Featured post

More Posts you might have missed on the other site

Here, in order from oldest to most recent are the not-exactly-numerous posts that have appeared on the other site in the past two and a half...

Tuesday, 4 June 2013

The Crisis of the State

[Here is the paper I pre-circulated for a really interesting conference called ‘Worlds in Motion: Rome, China and the Eurasian Steppe in Late Antiquity’ held last week at the Institute for Advanced Study in Princeton.  My great thanks to Michael Maas and Nicola di Cosmo for the invitation to attend.  For the doubtless hundreds of you who have been following this, it's the latest stage in my thinking about the issue of the state in late antiquity.  This begins with the same old same old about the decline of governmental apparatus and the rise of the aristocracy, but then moves on to address other, perhaps more subtle issues.  It comes to the same conclusion in the end, that to call western polities after 600 'states' is to rob the word state of any analytical value.  In discussion it was pointed out, possibly - probably? - rightly, that to concentrate on whether something was or wasn't a state distracts attention from important issues, but as there have been many works recently arguing that early medieval kingdoms were states, I am worried that this isn't an attempt to stifle opposition to the ... consensus.  Now read on.]


(I am aware that this is all very provisional and probably quite disjointed.  Some of the confusion may have righted itself in my mind by the end of May; some may not.  I crave indulgence for all that and look forward to your thoughts!  As this work stems from an on-going project, many of my references are to unpublished, unreferenced texts of lectures that I’ve posted on my blog.  Other referencing is very rudimentary but (I hope) indicative.  Apologies for those rough edges, too.)

For the last few years I have been looking at changes that took place in western Europe between the last third of the sixth century and the first half of the seventh.  This was a crucial period – as indeed it was in the eastern Mediterranean, as has long been known – but the fission of western European history into different national histories following the rupture of the unifying  grand narrative of Roman imperial history means that this change is less widely appreciated there.  Nonetheless, important change there was, recognisable in all the different regions of the West, in the different thematic types of history – economic, religious, social, religious, intellectual, military, political &c. – not only in the images that our sources present but in the forms which the evidence itself takes and in the quantitative extent and geographical spread of its survival.  

Clearly something major was going on, even if the details, the responses, and the trajectories of development were different from region to region.  A key aspect is change in the nature of political structures.  At least in the first quarter of the sixth century, perhaps the first third, up to the start of Justinian’s wars, conveniently beginning in 533, western European politics were still played out within an imperial framework – perhaps to an even greater degree than we have envisaged.

As a corollary, it is probably not surprising that government continued in recognisably Roman fashion.  For all that we are used to conceiving of them as ‘Germanic’ kingdoms in this period, it is very difficult indeed to find much that can cogently be called Germanic or even barbarian.  Certainly there were new elements in western rulership but these developed within a distinctively late imperial framework.  It is worth remembering how new kingship was and the extent to which it was being made up by political actors as they went along.  It may also be that even as late as the early sixth century it was not regarded as a permanent or even ideal solution to the problems thrown up by the fifth century, even by those occupying royal thrones.  This is not to suggest lumpen continuity of form, but a continuity of the tramlines within which change took place.  At least by the mid-seventh century, something important had changed in all of this.  

At the same time, changes occurred in the nature of social structures, most significantly with a growth in local aristocratic power.  Not unconnected to this were profound economic transformations.  Simultaneously, too, there were shifts in royal ideology and in aspects of religious life.  All these areas are closely interrelated and focus upon the interaction between central government and local power.  For those reasons I want to explore them under the rubric of the crisis of the state.  In terms of this workshop, the significance of this is that the change proposed will have little to do with the usually supposed catalysts for political change in the late antique West: the barbarian migrations and the western Empire’s political demise.  This means that it proposes a rather different chronology for change.  Traditionally, shifts in western political structures have been supposed to have occurred in the fifth century, with the migrations and the end of the Empire.  Realisation that drastic political change, as opposed to traumatic political events, did not happen then produced a counter-view of continuity across the period, perhaps right through to the Carolingian period.1  Similarly, in a Frankish context, old ideas of major transformation around the mid-eighth century with the dynastic change from Merovingians to the Carolingians have been replaced with an awareness of continuities across that divide.2  However, the idea of long-term continuity or slow evolution seems to me to be equally unsatisfactory.  I hope to demonstrate that something pretty dramatic did happen between the Fall of the West and the Rise of the Carolingians, but not when, or in the way that, we have been used to thinking.

There will be a weak and a strong thesis to this paper.  The weak thesis is that a crisis of the state occurred around 600; that changes took place which compelled a real shake up in the ways in which central and local power interacted, a critical moment which, whether or not it did, at least could have produced a breakdown of the state.  A supplementary to the weak thesis is that these changes killed off the ‘Roman World’ that is still so visible in, say, 525-30.3  One might entitle this ‘weak thesis’ ‘the end of the late antique state in the West’.  The ‘strong thesis’ is that the result of these changes was the end of political formations that can usefully be analysed as states in any way.  With a slight but important change in the word order, the strong thesis can be entitled ‘the end of the state in the late antique West’.

Theory (1)

Marxist and non-Marxist, Weberian definitions of a state converge on key issues.  Two examples from consciously opposed theoretical camps may serve to prove this.  First of all, Michael Mann:

The state is a differentiated set of institutions and personnel embodying centrality, in the sense that political relations radiate outwards to cover a territorially demarcated area, over which it claims a monopoly of binding and permanent rule-making, backed up by physical violence.4

Mann, of course, also argued that a state had to control all four of his sources of social power: ideological, economic, military and political.  He also, interestingly, completely skipped over the period that concerns me today.  Indeed, as far as I can see the whole 5-600 years of the early middle ages constitute the only era of recorded human political history in Europe that he did skip.  Nevertheless, he does say that some post-imperial states existed but that they were small and short-lived.5  I contend that he is wrong on both of these counts, certainly for the period before 600, less so thereafter.

