariots that were owned by the state and minutely supervised by its central apparatus. As in late medieval Europe, the aristocratic-military horsemen appear to have been maintained in various statuses and by a variety of remuneration methods, including direct payments and land allocation.33 The shift from chariots to horseback riding for military purposes (initially the two equestrian forms coexisted for a few centuries) barely affected the parameters of the fragmentation-centralization continuum of state structure. This continuum extended from the relatively rare cases of ‘pure’ feudalism for sustaining the expensive cavalry arm, to the semi-feudal, to more fully bureaucratic systems of more highly developed polities. Military horseback riding was first introduced into the civilizations of the ancient Near East during the ninth century BC. It was incorporated in a fairly centralized-bureaucratic form into the armies of Assyria. In Assyria, as elsewhere, the aristocracy had been mainly equestrian, riding first war chariots and later also horseback. However, despite periodical lapses, royal power was sufficiently strong to prevent the feudalization of the realm. Furthermore, as Assyria had become a huge tribute extracting machine, much of the cavalry during the later Empire (mid-eighth to seventh centuries BC) consisted of paid professionals/mercenaries. The procurement and raising in large farms of the tens of thousands of horses which the army required became a major state industry, bureaucratically run by a highly developed state apparatus.34

By comparison, in politically and economically less developed states on the Assyrian march, feudal forms variably intermixed with state-centered means and methods for raising troops. Iran is a major example. The Median state crystallized as an overlordship in 673 BC, in response to protracted Assyrian pressure, and allied with Chaldaean Babylon to finally destroy Assyria in the late seventh century. Like all

12

the powers of the region, Media had striven to emulate the Assyrian military system, combining shock and missile infantry, horsemen, and siege and engineering corps. Apparently all free men were liable for service, and the king’s wealth from booty, tribute, and tax also made it possible for him to retain some permanent household and garrison troops. The power of the landed aristocracy and its retinues, which provided most of the cavalry and much of the military leadership, thus appears to have been checked within a ‘mixed’ state structure. All the same, the aristocracy remained very powerful, and may have even gained in strength as it accumulated great wealth with the empire. Indeed, when the Median aristocracy grew dissatisfied with the monarch Astyages, who had attempted to curb its power, it switched its allegiance to the Achaemenid Cyrus of Persia, assisting Cyrus into the throne of a combined Persian-Median Empire (550 BC).35

However, under Cyrus and his successors the empire expanded over the entire ancient Near East, incorporating the region’s great centers of civilization. With the empire now in command of fabulous wealth, the commercial and literate resources of major urban centers, and a developed road system constructed by the state, Darius I (522-486 BC) turned it increasingly bureaucratic, curbing the power of the aristocracy. A central permanent army of perhaps 10,000 cavalry and 10,000 infantry was established, in addition to garrison troops positioned in key locations throughout the empire, some of them foreign mercenaries. Together with money payments, land allotment remained one of the methods for supporting semi-permanent troops of all arms, especially in the provinces. The king continued to grant large estates to his favorites, and the Persian-Mede landed-equestrian aristocracy in general remained rich and influential. However, it took its place as a service elite within the state’s apparatus, in the imperial royal palaces and capital cities or in ruling the imperial provinces.36 As empires periodically reemerged in subsequent Iranian history, they exhibited much the same features as their Median and early Achaemenid predecessors. In both the Parthian and Sasanian empires (247 BC-AD 224 and AD 224-651, respectively), the great landed aristocracy and its mounted retinues constituted the state’s elite fighting force, maintaining a strained balance with the central authority. The infantry mass, called up from among the free men, was secondary and subservient to the mounted aristocracy. The more these empires expanded to include important urban centers (mainly Hellenistic and Mesopotamian) beyond the Iranian upland plateau and the more the kings were able to rely on the taxation of land and trade and on other sources of recruitment, raising household troops and hiring foreign mercenaries, the more successful they were in tilting the balance in favour of the central royal power in these semi-feudal states.37 In the subsequent Turkic-Iranian empire of the Safavids (AD 1501-1736), the shahs undertook similar measures to curb the power of the tribal-feudal aristocracy.38 In the old Near Eastern centers of civilization and urbanism, the balance tilted even farther in favour of the central state authority. The fief system for sustaining cavalry was widely employed, generating processes of feudalization in some periods and regions. However, states overall were in command of relatively developed economic and administrative infrastructure and were thus more successful in keeping feudalization in check. Even the beneficium itself incorporated more advanced, financial, means in comparison to the landed fief of the European natural economy, often consisting of income from commercial and industrial enterprises (the Byzantine pronoia, Arab iqta, and Turkish timar).39

