From the very beginning, bureaucracy was a dirty word. The eighteenth-century economist, merchant, and official, Jacques-Claude Marie Vincent, Marquis de Gournay (also credited with coining the term laissez faire), complained that free trade in grain was being impeded by over-regulation. This “illness” of French government, he was later reported to have wittily remarked, was a kind of bureaumanie—the inevitable result of bureaucratie.1 Despite the efforts of sociologists and historians, “bureaucracy” as an analytical—or even descriptive—term has never quite escaped its polemical origins. (Late antique scholars might think here of “paganism” or “barbarians.”) A feeling of unease—and a sense that bureaucracy needs defending both as a useful category and as a positive institution—frames the best contributions to the excellent new collection, Empires and Bureaucracy in World History, edited and thoughtfully introduced by Peter Crooks and Timothy Parsons. The immediate interest for readers of Studies in Late Antiquity is to have Michael Whitby's useful contribution on the later Roman empire set alongside sixteen other chapters which examine a range of pre-industrial empires (the term is used loosely—the focus of this book is on understanding bureaucracy) including Song China, the Inca, the Ottomans, Byzantium, the Carolingians, and the medieval West, then moving through to Napoleonic Europe, British colonial India, and post-World War II Africa. The result is not (and not intended as) a continuous narrative; rather, Empires and Bureaucracy offers a set of often contrasting papers which encourage reflection on the assumptions of specialist historians of government at times too accustomed to the conventions of their own particular period.
Michael Whitby's essay (“The Late Roman Empire was Before All Things a Bureaucratic State”) takes its title, as he explains (129), from the opening sentence of chapter 16 (“The Civil Service”) of A. H. M. Jones, The Later Roman Empire 284–602: A Social, Economic, and Administrative Survey (3 vols, Oxford: Basil Blackwell, 1964).2 It opens with a sketch of the early empire as “government without bureaucracy” (130).3 “The Roman empire did not require such administrative complications because, overall, it worked” (133). “Amateurish” (130) and “somewhat gentlemanly” (146), “light-touch” (134) imperial rule was particularist, privatized, and piecemeal (134–135). It relied on local elites in a network of cities “internalizing Roman priorities” (131) or facing brutal reprisal. This is a familiar and widely accepted formulation, and one which relies heavily on culture (135) to do the hard work of empire in imposing some recognizable sense of “Romanness” on the Mediterranean world.
This “undergoverned” (130, 134, 139) state fractured in the third century as a result of military pressure along the northern frontier and from the Persians in the East. Recovery under the Tetrarchy and Constantine demanded “an intensification of central intervention” (137) in provincial affairs, key economic sectors, military supply, taxation, and religion. “The administrative personnel engaged in running the system expanded dramatically” (138). Rather than anatomize these officials and their departments (Jones's classic account of late Roman institutions still holds the field), Whitby—in the spirit of Empires and Bureaucracy—offers a defence of bureaucracy as more than a regrettable necessity.4 Bureaucracy strengthened empire (141). Along with court, army, church, and provincial and metropolitan elites, it “constituted one set of networks that helped bind the empire together” (144). It played a key role in ensuring the continued importance of traditional education (144). Nor should it too quickly be condemned as dysfunctional or corrupt. “There is clear evidence for buying office, nepotism and extraction of fees for services in the late empire, but it is wrong to assume that the urban administrations of the early empire were paragons of virtue (139) … The administrative system of the late empire was far from perfect: it certainly privileged wealth and connections, but it was not necessarily any less effective than what had pertained in the early empire” (140).