Secondly, John Haldon’s, from The State and the Tributary Mode of Production, a Marxist work written explicitly to counter Mann’s modified Weberianism:

[A state is] ‘a set of institutions and personnel concentrated spatially at a single point and exerting authority over a territorially distinct area.6

To take a third example, Chris Wickham’s definition of the state turns on five things: 7 
  1. The centralization of legitimate enforceable authority (justice and the army)
  2. The specialisation of governmental roles with an official hierarchy which outlasted the people who held official position at any one time;
  3. The concept of public power …;
  4. Independent and stable resources for rulers;5. A class-based system of surplus-extraction and stratification

This might broadly be acceptable as a definition.  Such a definition tallies reasonably well with those of other thinkers, including those who work on the middle ages, such as Susan Reynolds.8  It is broad enough to encompass a range of state forms, but also strict enough to rule out other forms of complex political organisation.  Yet, if the definitions given by Mann, Haldon and Wickham are uncontroversial, they apply badly to western Europe after c.600, as I hope to demonstrate.  

More recent work on putative early medieval states has moved too far in the opposite direction.  The success of early medieval regimes in ensuring that local holders of power bought into their legitimacy surely produced politically coherent kingdoms or polities (to use more descriptively neutral terminology) but a politically coherent polity, I would contend, is not necessarily a state.9  Nor is the eternal retreat of the term down the political chain of command – to lordships, logically, I suppose, ultimately to Germanic Hausherrschaft or Roman patria potestas – a solution.  The fashionable consensus view of medieval politics suffers from many problems, not least a lack of any sophisticated analysis of the reactionary and repressive work that the word consensus does, then and now.10  This cosily politically-conservative view simultaneously negates the violence involved in lordship and the violent competition that is all-too-visible in royal and aristocratic politics.  Ironically, although this view glosses over ideological implications of the word consensus it fails to address the actual weakness of political centre; it fails to see that the extent of negotiation involved in dealing with local power circuits speaks of a degree of governmental weakness that surely prevents any analytically apt use of the term state to describe early medieval kingdoms.  

There is also a simple materialist, economic point to make in the context of the early middle ages and that is that surplus and the means of generating it were finite.  Surplus came overwhelmingly from the land, generated by the finite resource of human and animal muscle power.  Land was finite, obviously, and technologies for making economically marginal landscapes agriculturally productive did not yet exist.  Nor did means of improving the agricultural yield of estates, even if yields may not always have been as knife’s-edge as we have often thought.11  The economy was largely non-monetary, keying rewards for political and other service and the means by which elite status was maintained to the produce of designated lands.  Commerce was small-scale, overall and there was no system of banking or state credit.  Thus in some sense the grant of surplus from one person or institution to another did in some sense mean the loss of some part of the basis of power and its transfer to someone else.  In this sense the old ‘zero-sum’ idea of the relationship between rulers and aristocrats has some fundamental merit and this is why the nature of such transfers and the articulation of the extraction of surplus are the location for so much of the dynamics of early medieval political change.12  The first area of change I want briefly to describe thus concerns the demise of what might be termed state apparatus and the growth of the power of the local aristocracy.

Evidence for Change (1): The Efficiency of State Apparatus

Some words on regional variation

In describing the changes that took place across the last third of the sixth century and the first half of the seventh I focus on Gaul/Francia, the area I know best.  There were of course important differences in detail from one area to another, which I cannot do justice to in the space available, even where I am aware of them!  At a higher level of analysis, though, it is also clear that trajectories of change were not identical everywhere.  

The essential outlines seem similar in England and in northern Gaul.  There are important similarities in Spain too although much remains to be clarified.  (I hope that I may have been able to clarify my own thoughts a little more by the time of the workshop…)  The traditional debate has been over the supposed weakness of the later Visigothic kingdom, an image that ought to fit the general outlines of the thesis I am proposing.  Nonetheless thirty years ago Roger Collins put forward a very convincing counter argument that seventh-century Spain should be regarded as a strong and centralised kingdom.  It may be that the local power and independence of the later Visigothic aristocracy was rather less impressive than we think and thus the dynamic of their relationship with the royal court somewhat different from what is often envisaged.  Similarly, in Lombard Italy something of a rethink of traditional ideas may be necessary.  By the eighth century the Lombards had created an impressively centralised kingdom in Italy but it is difficult to project this image confidently back to the late sixth or the start of the seventh.  The written evidence upon which it is based is almost entirely eighth-century.  Most of the seventh century lacks decent documentary records and – as in Spain – the archaeological data may paint a picture that is rather different.  

A strong possibility at a macro scale, therefore, is that there was a different trajectory of development in southern and northern Europe.  Through much of the sixth century, patterns of trade around the Mediterranean, however they may have declined in scale, remained broadly within a late Roman framework; certainly there was more long-distance commerce than is visible in the north-west of Europe at that time, and the exchange network reached round as far as the eastern shores of the Irish Sea.  Economic changes around 600 produced a change in patterns and a decline in the Mediterranean system, while long-distance trade and economy revived around the North Sea.  Those economic shifts might well be an important part of any explanation for different socio-political trajectories in north and south.  Although real economic decline in the Mediterranean seems to me only really to set in, and the contrast between North Sea and Mediterranean economies to become starkest, rather later, at the end of the seventh century, the seeds of that change may belong to my period.13

Another interesting development in the northern area is a break-down in the differences between regions that had once formed a part of the Roman Empire and those which had not.  Ironically this may be related to ideas about the end of the Roman world and the analogous effects this had on regions inside and outside the old limes.14  This is a point that has emerged as the result of some research into what is now Scotland, and the development of Pictish society and politics.15  One way in which the old differences were eroded is in the survival of evidence.  In the sixth century, large swathes of former barbaricum lack diagnostic archaeological evidence as well as written sources.  In the seventh, this changes dramatically.  Here, then, it might be that the general outlines of social and political development were more similar to those in my geographically and analytically central case-study of Gaul.

‘State Apparatus’

There might be something of a chicken-and-egg relationship between the growth of aristocratic power and the decline of the old state apparatus but I will talk about the latter first and then move on to other evidence for the increase in aristocratic authority in the localities.