13

The Byzantine Empire, for example, relied extensively on land allotment to both infantry and cavalry, with the expensive-to-maintain cavalry receiving five times more land than the infantry. However, the wealthy empire also paid for a strong central army (which expanded and shrank with the changing fortunes of the empire during its thousand year history) and foreign mercenaries, and was highly bureaucratic. Consequently, the relatively affluent holders of cavalry fiefs never really had the space to grow into feudal strongmen.40 In the wake of the Arab conquest in the seventh century AD, the lands of Islam were ruled by the elite tribal nomads and paid warrior contingents, centered on garrison towns. Later, however, the fief system was extensively practiced, tenuously balanced by the ruler’s household troops and paid mercenaries. Here and there, a decline of central power and processes of feudalization went hand in hand.41

The Ottoman Empire, too, resorted to extensive fief allocation for sustaining its mounted warriors (sipahi), which in its peak, in the sixteenth century, reached 100,000-120,000. But, again, the sultans possessed a strong central standing infantry force (janissaries), provincial garrison troops, and well-developed professional services, all kept on the empire’s enormous wealth. Furthermore, the Ottoman rulers were able to draw on the human resources of the subject provinces for creating the administrative machinery that governed the empire. Only during the decline of the empire did the sipahi win hereditary rights and greater domination of their localities, which resulted in deepening processes of feudalization. They also increasingly avoided military service.42 The structure and development of imperial polities on the Indian sub-continent was fairly similar.43 All this also accords with the trajectory of that model case of feudalism, the European. For if it was the backwardness of the economic and bureaucratic infrastructure of the Frankish state that set the process of feudalization in motion once the military demand for cavalry increased, it was precisely the growth of that infrastructure in the new European monarchies that progressively rolled feudalism back. It was this rather than any particular, externally or internally induced, economic crisis within the seigniorial-manorial ‘mode of production’ that brought about its decline.44 In southern Europe, where urbanism, trade, and money economy had not declined as much as in northern Europe, feudalization had been more effectively checked and more quickly receded. But in northern Europe, too, as feudalism reached its peak in the eleventh to twelfth centuries AD, the rise of the cities and the revival of trade would increasingly provide the state monarchs with both sources of revenue and administrative skills.

Thus, by the late Middle Ages, the monarchs were more and more able to expand their household troops, which they increasingly kept on salaries, hire foreign mercenaries, keep feudal levies in service for longer periods of time on cash payments, and revive national militia infantry armies of free men, both civic and rural. They increasingly imposed taxes and expanded the administrative apparatus designed to supervise over all this, as well as employing private entrepreneurs on market principles.45 They steadily grew in power vis-à-vis the feudal aristocracy, in a process that was as self-reinforcing as feudalization had been. Consequently, from the thirteenth century on, the European system no longer approximated the ‘pure’ feudal model, but transformed into the ‘semi-feudal’, estate, or corporate state; that is, it included strong feudal features and elements alongside other - civic, moneyed, and centralized-bureaucratic - methods for raising troops and ruling the country. As the process would run its course during the early modern period, the old feudal aristocracy would increasingly transform into a service aristocracy, manning the

14

upper echelons of the state’s machinery in the bureaucracy and armed forces. Thus, the erosion of Europe’s ‘pure feudalism’, like European feudalism itself, should not be viewed in isolation, but considered within a much wider, comparative, context and in light of the fundamental preconditions which brought feudalism into being.