Like many before him, including this reviewer,5 Whitby emphasizes the decisive shift in the scale of imperial government in Late Antiquity. To quote Peter Brown: “The contours of the new situation were unmistakeable and are well known. ‘Soft’ government was replaced by ‘hard’ government.”6 In the first two centuries C.E., Whitby reckons “probably only between 100 and 200 high-level (senatorial and equestrian) positions, and only a few hundred formal positions in central and provincial governments,” as against 20,000 senior administrative personnel in the later Empire (138 n. 33). The figures are fragile. Whitby is right that “the numbers are impossible to quantify” (138). As a sighting-shot, it is worth noting that in the early empire there were perhaps as many as 10,000 slaves and seconded soldiers performing administrative tasks, to which add (in Whitby's pleasing phrase) the “much less visible ‘undergovernment’” in, for example, the cities (139). It matters too that the senatorial vision of government, so neatly exemplified by Pliny the Younger's Letters and Panegyric, deliberately emphasized direct engagement with the emperor. At times, the vast palace on the Palatine seems rather understaffed. It is salutary to be reminded (by Deana Heath on “Bureaucracy, Power and Violence in Colonial India: The Role of Indian Subaltans”) that British imperial administration, which also prided itself on its light touch, relied on an Indian Colonial Service of a mere one thousand (white) men, supported by a state-funded workforce of one million (365). No doubt too the estimates for those discharging administrative duties in the later Empire should also be increased. In the end, however these numbers should be worked out (such back-of-the-envelope calculations are perhaps always better thought of as metaphors), they offer some very rough measure of the change in the structure and scale of imperial government. They might suggest—in contrast to the Principate—that there is a plausible case to be made to support Jones's declaration that “the later Roman Empire was before all things a bureaucratic state.” As Whitby observes: “Jones's perceptions of the late empire were undoubtedly influenced by the nature of imperial control and organization in the earlier stages of the Roman empire” (130).
But it is a claim which seems much less convincing in any broader context. That is the chief lesson to be learned from reading the other essays in Empire and Bureaucracy. As a start, it is worth pointing out that the low level of bureaucracy (whatever the headcount) in the early Roman Empire is not in broader historical terms an outrider. It still demands explanation, but not as something remarkable. It is not a unicorn. It takes its place with other empires which like the Inca (to quote Chris Given-Wilson's fascinating essay7) did not attempt to achieve “the maximum achievable degree of control over the peoples whom they conquered, leading … to the maximum achievable extraction of tribute and exploitation of labour and other resources” (81). In contrast with these “proto-bureaucratic” empires (to adopt Whitby's useful term (134)), other imperial regimes have depended, to some greater degree, on: “routine administrative activity delegated to office holders (who are often, but not always, professional career administrators), conducted on the basis of records … with some differentiation and specialization of offices that are organized hierarchically and are reliant on systems of communications” (17–18).
Crooks and Parsons' working definition of bureaucracy deliberately moves away from Max Weber's much discussed “ideal type” with its emphasis on institutional independence, impartiality, rationality, legal fixity, and technical superiority: “The fully developed bureaucratic mechanism compares with other organizations exactly as does the machine with the non-mechanical modes of production. Precision, speed, unambiguity, knowledge of the files, continuity, discretion, unity, strict subordination, reduction of friction and of material and person costs – these are raised to the optimum point in the strictly bureaucratic administration.”8 The editors' stripped-down formulation trades some of Weber's complexity and sophistication against a welcome slackening of the ties (assumed or implied) between bureaucracy and modernity (18–20). Imperial bureaucracies which fall far short of the ideal type are not so immediately vulnerable to criticism as corrupt or inefficient. To quote Crooks and Parsons:
The case studies presented in this volume suggest that the norm was a more tangled imperial administrative structure and hit-and-miss bureaucratic process. Moving rather schematically outwards from the centre, the chapters suggest a general tendency towards a small and ramshackle organization at the imperial core; uncertain connections between metropole and colonies resulting in ‘negotiation’ rather than top-down imperative control; bureaucrats on the peripheries unable to penetrate too far without undermining their own authority (18).