First, tax.  How sixth-century tax was levied is a difficult matter.  Much of the West, especially north of the Loire, was effectively non-monetary; everywhere north of the Pyrenees, outside Marseille, lacked small denomination currency.  A regular standing army and large central bureaucracy, the recipients of much of the late imperial government’s taxation in money or in kind, no longer existed.16  In a fifth- and sixth-century context, though, the point of taxation may have been less about revenue than about patronage.  That is to say that most tax remained in the areas where it was levied, as a salary for royal officers there.  Whatever was levied as bullion, possibly in the form of old coins, might have been passed on to the centre.   Nonetheless, the fact was that, however revenue was raised and whatever it was used for, post-imperial governments taxed.  That is crystal clear from the sources, which not only mention taxes and revolts against raises in the tax-rate or what was held to be against unjust taxation, but also refer to tax-lists.  It has also long been known that some late Roman taxes have lineal descendants in the dues levied by the lords of ninth-century estates.  Other – lesser – duties and obligations that the king could still call upon, at least in theory, at that date also seem to derive from late imperial taxation.  In 1982, Walter Goffart proposed that, as the lords had long since appropriated the more important revenues, immunities from these minor revenues and obligations were all that was left to ninth-century kings to reward their followers.  Goffart has subsequently adopted a more extreme position, whereby all late Roman taxation continued without a break into the ninth century.17  This aligns him with French historians like Elisabeth Magnou-Nortier and Jean Durliat but the argument for extreme fiscal continuity remains a minority position.  

Goffart’s 1982 argument was more persuasive, especially in its contestation that taxation itself withered around 600 in Gaul.  If there was any post-imperial taxation in Britain, and there’s really no good reason to suppose that there wasn’t (other than the silence of the documentary record, a record so nearly non-existant that it is silent about just about everything), then similarly, by the time one gets to the earliest surviving written data, forms of revenue have clearly followed the same path, with more or less the same survivals, as in Gaul.  In the earlier seventh century, references begin to immunities, into which royal officers could not enter to collect whatever dues remained to the fisc.18  The same seems to be true in the north of Italy in the Lombard kingdom, although we could hardly claim to be well-served by relevant evidence.  Paul the Deacon’s story about the distribution of the northern Italian Roman population among the Lombards as tributarii might be an index that taxation was still in operation or it could be a sign that Italy was already seeing a development in the nature of military service that is better documented in Gaul very slightly later.19  Given the political circumstances, the latter may well be preferable.  The fate of taxation in Spain by the seventh century is an area where I need to do more research.  Explaining this change means dealing with more than the practical difficulties of collecting revenue in kind, especially given that taxation appears to have withered precisely when something of an economic revival took place in north-western Europe, including the reintroduction of coinage.  

Another crucial factor, well evidenced in all areas under discussion, is the spread of a political/ethnic identity hitherto associated with military service and at least partial tax-exemption, to more or less complete equivalence with legal free status.  By this I mean Frankish identity in northern Gaul, Gothic in Spain, Lombard in northern Italy and English in lowland Britain.  The issue of identity is one to which I shall return.

The allusion to military service brings us to another principal means by which late antique states impinged on the lives of their subjects, enabling a central government to ensure that its authority penetrated local communities.20  The standing army withered in western Europe in the fifth century, at least in its usually understood, Roman, form.  Nevertheless, that being said, later fifth- and sixth-century armed forces had clearly developed from the last imperial armies.  As just intimated, they were founded around the levying of a group of privileged freemen whose status was probably inherited, or inheritable, and based on a real, or claimed, ‘barbarian’ ethnicity.  This situation pertains in almost all of the areas where we have clear written evidence and probably applied in those where we don’t, if one can extrapolate from the symbolism of artefacts placed with the male dead in areas like Anglo-Saxon England.  As far as one can tell (in the areas with written evidence), these individuals were called out within administrative districts, civitates and commanded by royal officials.  Because of the importance of their patronage and the access to power that military service provided, Kings continued to be able to use these armies as independent coercive forces against recalcitrant aristocrats and other rebels.  

As far as the Frankish kingdoms are concerned this means of raising an army had more or less completely eroded by the mid-seventh century.  By that date it seems clear that armies were being raised from aristocrats and their clienteles.  Instead of being described as drawn from particular administrative units, civitates, the elements of Frankish armies start being described as scarae – shearings off – and the implication seems to be that although the idea of military service as a general obligation had far from disappeared, it was now heavily moderated by ties of dependence and lordship.  A description of an army campaigning in Burgundy about a century later sums up the new situation ‘a multitude of magnates and a great band of their followers’.21  A similar process towards an army with aristocratic retinues has been identified in Gothic Spain and at this date Anglo-Saxon armies seem to be broadly similar.22  Indeed the military history of Anglo-Saxon England runs along a very similar course to that of early medieval Francia.  The heads of these bands do not command simply by virtue of royal office, as had been the case in the sixth century, but from their socio-economic standing.

At about this time, the immunities alluded to earlier are much more frequently documented, in forms that apply to aristocratic or church estates.  Kings envisaged areas into which their officers could not enter unbidden, either to collect revenue or, it would seem from the inclusion of the haribann in the list of things they weren’t allowed to collect, enforce military service.

The last area of state apparatus that I want briefly to discuss is the law.  Obviously, the Roman state had applied its law universally by means of its appointed officers, even while admitting the private jurisdiction of fathers, slave-owners and other lords over certain types of tenants, while admitting the jurisdiction of army courts and while syphoning off some classes of case to ecclesiastical courts.  The same system persisted through the sixth century, although episcopal courts may have broadened their competence and (in some areas) military jurisdiction morphed into the class of laws dealing with relationships between ‘Romans’ and ‘barbarians’.  Generally speaking, the state administered the law through its counts, centenarii and so on, with provision for appeals from one level of court to the next, possibly as far as the king himself.  The legal category of lèse majesté remained to create what we might consider as a ‘state of exception’.