It should be noted that if the central authority sometimes allied with the free population - both civic and rural - to curb feudal-aristocratic power, it engaged here in a fine balancing act, for it even more routinely formed a common front with the aristocracy to keep the populace in its place. In most societies, the aristocracy was indispensable to the state for both its paramount military role and as the upper stratum of a social-political system based on the subservience of masses of tribute paying agrarian producers.46 Japan offers a particularly interesting example in this respect. As in Europe, Japan’s feudal system was transformed in the fifteenth and sixteenth centuries, as the major territorial lords (daimyo) raised strong infantry armies of commoners, armed with pikes, halberds, crossbows, and, from the 1540s, muskets. These armies increasingly dominated the battlefield, eclipsing samurai military ascendancy. However, as Japan was united by these means under the strong central government of the Tokugawa Shogunate (1600-1868), the peculiar mixed absolutist-feudal regime now established, while keeping the regional lords closely in check, also monopolized military force in the hands of the samurai, abolishing budding municipal autonomies and completely disarming the populace.47 The survival of this regime (which also suppressed firearms) was only possible, of course, because of Japan’s isolation from the rest of the world, which freed her from the external constraints of power politics, at least for a time. For the sake of comparison, it is finally worthwhile to examine the most centralized-bureaucratic political-military regimes. In China, for example, in a process which in many ways resembled that of late medieval Europe, the complete feudalization and fragmentation of the system during the Spring and Autumn period (722-481 BC) was reversed during the Warring States period (fifth century to 221 BC), as central states’ authorities, largely drawing on the economic and human resources of emergent urbanism, consolidated the realm into a small number of increasingly centralized, bureaucratically run, states. The largest of these states possessed chariot forces which numbered in the thousands and which were increasingly controlled by central administrations. Furthermore, the growing centralized-bureaucratic monarchies also created infantry armies of conscript peasants, organized and commanded by the state. Conjointly, the peasants were freed from their subservience to the aristocracy and were granted private possession of their land. This was a central element in the process by which feudalism was crushed and huge state armies were created, vying with one another for supremacy, until the State of Ch’in, which had pushed this process the farthest, conquered all others and united China. In the strongly bureaucratic united empire of the Ch’in (221-206 BC) and Han (from 206 BC), the cavalry (which had replaced chariotry) was just another arm, and, as in Assyria, a special state bureaucracy took care of the procurement and raising in huge state farms of the army’s hundreds of thousands horses.48

In Rome, too, the aristocracy constituted the cavalry during much of the Republic, though in warfare among city-states, in rugged terrain, and over relatively short distances, the cavalry’s military prowess was inferior to that of the infantry. All the same, as the Roman army was professionalized during the late Republic and under the Empire, the cavalry, as in Imperial China, became just another arm. In both the above cases of centralized bureaucratization, aristocratic social supremacy and the mounted arm became largely disassociated from one another. Rather than consisting

15

of knights and cavaliers of all sorts and designations, the mounted arm would become simply cavalry. Conclusion There are two main problems with the accepted ‘definitions’ of feudalism. First, these definitions mostly abstract the north-western European model of the Frankish realm, ‘from the Rhine to the Loire’, which is taken as standard. Scholars have not really been able to make up their minds as to how to regard ‘deviations’ from this model, large or small – be it in other parts of medieval Europe (such as Norman England, Southern Europe, and Eastern Europe), in Japan, or in a large variety of ‘semi-feudal’ societies. Having catalogued the differences, they have been left unsure whether or not it is still legitimate to speak of feudalism in such diverse cases. What has been missing is a deeper understanding of the feudal in feudalism, which might then be detected to a larger or smaller degree in a whole range of historical societies, and in relation to which one trait or another of these societies could be explained. As this article has attempted to show, at the basis of the feudal phenomenon was the principle of sustaining the state’s mounted warriors on revenue-yielding property, overwhelmingly land, with the state either intentionally allocating the property or relying on existing property holders, or both. This principle was fairly widespread because it was one of the main options open to the state’s central authority for financing what was probably the most vital and costly item on its budget. The principle had a lot to recommend for itself. By directly linking warriors (and other service providers) to sources of revenue, the state was able to shortcut and discard the need to circulate revenue through the whole complex, expensive, and cumbersome intermediate medium of administrative bureaucracy. Furthermore, the benefactors served as a sort of managerial stratum of the allocated resources. It is therefore not surprising that the principle was commonly practiced in pre-modern societies. On the other hand, resource allocation had a serious disadvantage: sitting on the resource, the benefactors over time tended to appropriate it, while also striving to avoid as much as possible their obligation of service. When applied to the upkeep of horsemen, and only then, the principle of resource holding had the potential of generating feudalism.

Indeed, the failure to fully grasp the equestrian element in that process has been another major deficiency in scholarly treatment of feudalism. Since horsemen, particularly fully armoured ones, were expensive to maintain, several times more expensive than infantrymen, their property holdings in any society had to be that much larger than the so-called ‘infantry fief’. Ipso facto this transformed the landed professional horsemen into a social-economic elite, and vise-versa. Given the right conditions, they were thus able to grow into local strongmen, usurp political and judicial powers from the central state’s authority, and reduce the peasants in the surrounding countryside to servitude. This was the process of feudalization, by which a predominantly military system was being transformed into a comprehensive socio-political regime. Large non-urban state-societies with small-scale non-mercantile natural economies, which in addition possessed a special strategic preference for the mounted arm, were particularly vulnerable to such processes of feudalization. Lacking a sufficiently developed administration and tax collecting bureaucracy, they found revenue from individual resource holding – in their case, invariably land – the most natural, if not the sole viable principle for sustaining and administering a large number of costly horsemen. Furthermore, because of their very same deficiency, such states were the least capable of resisting a gradual takeover of local authority by the

16

horse elite and by the landed-feudal hierarchy. It is in such societies, then, that ‘pure’ feudalism tended to grow.