Perhaps what matters is the extent to which empires have relied on bureaucracy with the aim of increasing extractive capacity or marshaling resources in frontier provinces, and how that strategy has played out against alternative/competing/conflicting means of promoting integration or coherence. In any case, it is clear from the essays in this volume that the relationship between empire and bureaucracy is a matter of delicate degree rather than a long, and frequently frustrated, march towards rationality and the nation-state. It is somehow comforting for ancient historians to learn from Frederick Cooper's essay (“From Chief to Technocrat: Labour and Colonial Authority in Post-World War II Africa”) that as late as the 1950s the colonial administration in French West Africa was able to maintain a register of birth and deaths only within ten kilometers of a recording center (408).
The more extensive bureaucracy promoted by the Song dynasty (960–1276 C.E.) is explored in Patricia Ebrey's essay (“China as a Contrasting Case: Bureaucracy and Empire in Song China”). A “remarkably well-organized and effective” (47) imperial administration was a key element in an effort—less successful in frontier regions (48)—to achieve empire-wide uniformity in the basic elements of government: law, taxation, currency, and education. Ebrey details a hierarchical and rule-bound system of 30,000 well-paid senior officials (subject to periodic internal review and regularly rotated between 350 prefectures), and 200,000 subordinates generating a mass of documentation (40–47). Especially noteworthy are the examination system (which used candidate numbers and had answer papers copied so that handwriting could not be recognized by examiners), the various rituals of official life (which “helped create a basic acceptance of the institutional hierarchy on the part of the educated class” ), and the extensive “treatises, histories, legal codes and institutional reference books [that] gave those both at court and in the provinces a way to think about problems as they arose” (45). These elements are recurrent themes in Empires and Bureaucracy. Particularly fascinating (to offer a very brief selection with late-antique historians in mind) are the discussions of handbooks for bureaucrats and the literature on statecraft in the early Arab caliphates (72–78);9 the civil and military administration developed under Charlemagne (181–194);10 the “precocious” and extensive record-keeping of the medieval English state and its French territories (the so-called Angevin empire) (197–199, 207–213, 264–273);11 the vast (and still surviving) documentation of the sixteenth-century Ottoman empire which underwrote a reforming push towards “the routinization of function, salary and rank, merit-based promotions and frequent rotation of office” (107–116);12 and Napoleon's creation of a cadre of loyal and highly-trained imperial administrators (356–362).13
It is difficult to argue that later Roman bureaucracy can easily match these examples in terms of (for a start) professionalism, record-keeping, or organization. Not even late antique texts like the Notitia Dignitatum—key to Jones's institutional reconstruction—bridge the gap. This splendidly illustrated, (likely) early fifth-century text presents detailed rank-ordered lists of administrative departments and troop dispositions in both East and West. But it is doubtful that the Notitia should be read as a straightforwardly descriptive handlist of later Roman bureaucracy. That would be to ignore its ideological concern to project a sense of order, security, and Mediterranean unity: “a reformation of the oikoumene in bureaucratic terms.”14 It is noticeable too that, from the point of view of those involved in government setting down their own ideals and practices, surviving fourth- and fifth-century literature seems to run thin in terms of administrative primers, reference manuals, or advice to officials. Similarly striking—at least as a matter of degree—is the extent to which later Roman bureaucracy was significantly shaped by the payment of money for access (reinforced by low official salaries), by shifting and overlapping responsibilities, by (as Michael Whitby notes) the “fluidity and the competition that existed between administrative rivals” (143), and by the sometimes unpredictable intervention of emperors seeking both to preserve their autocratic authority and to maintain strong centralized networks of imperial patronage. Even the commitment to record-keeping can be questioned. It should perhaps be more often observed that a consolidated and ordered (chronologically and by topic) collection of imperial edicts was not available until the Theodosian Code was issued in 437. In other words, for a century after the death of the Emperor Constantine later Roman administration—however its elaboration is to be envisaged—operated without an agreed, authorized, and available (if only to officials and judges) codification of imperial rulings. To quote John Matthews: “The Theodosian Code [was] the first fully official attempt since the publication of the Twelve Tables by a Roman government to collect its own legislation.”15