Obviously the law has never applied – and still does never apply – evenly across all of the classes of a polity, whatever the theory may have dictated.  It was a commonplace under Rome that the rich man could buy himself all sorts of judicial rulings and influence court decisions where his interests were concerned.  Plus ça change…  Nonetheless it is interesting that seventh-century law shows a much greater concern with attempts by aristocrats to interpose themselves between the officers of the court and the cases they were trying.  Visigothic law mentions the disruption of court proceedings with various forms of ‘tumult’, sometimes specified to be carried out by the powerful patrons of one or other litigant.23  Frankish laws make special provision for the trial of a freeman in obsequio to another freeman (a relationship, by the way, not even recognised in sixth-century legislation) apparently aimed at preventing more powerful ingenui from sheltering their followers from the law.24  These may be no more than differences of nuance from earlier situations, if that.  Nevertheless they seem to point towards broader and more significant trends.  Again, the creation of immunities is very significant.  There were now formally-recognised gaps in the coverage of the law of the realm, no-fly zones, spaces where the royal writ did not run.  This seems to me to be something much more significant than a simple extension of patria potestas or Hausherrschaft.  If one is looking for the basis of what I am increasingly coming to think of the unworked, federalist nature of early medieval politics, it is probably to be sought in the immunity.

By way of concluding this section, one can argue that the aristocracy, had managed to position itself between the kings or the central government and the local population, in the extraction of surplus, the raising of armed forces and frequently the operation of the law.   It is clear that this did not affect the cohesion – or at least the basic unity – of the western kingdoms, north or south of the Pyrenees, for different reasons, but it does equally make it clear that there was a marked decrease in the ability of whoever controlled the centre of the realm to enforce his writ in the locales of the kingdom.  This decrease is marked in the actual institutions of the kingdom, not simply a contingent, practical difficulty.  Many states go through phases where their ability to make their writ run is seriously compromised, but without the formal apparatus of government changing.  

Aristocratic power

Again, my comments will focus on Gaul but, although this is not a generally-held view, I am increasingly of the opinion that the same general development took place in Britain and elsewhere in the north-west, inside and outside the former imperial limes.  These changes focus upon a tightening or intensification of local power networks.  As before, Spain remains a difficult region to categorise although some issues of the Italian situation may be less problematic.  In the latter area the direction of change could very well be quite the opposite of that in Gaul.

That there was an increase in the authority of local and regional aristocrats in northern Gaul around 600 is difficult to gainsay.  I have long argued for the point of view that sees the northern Gallic or Frankish aristocracy as essentially a service aristocracy without local bases of power that securely confirmed them in their élite status and allowed the transfer of the latter from one generation to another, independently of an association with the Merovingian state.  Written sources and various forms of archaeological data converge to give this impression.  Historiographically, this has long been a matter of debate and opinion remains divided.  There is no space to go into the details here.25  What matters is that there was a significant change around 600 (which one would have to concede whether or not one accepted my interpretation of it).  Legal evidence and changes in the archaeological record make clear the existence of a more stratified society with a more secure control over local surplus by the aristocracy.26  Other written (and some archaeological) sources suggest that this aristocracy was able to project its power more confidently into the future, to evade local legal custom and separate itself from the community in matters of burial.  The evidence for an economic upsurge and a revival of urban sites in this period underlines this, as – obviously – do the ways by which the aristocracy was able to interpose itself between the operation of the central government and the power circuits of local society.  Central to this change, as I have argued previously,27 was a change in the means by which surplus was articulated to support the position of the dominant class.  This is a change – in Wickham’s terms – from tax-based power to land-based, although it should be noted that my chronology for this change differs from his.  The relative local power of the southern Gallic – Aquitanian – nobility does not seems to have changed as drastically, although its relationship to central government may have shifted importantly in some of the same ways, as I will suggest further below.

In Anglo-Saxon England, the absence of fifth-and sixth-century written data makes precise discussion of the transformation difficult, but the archaeological record shows considerable similarities with that in northern Gaul and the documents that emerge in the course of the seventh century suggest that the changes brought analogous results to those in Francia.  The standard view of the period has tended to envisage the appearance of kingship in this period, but the evidence for such a supposition is weak and marred by too insular a focus.  It is as likely that what the decades around 600 witnessed was the fragmentation of larger political units into smaller, more intensive lordships.  In England some of these became known as kingdoms but their relationship to a larger political arena does seem to have remained.28  In northern Britain, changes in the archaeological record around 600 have also frequently been ascribed to a growth in the Pictish kingdom, a step on a road to a ‘Pictish state’, but again the material might rather reflect a shift from extensive kingdoms to smaller more intensively governed political units.29  Widespread changes well-documented in Ireland at about the same time might represent the final working through of a similar process.30

As noted, the nature of the Spanish aristocracy is less clear than we might have thought.  Wickham correctly pointed out the disjunction between the impression of wealth that one might obtain from the written sources and the rather different picture painted by the material cultural record.31  He also plausibly drew attention to some possibly misleading terminology in Spanish documents.  Spanish villas do not seem to have survived much beyond the start of the sixth century and it may be that communities became more closely grouped around churches which in turn were new foci for aristocratic display and authority.  It would nevertheless appear that aristocrats remained closely bound to their communities.32  An apparent spread of the, admittedly muted, custom of furnished burial might suggest that displays of status to local society were important in transmitting status from one generation to the next.  The difference between sixth- and seventh-century distribution of grave-goods burial might, however, be skewed by a concentration in the former period upon burials supposed to mark Gothic settlement.  Grave-goods burials without such material culture are more widespread across the peninsula.  Provisionally, I propose the fairly anodyne conclusion that the seventh-century Spanish aristocracy was significantly different in nature from its fifth- and possibly sixth-century precursors.  Its relative wealth and independence are difficult to establish but the sources suggest a number of analogous changes, during the seventh century, to those occurring in Francia.  The impossibility of hammering the Spanish situation into a neat framework of relative growth and decline may be a good way of introducing my second area of analysis.  Nonetheless, to conclude this first section, the evidence seems fairly clear that the ability of seventh-century governments across to the West to penetrate local society in order to extract surplus, maintain a coercive force and ensure that its writ ran was significantly curtailed by comparison with the sixth century.