More developed state-societies, too, variably applied the principle of land tenure for sustaining horsemen. However, the more developed a country’s urban-mercantile economy and the state’s administrative apparatus and tax-collecting system tended to be, the more the state possessed other means for raising, sustaining, administering, and commanding troops. Consequently, it was that much more able to resist processes of feudalization. Feudalism, that is, the usurpation of political and juristic power from the central authority by landed-equestrian warriors and lords, would thus be variably checked within ‘mixed’, semi-feudal regimes, or fail to materialize in more fully bureaucratic regimes.

17

1 Most notably: Georges Duby, The Early Growth of the European Economy: Warriors and Peasants from the Seventh to the Twelfth Century, London: Weidenfeld and Nicolson, 1974; Guy Bois, The Transformation of the Year One Thousand, Manchester: Manchester UP, 1992; Susan Reynolds, Fiefs and Vassals: The Medieval Evidence Reinterpreted, Oxford: Oxford UP, 1994. For a critique see e.g. Dominique Barthélemy, La Mutation de l’an mil a-t-elle eu lieu?, Paris: Fayard, 1997. 2 For such skepticism see e.g. Elizabeth Brown, ‘The Tyranny of a Construct: Feudalism and Historians of Medieval Europe’, American Historical Review, 79 (1974), 1063-88. 3 Mark Bloch, Feudal Society, London: Rouledge, 1961, 447. 4 E.g. Bloch, Feudal Society, 446; F. L. Ganshof, Feudalism, London: Longmans, 1964, xv. 5 A useful summary of views and definitions can be found in John Ward, ‘Feudalism: Interpretative Category or Framework of Life in the Medieval West?’, in E. Leach, S. Mukherjee, and J. Ward (eds.), Feudalism: Comparative Studies, Sydney: Pathfinder, 1985, 40-67. 6 Elman Service, Origins of the State and Civilization, New York: Norton, 1975, 81-3; Allen Johnston and Timothy Earle, The Evolution of Human Societies: From Foraging Group to Agrarian State, Stanford: Stanford UP, 1987, 249. But see Bloch, Feudal Society, 446 (and chs. 9-10); Ganshof, Feudalism, xv; also, W. G. Runciman, A Treatise on Social Theory, Cambridge: Cambridge UP, 1989, ii. 208, 368; Kristian Kristiansen, ‘Chiefdoms, States and Systems of Social Evolution’, in T. Earle (ed.), Chiefdoms: Power, Economy and Ideology, New York: Cambridge UP, 1991, 16-23. 7 Well-noted by Rushton Coulborn, in his (ed.), Feudalism in History, Princeton: Princeton UP, 1956, 7, 186. This still valuable collection is all too easily dismissed as outdated nowadays. 8 Max Weber, Economy and Society, New York: Bedminster, 1968, 231-6, 964, 968-71, and chs. xii-xiii. 9 Otto Hintze, ‘Wesen und Verbreitung des Feudalismus’ (1929), reprinted in his Feudalismus – Kapitalismus, Göttingen: Vandenhoeck, 1970, 12-47, esp. 14, 22; Bloch, Feudal Society, 446 (though horsemanship is mentioned as somewhat almost incidental on p. 444); Ganshof, Feudalism, xv. In Weber’s extensive discussion of feudalism in Economy and Society, 255-62, 1070-85, equestrianism is mentioned only once (p. 1077) and inconsequentially. Coulborn, Feudalism in History, 8-9, comes closest, but still refers to equestrianism outside his definition of feudalism (p. 4-6) and as a prevalent rather than essential feature of that phenomenon. Jean-Pierre Poly and Eric Bournazel, The Feudal Transformation, 900-1200, New York: Holmes, 1991, who stress castles and the devolution of power to the local strongmen, but not the equestrian feature of the system, are typical of the more recent literature. 10 H. Bruner, ‘Der Reiterdienst und die Anfänge des Lehnwesen’, Zeitschrift der Savigny Stiftung für Rechtgeschichte. Germanistische Abteilung, 8 (1887), 1-38; Lynn White, Medieval Technology and Social Change, Oxford: Oxford UP, 1962, 1-38. For a criticism of White’s thesis see e.g. Bernard Bachrach, Armies and Politics in the Early Medieval West, Aldershot: Variorum, 1993, chs. xii, xiv, xvii. 11 Data on various societies are often obscure and interpretations vary, but the general picture is clear enough. For the Achaemenids see The Cambridge History of Iran, ii. ed. I. Gershevitch, Cambridge: Cambridge UP, 1985, 281, 573-6. In Solon’s system of classes, horsemen possessed almost double the income of well-to-do farmers who