When set against the range of imperial governments surveyed in Empires and Bureaucracy, this brief, and knowingly abbreviated, survey ought at the very least to provoke hesitation or qualification in any description of the later Roman Empire as a “bureaucratic state.” Of course, comparisons need cautious consideration. It is all too easy for late antique historians to be wary and sophisticated readers of the Notitia Dignitatum, while taking contemporary claims of an efficient and well run administration in other societies too much on trust. The comparative historical enterprise is always likely to fail at some level of detail or critical expertise. But that does not necessarily undercut the importance of the impressions that remain. These might also perhaps suggest that those who focus on later Roman bureaucracy's deficiencies, as part of a wider tale of decline and fall, are, somewhat paradoxically, too optimistic in assessing its likely impact as a disruptive force. To quote Len Scales (on the “parchment imperialists” of the late medieval Holy Roman Empire, whose lack of a developed administration is sometimes cited as an index of the empire's weakness): “The proliferation or perpetuation of bureaucracy is no more a necessary symptom of an empire's health than limited institutions reliably indicate its impending collapse” (248).16 Likewise, to suggest, following Michael Whitby, that later Roman bureaucracy “was not necessarily any less effective than what had pertained in the early empire” (140) is to miss the point. It too swiftly sets aside both the differences in the structure, scale, and ambitions/demands of government, as well as the radically changed circumstances of imperial rule. To return to Chris Given-Wilson on the Inca: “it may well be the case that a more comprehensive (and hence probably more intrusive) bureaucracy was unnecessary and might well have proved counterproductive” (99). Would the early empire have been more “effectively” governed if Augustus, or Vespasian, or Trajan had developed a late antique style bureaucracy?
Whitby's conclusion is rather muted (and perhaps he would not disagree with too much of the above). “There was, however, no rapid or designed bureaucratization of government: arrangements were changed in different areas at different times in response to particular needs, with the result that Roman arrangements were characterized by a certain diversity and fluidity” (146). That would seem to stake out a position knowingly removed from Jones's claim (and Whitby's own chapter title): “the later Roman Empire was before all things a bureaucratic state.” To be fair, Whitby had walked back “before all things” in his opening paragraph: “one might quibble about the emphasis” (129). That reservation deserves an even greater distance. There is no doubt that The Later Roman Empire 284–602: A Social, Economic, and Administrative Survey is a monumental achievement of meticulous scholarship. It does precisely what it says on the can. It is principally and unapologetically a detailed account of late Roman institutions. It could have been read with pleasure by the officials who compiled the Notitia Dignitatum. Above all, this is a survey of “the later Roman Empire” (a hard, institutional expression) and not of a softer, more diffuse “Late Antiquity.”
From that point of view, Jones's neat encapsulation of the later Roman Empire as a bureaucratic state demands twofold revision. First, it is an unhelpful and misleading over-simplification. In a broad comparative context—as the intelligent and interesting essays in Empires and Bureaucracy in World History make clear—it overstates the contrast with the Principate and risks exaggerating the effective reach of late-antique government. As Whitby's conclusion suggests, in many respects the later Roman Empire (and perhaps especially in the fourth century) is not a particularly persuasive example of a bureaucratic state. Secondly, Jones's apothegm springs an institutional trap. As the best history-writing (of Late Antiquity and more widely) in the last generation has amply demonstrated, there is no good reason to assume—at least without hesitation or argument—that the history of the state should define a period. In the fifty years since Jones's The Later Roman Empire, the study of late antique institutions has perhaps been underserved, even so it seems unlikely that it will re-emerge at the core of the history of (to reclaim a term) the later Roman Empire. Whitby—particularly given his attractive emphasis on the significant role of court, army, church, and elites (144–146)—rightly retreats from his starting-point. Simply put, the raw claim that “the later Roman Empire was before all things a bureaucratic state” inflates both the importance of the state and the importance of bureaucracy to the state.