Theory (2)

In a purely Gallic context my explanation long focused more or less exclusively upon the issues just discussed: competition for resources and authority between (to put it crudely) kings and their aristocrats, with the aristocrats winning out.  The catalyst for the change was the series of royal minorities in Gaul that began in 575.33  There might yet, descriptively, be something in this but the idea has more recently seemed to me to be inadequate.  Why should a loosening of the royal grip on patronage etc. automatically have led not only to aristocrats rushing to establish themselves more securely (which might seem reasonable) but also to changing their relationship with central authority so that the former operated less efficiently than before?  Roman aristocrats had been incredibly powerful without feeling the need for this sort of conflict or competition with the state.  My analysis began to shift towards a more ideological focus, to concentrate – again to borrow a phrase from Wickham – on how ‘the state lost consent’.34  Probably not coincidentally that led me to think more about Gramscian ideas of hegemony and the differences between political and civil society, ideology and so on.35  Ultimately that produced a greater concern to link material change to transformations in ideas.36    

The problem I encountered is well summed up by Mike Braddick in his discussion of the early modern English state.37  An emphasis on central power and force tends to play down the reality of states in much of history.  Central power inter-relates with other circuits.   It is this interaction that is crucial.  This is what recent work on the putative early medieval state has stressed but, to my mind, drawn unconvincing conclusions from.38  The issues remain of the degree to which local power circuits rely upon the central government for their articulation and functioning and the extent to which the central authority can over-ride local powers to impose its will.  This relationship is, of course, dynamic.  In the final analysis force is crucial39 but a concentration upon it is potentially misleading, as Braddick notes.  A state that must routinely employ force to ensure that its writ runs would today be classed as ‘failing’.  On the other hand, a state may not be able to employ its coercive force because of political difficulties, without ceasing to be a state.  If an army stands aside in the face of a political coup, or joins the rebels, we may witness failure a régime’s failure but not the end of a state, unless the institutions governing the army’s existence collapse with it.  This point has been central to numerous analyses, such as Althusser’s.40  

We must, therefore, not simply discuss changes in what one might think of as the ‘headline’ issues of state authority – fiscal imposts, military service, the operation of the law – but to look at more subtle, but perhaps also more important areas of interaction between the local and the central.  We must explore a slightly different range of issues, including many of those which have been, rightly, discussed by those recently wanting to adopt the terminology of the state as valuable for early medieval history.  These will include the spatial nodes of political action and their relationship to the government, office and of the legitimacy of local authority, political subjectivity and the character of the political community.  In all of these areas, I propose a major shift between the sixth and the seventh century which greatly changes how we might use the term ‘state’.  Indeed I contend that they make the term inappropriate.  The discussion has a certain contemporary resonance with what Rancière has described as the retreat of the political.41  The British Conservative Party has wedded itself to a process of replacing ‘the Big State’ with ‘the Big Society’ (no one quite knows what this means; essentially it is a form of post-industrial feudalism…).

Most of the issues mentioned above are fairly self-explanatory.  The last two, however, may require a little more by way of introduction.  By political subjectivity I mean the ways in which a socio-political actor is formed.  What (if any) part does the state play in the formation of the subject?  Closely related to this issue is that of the nature of the political community.  The existence of community has seemed enough to some analysts to justify the validity of the concept of the state.  However, as with the related term consensus this requires closer scrutiny.  In this context I will employ some thinking by Jean-Luc Nancy in his joint work (with Christophe Bailly) La Comparution and his La Communauté Désoeuvrée.42  I am not going to use this in any direct way, for many reasons.  It emerged in the aftermath of communism’s collapse in the context of the question of what the Left could retain of the communist project, and how.  Nonetheless, I will employ some issues arising from this discussion as a springboard to opening up some areas of debate apparently foreclosed by current discussions.  

One point of contact – one that increasingly strikes me as an important and fruitful line of investigation – is the renegotiation of terms of community in a period when long-standing bases for such have been swept away.  Another is the opening up and exploration of the interstices between consensus, integration and assimilation (politically-laden terms) and communal identity, and competition, coercion and alterity.  What attracts me in Nancy’s work is the deconstructive move involved in his notion of désoeuvrement, the opening of a space for an understanding of community in difference, community that is not the object of a work.  I am not going to push any of the ‘respect for the other’ that comes with Nancy’s comparution, even if there is use to be made of this.  The idea of political community not based around overarching identity or collectivity or perhaps even hegemony in the Gramscian sense, but on more contingent constellations of social and political actors, might bear exploration.  The attempts to create overriding identities to unite political communities is another dynamic of medieval political action.

Evidence of Change (2)

The spaces of the political

In the Empire and the post-imperial sixth century, the areas in which politics was enacted were places connected with the state, the imperial or royal court (or courts), the cities and other administrative nodes of the realm, peripatetic courts of presided over by royal officers, the army: unproblematically public spaces.  The cities and the lesser administrative centres (the vici, castra and castella) were the location for competitive aristocratic displays.  The episcopal network was, largely, based upon that of the administration of the late Roman Empire, as is well known.  Most church foundation in this period took place in the cities and smaller towns.  Estate churches were known but even monasticism was predominantly (not exclusively) an urban phenomenon.  Bishops fought hard (if not always very successfully) to preserve their prerogatives in the areas of preaching, baptism and so on.  The overwhelming majority of Christian cults were focused upon the cities, dominated by the bishops.  Rural cults, such as that of Saint Julian of Brioude were still intimately connected to the bishops and their cities (in this case Clermont).  Thus it is unsurprising that, outside the royal court/s, cities were the principal focus of politics.  Kings strove to maintain their ability to control these spaces, or their bishops.  Remember the famous complaint attributed (albeit problematically) to Chilperic I by Gregory of Tours, that all authority had passed to the bishops in their cities.43  This seems true of all formerly imperial areas.  Britain is usually cited as an exception but it is worth noting that we suffer from an absolute dearth of relevant evidence.  Roman cities could still have served as the foci for politics.  Canterbury, for example, appears to have been a royal centre at the end of the sixth century and York is attested in a similar role a generation or so later.  The high-status occupation of the baths basilica at Wroxeter in the post-imperial era hints at similar processes.   In the non-urbanised areas beyond the former limes it is difficult to draw any conclusions.  Comparing the distribution and nature of fortified centres north of the Antonine Wall, however, where fortified centres occupied are fewer and larger at first and then, around 600 become more numerous but smaller, possibly suggests that politics were largely enacted at the major centres of large kingdoms.  