18

owned a pair of oxen, the backbone of the hoplite army (300-500 as against 200-300 measures): Aristotle, Constitution of Athens, 7.3-4; Plutarch, Solon, 18. On average, Roman equestrian colonists were allocated twice as much land as infantrymen, while their census income during the late republic was supposedly ten times greater: Livy 35:9, 35:45, 37:57, 45:34; Alexander Yakobson, Elections and Electioneering in Rome, Stuttgart: Franz Verlag, 1999, 43-8. Cavalrymen are reckoned to have been allocated four times more land than infantrymen in Byzantium, and the special heavy cavalrymen maybe sixteen times more: Warren Treadgold, Byzantium and Its Army 284-1081, Stanford: Stanford UP, 1995, esp. 23-5, 171-9; also John Haldon, Warfare, State and Society in the Byzantine World 565-1204, London: UCL, 1999, 128. For medieval Europe see e.g.: Bernard Bachrach, Early Carolingian Warfare, Philadelphia: U. of Pennsylvania, 2001, 55: property over 10-18 ha for infantry, 120-216 ha for cavalry; Philippe Contamine, War in the Middle Ages, Oxford: Blackwell, 1986, 88: 2.5 to 4 times more property for knights than for heavy infantry in Henry II’s English army; Charles Oman, A History of the Art of War in the Middle Ages, New York: Burt Franklin, 1969, ii. 127: five times more property for a knight than for an archer on the eve of the Hundred Years War. 12 Weber, Economy and Society, 1071-2. 13 Rent versus tax as the essence of feudalism has been rightly suggested by Chris Wickham, ‘The Other Transition: From the Ancient World to Feudalism’, Past and Present, 103 (1984), 3-36. He does not discuss the necessary preconditions for this change, attributing it to the strength of a tax-evading aristocracy. However, the Merovingian state was not stronger vis-à-vis its aristocracy than the Carolingian rulers who launched the process of feudalization. 14 Cf. the thoroughly idealist interpretation of Poly’s and Bournazel’s generally admirable, The Feudal Transformation, 81, even though they believe that the act represented ‘breathtaking realism’. 15 F. Joüon des Longrais, L’Est et L’Ouest, Paris: Institut de Recherches d’Histoire Étrangere, 1958; John Hall, ‘Feudalism in Japan – A Reassessment’, in his and M. Janssen (eds.), Studies in the Institutional History of Early Modern Japan, Princeton: Princeton UP, 1968, 15-51; id. Government and Local Power in Japan 500 to 1700, Princeton: Princeton UP, 1966; Peter Duus, Feudalism in Japan, New York: McGraw-Hill, 1993; Jeffrey Mass, Warrior Government in Early Medieval Japan, New Haven: Yale UP, 1974; id., ‘The Early Bakufu and Feudalism’, in his (ed.), Court and Bakufu in Japan, Stanford: Stanford UP, 1982, 123-42; Ishii Ryosuke, ‘Japanese Feudalism’, Acta Asiatica, 35 (1978), 1-29; Marius Jansen (ed.), Warrior Rule in Japan, Cambridge: Cambridge UP, 1995 (compiled from vols. 3 and 4 of The Cambridge History of Japan). 16 Ganshof, Feudalism, 51-9. 17 Again see the references in Ward, ‘Feudalism: Interpretative Category’, 45. Also Polly and Bournazel, 7, 9. 18 Contributing to this misconception are otherwise knowledgeable studies such as Oman, A History of the Art of War in the Middle Ages, i. 13-14; J. Verbruggen, The Art of Warfare in Western Europe during the Middle Ages, Woodbridge: Boydell, 1997, 5. But see e.g. Arther Ferril, The Fall of the Roman Empire: The Military Explanation, London: Thames and Hudson, 1986, 7-8, 60, and passim; also White, Medieval Technology, 6-7, for he connects equestrian supremacy to the later introduction of the stirrups. 19 Herwig Wolfram, History of the Goths, Berkeley: U. of California, 1988, 127, 217.