A radical change in this pattern occurs around 600.  In various areas of the north-west, high-status rural sites appear.  Many of these focus on large halls, the political nature of which is well known from all sorts of early medieval literature.  More importantly, the seventh century, across the west, is the great moment of the foundation of rural churches and monasteries – from Ireland, via England and Francia, to Spain and Italy.  This is closely bound up with the processes we have encountered already, not least the creation of immunities and the increasing security of control over land.  The use of monastic foundations to cement local power and authority, to evade the demands of custom and the law, is attested all over Christian western Europe.  As is well known, familial rural monasteries became the political nodes of seventh-century Francia.44  The royal courts, naturally, remained crucial foci for political action and towns continued to play a role; indeed in the north-west this was a period which saw them begin to revive after perhaps three centuries of stagnation and decline.45  The extent to which they were political rather than economic foci might, however, be questioned.  The earliest northern Gallic charters do not suggest that towns were commonly used for the sorts of gathering that discussed transactions like land grants and so on.  For these, rural churches and monasteries or the centres of estates were more common, and that is even true of royal charters.  That is of true of England too, though the lack of earlier evidence makes it impossible to assess the extent to which this represented a change.  One might, then, suggest that, spatially at least, the circuits of local power ran through networks that were based upon the private – local aristocratic families, their lands, churches and other foundations – more than upon the public, administrative centres of the preceding era.

Identity and subjectivisation

It is absolutely crucial to consider the state’s role (if it had one) in the formation of social and political subjects.  Were identities related to, and formed by, a relationship with the state?  Two key areas are gender and ethnicity.


The Roman Empire had been a heavily gendered edifice.46  The very concept of masculinity had been, via its oppositions to barbarism as well as to the feminine, tied to the notion of being Roman.  What made a man superior to woman (or animal or barbarian) was also what made Roman government superior to that of the nations round-about.  The model of civic masculinity was, at least at the level of the social élites, tied into participation in government.  The idea that one was not suitable to govern also implied an inability to govern oneself and family (and vice versa) and called one’s masculinity into question.  I have argued that in the fourth century a rival, ‘martial’ model emerged in the Roman army, often in self-conscious opposition to traditional civic masculinity, stressing ferocity and other normally un-Roman characteristics.  What, however, legitimised and indeed permitted this model was its connection with the imperial government and especially the emperor.47  This has a close relationship with ideas of ethnicity in the immediately post-imperial era, to which I shall return in a moment.  The martial model steadily gained over the civic from the fifth century onwards but seems to have decisively won out – apparently quite suddenly – in the decades around 600.  

By the seventh century, outside the church, the warrior was the only real lifestyle choice open to the western aristocratic male.  I suspect that that comparatively sudden swing stemmed from the end of the possibility of relating both models to an involvement in a Roman (or sub-Roman) state, the situation that had held the two in relationship (or a dynamic tension) to each other.  I’ve previously discussed sixth-century gender as a curious moment wherein there were two poles of gendered social organisation arranged with reference to a shared, vanishing centre.48 The seventh sees (in my view) gendered restructured around a single masculine with femininity discussed with reference to its proximity to or distance from that.  Stress appears to be laid on male heads of families, a conclusion possibly supported by material cultural change as well as in the documentary record.  [An important variation here, however, might be suggested by some changes in later seventh-century Anglo-Saxon burials, with an emphasis on lavishly furnished female graves.]  It seems to me that the construction of masculinity no longer had any relationship to a function, position in or relationship to any particular political unit.  This point might be underlined with reference to the development of ethnic identity.


There was a well-known functional aspect to late Roman and immediately post-imperial ethnic identity, as already alluded to.  Romans paid tax and staffed the civil bureaucracy and Church; barbarians served in the army.  This can be seen clearly in sixth-century naming practices.49  The point underlines the linkage of identity to a place or role within the functioning of the state and indeed the construction of masculinity, as in the immediately post-imperial era, references to barbarian ethnic identities in relation to women seem very few.50  

There were numerous reasons for the spread of non-Roman identity in the later sixth century: legal privilege, tax exemption, and so on.  I suspect, too, that after the wars of Justinian Roman identity might have lost much of its political value.51  This is associated with the demise of the civic model of masculinity.  By the time that Lex Ribvaria was issued, all Romani had to have a Ripuarian (Frankish) freeman speak for them at law.  The category had gone from being a parallel free society to being a class of the semi-free.52  By the seventh century this sort of ethnic change was general in the west, as mentioned already.53

This change was closely linked to that in the raising of armies, which could no longer be carried out practically on the basis of barbarian identity.54  As noted, the process now took place ‘vertically’ down chains of dependence.  This itself meant that the role of military service in subject-formation was now frequently related to ties of personal dependence or allegiance rather than a role in the state.  I suspect that change in ethnicity worked simultaneously in opposing ways.  On the one hand, non-Roman identity was now so widespread among the free population that it could no longer serve to delineate successfully a particular socio-political group with a particular relationship to the government of the realm.  On the other, presumably connected to the fact that it was now shared by women and children and clearly inherited rather than achieved (as it might have been earlier), it was now personal, so that an individual (of a particular status at least) was able to claim the personality of the law.55  There was thus no necessary relationship between ethnicity and the polity or a function within it.  [All this – and much else – would be very different in eighth-century Lombard Italy but as I said earlier, the grounds for back-projecting such a picture to the early seventh or late sixth centuries are slim, and the eighth-century Lombard kingdom – state? – might have been a comparatively recent development.]