19

20 Edward Shoughnessy, ‘Historical Perspectives on the Introduction of the Chariot into China’, Harvard Journal of Asiatic Studies, 48 (1988), 189-237; M. Loewe and E. Shaughnessy (eds.), The Cambridge History of Ancient China, Cambridge: Cambridge UP, 1999; Cho-Yun Hsu and Katheryn Linduff, Western Chou Civilization, New Haven: Yale UP, 1988; Herrlee Creel, The Origins of Statecraft in China, I: The Western Chou Empire, Chicago: Chicago UP, 1970, esp. ch. 11; Li Xueqin, Eastern Zhou and Qin Civilizations, New Haven: Yale UP, 1985; Mark Lewis, Sanctioned Violence in Early China, New York: SUNY, 1990; Derk Bodde, ‘Feudalism in China’, in Coulborn, Feudalism in History, 49-92. 21 Whereas Japanese feudalism, more or less comparable to the European, was postulated by early Japanese scholars and recognized by their Western peers (e.g. Hintze, Bloch, Ganshoff, Coulborn), more recent scholarship on Japan has become increasingly skeptical regarding the European and even feudal analogy, paradoxically, as European feudalism itself has come under increasing criticism. What has actually been shown, however, independently in both cases, is that in both of them the growth and formalization of feudalism was more complex and uneven than earlier held, which, if anything, would seem to reinforce rather than weaken the analogy. For Europe see n. 1 above, and regarding nobles, miles, and knights e.g.: Georges Duby, ‘The Origins of Knighthood’, in his The Chivalrous Society, London: Edward Arnold, 1977, 158-70; Pierre Bonnassie, >From Slavery to Feudalism in South-Western Europe, Cambridge: Cambridge UP, 1991, esp. 195-217. For Japan, in addition to the authorities cited in n. 15, see: William Farris, Heavenly Warriors: The Evolution of Japan’s Military, 500-1300, Cambridge Mass.: Harvard UP, 1992; Karl Friday, Hired Swords: The Rise of Private Warrior Power in Early Japan, Stanford: Stanford UP, 1992. 22 Montesquieu, The Spirit of the Laws, 30:1. 23 Bloch, Feudal Society, 441-7; Ganshof, Feudalism, xvi-xvii. 24 Voltaire, Fragments sur quelques révolutions dans l’Inde, cited in Bloch, Feudal Society, 441; Max Weber, The Agrarian Sociology of Ancient Civilizations, London: Verso, 1998 [1909], 38-9; extensively developed in his Economy and Society, 255-62, 1070-85; also Runciman, A Treatise of Social Theory, ii. 158. Runciman’s awareness that the state’s ability to maintain a central tax system and non-feudal troops constituted the key to resisting feudalism makes his study of particular cases more circumspect; yet he remains with a binary, feudal/non-feudal, frame. 25 Hintze, ‘Wesen und Verbreitung des Feudalismus’, and Coulborn and the other contributors to Feudalism in History come close to the same view. 26 Marxists have found China confusing in this respect. For a survey of Chinese and Soviet Marxist literature see: Arif Dirlik, ‘The Universalization of a Concept: From “feudalism” to “Feudalism” in Chinese Marxist Historiography’, in T. Byres and H. Mukhia (eds.), Feudalism and Non-European Societies, London: Frank Cass, 1985, 197-227; Colin Jeffcott, ‘The Idea of Feudalism in China and Its Application to Song Society’, in Leach, Mukherjee, and Ward, Feudalism, 155-74; Derk Bodde, in D. Twitchett and Michael Loewe (eds.), The Cambridge History of China, i. 22-3. For a recent Marxist example see Li Jun, Chinese Civilization in the Making, 1766-221 BC, London: Macmillan, 1996. 27 This is excellently pointed out in Chris Wickham, ‘The Uniqueness of the East’, in Byres and Mukhia, Feudalism and Non-European Societies, 172-5; also Bodde in Coulborn, Feudalism in History, 83-92.