The one area where there was no such change (as yet) was southern Gaul, Aquitaine.  Yet that area may reveal what seem to be more general tendencies.  The sixth-century Merovingian state had operated in such a way that regional élites travelled to the north of Gaul to the courts of the Merovingian kings.  The kings themselves, although they travelled around, did so within a comparatively tight circuit (certainly by comparison with their better documented eighth- and ninth-century Carolingian successors) in the Paris basin.  They rarely strayed beyond this, with the exception of Guntramn of Burgundy’s occasional forays in the northern Rhone valley and Austrasian kings’ visits to the Rhine.  Seventh-century Frankish politics show an important shift, as regional élites increasingly demand a more locally-present ruler.  This is not the case in Spain, importantly, although developments in Septimania later in the century might have had analogous roots and the large, evidently semi-autonomous region controlled by Tudmir in the early eighth century could point a similar way.56  It is possible that similar tendencies might have lain behind the appearance of smaller realms in England57 and the drift out of effective Frankish hegemony of the regions right of the Rhine.  A similar evolution possibly took place in northern Britain.58

The political community

Thus, all in all it seems to me that the political had retreated from the public sphere of the state to the private sphere of the interaction of aristocratic families and their adherents.  This is not in any way to deny the fact that this interaction was still played out for the control of, and actions legitimised by a connection with, the royal court.  This ensured that the seventh-century kingdom continued to be a coherent structure.  Yet that kingdom seems not to be a political community, or it is not the working out of community.  Identity appears centred upon family and upon features unconnected with the realm: ethnicity, perhaps, and religion.  It may be this that enhanced the attempts to create community through marriage alliances.  Even here, though, the great aristocratic clans beloved of earlier generations of historians seem to be historiographical creations rather than necessarily effective political structures.  Contingency is the watchword of a seventh-century politics dominated by shifting alliances of aristocratic groups.  In Francia and elsewhere, the political community is ‘unworked’ in Nancy’s terms.  There is no over-riding shared identity that shapes it, or which it produces.  This may even be true in Spain, and marked by the on-going (yet ultimately unsuccessful) attempts to clothe the kingship in Christian and other legitimising ritual.  Deconstruction of the notion of community thus offers us a way of comparing developments in different areas without being side-tracked by ideas of coercion, taxation and so on.

Interpretation and conclusion

In considering what produced these shifts, my thinking has come to converge upon the importance of the wars of the mid-sixth century, Justinian’s ideology of reconquest and the fall-out from that for westerners.  There’s no space to discuss this in detail, but – in brief – one of the most important effects of Justinian’s wars was surely to bring home to anyone in the west that they no longer lived in the Roman Empire.  The emperor had made it very clear that their lands had been lost to the Empire and needed reconquering, and had waged long, bloody, destructive warfare in an attempt to do so.  The bases of all sorts of identities and authority – centuries-old patterns of thought based around traditional, Roman sources of power, ideology, and behaviour – were cut away, potentially projecting the West into a sort of semiological hyperspace.  It is no coincidence that – as has long been known – the late sixth and seventh century saw an important move from traditional classical to biblical, especially Old Testament exemplars.59  That in turn underpins all sorts of other changes, encompassed in my discussion.  For people accustomed, after three hundred years or more, to the idea that the Roman Empire was commensurate with the sixth age, the awareness of living after Rome understandably produced considerable anxieties.  This conjuncture of circumstances produced, or enabled, a springing apart of old identities and communities and their reconfiguring in forms that may have been entirely contingent and expected only to be short-lived, but which turned out to be a blueprint for medieval politics.


1 The various volumes of the Transformation of the Roman World project, published by Brill in the late 1990s and early 2000s were diverse in approach but it’s not unfair to say that many contributions advocated this sort of continuity approach.  See also, A.M. Cameron, The Mediterranean World in Late Antiquity.

2 Cp. P.J. Fouracre, Francia in the Age of Charles Martel.

3 Here there may be some points of contact with J.J. O’Donnell, The Ruin of the Roman Empire.

4 M. Mann, The Sources of Social Power, vol.1, p.37.

5 Mann, The Sources of Social Power, vol.1, p.390.

6 J. Haldon, The State and the Tributary Mode of Production, pp.32-33

7 C.J. Wickham, The Framing the Early Middle Ages

8 S. Reynolds, ‘The historiography of the medieval state’, in M. Bentley (ed.) The Routledge Companion to Historiography.

9 S. Airlie & W. Pohl (ed.) Staat im frühen Mittelalter.

10 E.g., above all, the oeuvre of Janet L. Nelson, e.g. Politics and Ritual in Early Medieval Europe; ead., The Frankish World.

11 I am grateful to Dr Jonathan Jarrett for advice on this point.

12 For critique of the zero sum model see T. Reuter & C.J. Wickham, ‘Introduction’ in W. Davies & P.J. Fouracre (ed), Property and Power in the Early Middle Ages.

13 E.g.. the chapters by S.T. Loseby & S. Lebeq in P. Fouracre (ed.) The New Cambridge Medieval History 1.

14 This point is discussed further at the end of this essay.

15 See below, n.29.

16 See, e.g. C.J. Wickham, The Framing of the Early Middle Ages.

17 W. Goffart, ‘Old and new in Merovingian taxation’, Past and Present 1982; cp. id. ‘Frankish military duty and the fate of Roman taxation’ EME 2008.

18 B. Rosenwein, Negotiating Space. Power, Restraint and Privileges of Immunity in Early Medieval Europe (Manchester, 1999).

19 Paul, HL, 2.32

20 Details of the changes in military organisation in this period can be found in G. Halsall, Warfare and Society in the Barbarian West, 450-900 (London 2003); I would now modify the account in some details.

21 multitudine primatum et agminum satellitum plurimorum; Fred. Cont., 24

22 D. Pérez Sánchez, El Ejército en la Sociedad Visigoda (Salamanca, 1989); R. Abels Lordship and Military Obligation in Anglo-Saxon England (London, 1988).

23 Visigothic Law 2.2.2, 2.2.9 both issued by Chindasvinth.

24 Lex Ribvaria 33.1

25 Wickham, Framing of the Early Middle Ages, is the most recent and sophisticated attempt to argue for the existence of a wealthy and independent Frankish aristocracy before the seventh century.  For the unreferenced draft of an article attempting to counter Wickham’s argument see: http://600transformer.blogspot.co.uk/2012/02/officers-or-gentlemen-frankish.html (with links to 3 subsequent instalments).  A problem with Wickham’s argument is that he does not take into account the lack of wealth and independence of the late Roman aristocracy in the region: http://600transformer.blogspot.co.uk/2011/11/genesis-of-frankish-aristocracy-part-1.html (another unreferenced draft; again there are 3 subsequent parts, linked from the first).