20

28 Again, cf. Coulborn and the authors of the individual case studies in Feudalism in History, beginning with Burr Brundage, ‘Feudalism in Ancient Mesopotamia and Iran’, 93-119. 29 Most of the evidence comes from the vassal state of Arrapha rather than from Mitanni itself: Timothy Kendall, Warfare and Military Matters in the Nuzi Tablets, Ann Arbor: University Microfilms, 1974; also, Annelies Kammenhuber, Hippolgia Hethitica, Wiesbaden: Harrassowitz, 1961; id., Die Arier im Vorderen Orient, Heidelberg: Winter, 1968; Manfred Mayrhofer, Die Arier im Vorderen Orient – ein Mythos?, reprinted in his Ausgewählte kleine Schriften, Wiesbaden: L. Reichert, 1979, 48-71; Gernot Wilhelm, The Hurrians, Warminster: Aris, 1989. 30 Alan Gardiner, The Kadesh Inscriptions of Ramesses II, Oxford: Oxford UP, 1960, P80, P130-5, P150-5. 31 Michael Beal, The Organization of the Hittite Military, Heidelberg: Winter, 1992. 32 Alan Schulman, ‘The Egyptian Chariotry: A Reexamination’, Journal of the American Research Center in Egypt, 2 (1963), 75-98; id., Military Rank, Title and Organization in the Egyptian New Kingdom, Berlin: Hessling, 1964, esp. 59-62. 33 Kendall, Warfare and Military Matters in the Nuzi Tablets; Roger O’Callaghan, ‘New Light on the maryannu as Chariot-Warriors’, Jahrbuch für kleinasiatische Forschung, 1 (1950), 309-24; A. Rainey, ‘The Military Personnel at Ugarit’, Journal of Near Eastern Studies, 24 (1965), 17-27; H. Reviv, ‘Some Comments on the Maryannu’, Israel Exploration Journal, 22 (1972), 218-28; Michael Heltzer, The Internal Organization of the Kingdom of Ugarit, Wiesbaden: L. Reichert, 1982, esp. ch. 6 and pp. 111-5, 127, 192-4; Robert Drews, The End of the Bronze Age: Changes in Warfare and the Catastrophe ca. 1200 BC, Princeton: Princeton UP, 1993, 104-13, esp. 112. 34 J. Postgate, Taxation and Conscription in the Assyrian Army, Rome: Biblical Institute, 1974, esp. 208-11; Stephanie Dalley, ‘Foreign Chariotry and Cavalry in the Armies of Tiglath-pileser III and Sargon II’, Iraq, 47 (1985), 31-48. 35 Herodotus i. 96, 98, 101, 103, 123, 127; I. Diakonoff, ‘Media’, in The Cambridge History of Iran, ii. 36-148. Also, Muhammad Dandamaev and Vladimir Lukonin, The Culture and Institutions of Ancient Iran, Cambridge: Cambridge UP, 1989, 55. For the problems and interpretation of the scant evidence see Helen Sancisi-Weerdenburg, ‘Was There Ever a Median Empire’, Achaemenid History, 3 (1988), 197-212. 36 J. Cook, The Persian Empire, London: Dent, 1983, esp. 53, 101-12; Dandamaev and Lukonin, esp. 138-52, 222-34. 37 E. Yarshater (ed.), The Cambridge History of Iran, iii. chs. 2, 4; Josef Wieshöfer, Ancient Persia: From 550 BC to 650 AD, London: Tauris, 1996. 38 The contributions by H. Roemer and R. Savory, in P. Jackson and L. Lockhart (eds.), The Cambridge History of Iran, Cambridge: Cambridge UP, 1986, vi. 264-6, 344, 363-7. 39 This difference between ‘occidental’ and ‘oriental’ feudalism has already been stressed by Max Weber, Economy and Society, 259-62, 1073-77. 40 John Haldon, ‘The Feudalism Debate Once More: The Case of Byzantium’, The Journal of Peasant Studies, 17 (1989), 5-40; id., ‘The Army and the Economy: The Allocation and Redistribution of Surplus Wealth in the Byzantine State’, Mediterranean Historical Review, 72 (1992), 133-53 (which runs close to my own thesis); id., ‘Military Service, Military Lands, and the Status of Soldiers: Current Problems and Interpretations’, Dumbarton Oaks Papers, 47 (1993), all reprinted in his State, Army and Society in Byzantium, Aldershot: Variorum, 1995, chs. iv, vi, vii.

21

Also, Ernst Kantorowicz, ‘Feudalism in the Byzantine Empire’, in Coulborn, Feudalism in History, 151-166; E. Patlagean, ‘”Economie paysanne” et ‘feodalite byzantine”’, Annales E.S.C., 30 (1975), 1371-96; Hélène Ahrwiler, ‘La “pronoia” à Byzance”, in Structures féodales et féodalisme dans l’occident méditerranéen, Paris: CNRS, 1980, 681-9; Treadgold, Byzantium and Its Army, esp. 23-5, 171-9; Mark Bartuisis, The Late Byzantine Army: Arms and Society, 1204-1453, Philadelphia: U. of Pennsylvania, 1992, esp. 157-60, 164-5. 41 C. Bosworth, ‘Recruitment, Muster, and Review in Medieval Islamic Armies’, in V. Parry and M. Yapp (eds.), War, Technology and Society in the Middle East, London: Oxford UP, 1975, 59-77; Claude Cahen, ‘Technique et organisation socio-militaire dans le monde musulman “classique”’, in Structures féodales et féodalisme dans l’occident méditerranéen, 691-8; Hugh Kennedy, ‘Central Government and Provincial Elites in the Early “Abbasid Caliphate”’, Bulletin of the School of Oriental and African Studies, 44 (1981), 26-38; Patricia Crone, ‘The Early Islamic World’, in K. Raaflaub and N. Rosenstein (eds.), War and Society in the Ancient and Medieval Worlds, Cambridge Mass.: Harvard UP, 1999, 309-332. In his The Armies of the Caliphs: Military and Society in the Early Islamic State, London: Routledge, 2001, 59-95, Hugh Kennedy claims that the iqta was of marginal significance and not conditioned on service and that the Caliphs kept their armies on payment from tax revenue; the evidence, however, is far from clear and, in any case, things changed with the decline of the Caliphate. 42 Abdul Karim Rafeq, ‘The Local Forces in Syria in the Seventeenth and Eighteenth Centuries’, and M. Yapp, ‘The Modernization of Middle Eastern Armies in the Nineteenth Century: A Comparative View’, in Parry and Yapp, 277-307, 330-366 (esp. 343-56); Nicoara Beldiceanu, ‘Le Timar dans l’état ottoman’, in Structures féodales et féodalisme dans l’occident méditerranéen, 743-53. Again, my argument regarding ‘semi-feudalism’ in the lands of Islam largely agrees with that of Wickham, ‘The Uniqueness of the East’, in Byres and Mukhia, Feudalism and Non-European Societies, 175-82. 43 Daniel Thorner, ‘Feudalism in India’, in Coulborn, Feudalism in History, 133-150; the contributors to Byres and Mukhia, Feudalism and Non-European Societies; and the contributors to Leach, Mukherjee, and Ward, Feudalism: Comparative Studies. 44 E.g. Guy Bois, The Crisis of Feudalism: Economy and Society in Eastern Normandy c. 1300-1550, Cambridge: Cambridge UP, 1984. 45 The literature on all this is vast. For the military aspect see e.g. Philip Contamine, War in the Middle Ages, Oxford: Blackwell, 1984, 77-101, 115-8, 150-72; Michael Prestwitch, Armies and Warfare in the Middle Ages: The English Experience, New Haven: Yale UP, 1996; Maurice Keen (ed.), Medieval Warfare, Oxford: Oxford UP, 1999, chs. 6, 7, 10, 13. 46 See Montesquieu’s classic formulation in The Spirit of the Laws, 1:4; and John Hall, Power and Liberties: The Causes and Consequences of the Rise of the West, Oxford: Basil Blackwell, 1985, 42, with respect to China. 47 Duus, Feudalism in Japan, 67ff; Hall, ‘Feudalism in Japan’; Ryosuke, ‘Japanese Feudalism’; Perry Anderson, Lineages of the Absolutist State, London: NLB, 1974, 413-20, 435-61; S. Eisenstadt, ‘Tokugawa State and Society’, in his Japanese Civilization: A Comparative View, Chicago: U. of Chicago, 1996, 184-218. 48 Lewis, Sanctioned Violence in Early China, 61-5; id., ‘Warring States: Political History’, in Loewe and Shaughnessy, The Cambridge History of Ancient China, ch. 9; Twitchett and Loewe, The Cambridge History of China, vol. i; also Hans Bielenstein,

22

The Bureaucracy of Han Times, Cambridge: Cambridge UP, 1980; and for the horse bureaucracy under the T’ang: Jacques Gernet, A History of Chinese Civilization, Cambridge: Cambridge UP, 1996, 248-51; Denis Twitchett, ‘Tibet in T’ang’s Grand Strategy’, in H. van de Ven (ed.), Warfare in Chinese History, Leiden: Brill, 2000, 135-6.