26 G. Halsall, Settlement and Social Organisation. The Merovingian Region of Metz (Cambridge 1995), ch.9.

27 G. Halsall, ‘From Roman fundus to Carolingian Grand Domaine: crucial ruptures between late antiquity and the middle ages.’ Revue Belge de Philologie at d’Histoire 90 (2012), pp.273-98. See also G. Halsall, ‘Villas, territories and communities in Merovingian northern Gaul’ in People and Space in the Middle Ages, 300-1300, ed. W. Davies, G. Halsall & A. Reynolds (Turnhout 2006), pp.209-31.

28 Some of the issues discussed here, not least the problems with the usually adopted thesis of kingdom formation, can be found in my recent (semi-academic) volume Worlds of Arthur: Facts and Fictions of the Dark Ages (Oxford, 2013).

29 G. Halsall, ‘Northern Britain and the Fall of the Roman Empire’.  The Mediaeval Journal 2.2 (2012), pp.1-25.  See also Worlds of Arthur; http://600transformer.blogspot.co.uk/2010/10/woad-less-travelled-archaeology-of.html Cp., for the currently accepted view, S. Driscoll, ‘Power and authority in early historic Scotland: Pictish symbol stones and other documents’, in State and Society: The Emergence and Development of Social Hierarchy and Political Centralization, ed. Gledhill, J., Bender, B., & Larsen, M.T., (London 1988), pp.215-36.

30 E. H.C. Mytum, The Origins of Early Christian Ireland (London, 1992) for description.

31 Wickham, Framing the Early Middle Ages.

32 See, e.g., the oeuvre of A. Chavarría Arnau.

33 E.g. G. Halsall, Settlement and Social Organisation.

35 A. Gramsci, Selections from the Prison Notebooks

37 M.J. Braddick, State Formation in Early Modern England, 1550-1650 (Cambridge 2000), ch.1 [I am enormously grateful to my student Laura Salvage for pointing me at this.].  For a view of the mainland European situation, stressing the dynamics from above and below of the state penetrating local society and which has been influential in the development of my thinking, see J. Glete, Warfare at Sea, 1500-1650. Maritime Conflicts and the Transformation of Europe (London, 2000); id., War and the State in early Modern Europe: Spain, the Dutch Republic and Sweden as Fiscal-Military States, 1500-1650 (London, 2002).

38 The most serious example is M.J. Innes, State and Society in the Early Middle Ages, whose argument is belied at every turn by basic factual errors and egregious misreadings of documents.  For (believe me!) mild critique, see Halsall, Warfare and Society, pp.78-81 and references.

39 See, e.g., the discussion of the state in R. Geuss, Philosophy and Real Politics (2008).

40 L. Althusser, On Ideology

41 J. Rancière, La Haine de la Démocratie

42 J.L. Nancy, La Communauté Désoeuvrée (trans. The Inoperative Community (Minnesota, 1991); C. Bailly & J.L. Nancy, La Comparution .

43 Gregory, Histories 6.46.

44 See, for example, R. Gerberding, The Liber Historiae Francorum and the Rise of the Carolingians (Oxford, 1987).

45 There is a long bibliography of studies of western urban and regeneration, e.g.: G. Brogiolo & B. Ward-Perkins (ed.), The Idea and the Ideal of the Town between Late Antiquity and the Early Middle Ages (Leiden 1999); G. Brogiolo, N. Christie & N. Gauthier (ed.), Towns and their territories between Late Antiquity and the Early Middle Ages (Leiden 2000); N. Christie & S.T. Loseby, (ed.) Towns in Transition: Urban Evolution in Late Antiquity and the Early Middle Ages (Aldershot, 1996); R. Hodges & B. Hobley, The Rebirth of Towns in the West, AD 700-1050 (CBA: London); 

46 J.M.H. Smith, ‘Did women have a transformation of the Roman world?, Gender and History; G. Halsall, ‘Gender and the end of empire’, Journal of Medieval and Early Modern Studies.

47 A point thoroughly and convincingly discussed in the work of M. Kulikowsky. E.g. Rome’s Gothic Wars From the Third Century to Alaric (Cambridge, 2006)

48 Again I can only offer a reference to a preliminary draft:  http://600transformer.blogspot.co.uk/2013/02/classical-gender-in-deconstruction.html

49 A famous example is the brother of Bishop Nicetius of Lyon, who bore the name Gundulf.  Gundulf was a domesticus and later a duke at the Austrasian Frankish court and presumably took (or rather added) this name on going into secular service just as his great-nephew, Gregory of Tours, added the name Gregory upon entering the church.

50 G. Halsall, Barbarian Migrations and the Roman West (Cambridge 2007), pp.485-7.  Mark Handley has drawn my attention to a couple more barbara on inscriptions since then but not enough to really change the picture.

51 See further, below in the conclusion.

52 Lex Ribvaria.

53 This makes me think that the references to the Welsh as a subordinate population in the early 8th century Laws of Ine of Wessex might similarly be the outcome of a longer process with a more important shift around 600 and thus have no necessary implication for the fifth and sixth centuries, as is often supposed.

54 I have added to the discussion of this in Warfare and Society in my chapter for the forthcoming Cambridge History of War, ed. A. Curry.  This volume has been forthcoming for some time, however, and I am unsure what its current status is.

55 E.g. Lex Ribv.

56 R. Collins, The Arab Conquest of Spain (Oxford, 1989) esp. pp.42, 149, 202-3.

57 Again, following the hypothesis outlined in Worlds of Arthur, ch.11.

58 See above, n.29.

59 This can be seen no more clearly than in the difference between Gregory of Tours’ Clovis and the Clovis who emerges from documents of his own time.  Though this is not the conclusion drawn, it is thoroughly illustrated by W.M. Daley, 'Clovis: How barbaric, how pagan?' Speculum 69 (1994):619-64.