It is a great pleasure and an honour to be writing for the fortieth anniversary volume of Byzantine and Modern Greek Studies. As editor of the journal for some twenty years, from 1984 until 2004, I have watched the journal grow in stature and in esteem over that period, and I am delighted to see it continuing to do so in the hands of its current editors. In the first issue I edited, I also contributed an article that attempted to reconcile some very different approaches to the history of Byzantine society and culture, or at least, to show that such different approaches were not necessarily mutually exclusive. If now rather out-of-date in its content, that article remains a useful baseline for discussing the relationship between empirical research and writing and theoretical reflection.Footnote ¹‘“Jargon” vs. “the facts”‘? was a comment about the confrontation that at the time appeared to exist between, very broadly speaking, those who were interested in questioning the theoretical assumptions underlying and informing their research, and those who were not interested in such debates, preferring to see them either as irrelevant or as inaccessible. In my concluding remarks, I suggested that Byzantine Studies in the mid-1980s was in the process of what T. S. Kuhn would have called a ‘paradigm shift‘, that is to say, a process through which a traditional set (or sets) of assumptions and priorities, as well as theories and approaches, is replaced by different sets of ideas. While the changes in the nature of the subject that have occurred since then have not been particularly marked, there have nevertheless been some interesting and important developments that have altered the framework within which some ways of looking at the medieval eastern Roman world are carried on. The so-called ‘linguistic turn‘, for example, pushed Byzantinists, in particular, scholars of Byzantine literature and visual culture, to grapple with various aspects of what might very broadly be termed post-modernist and post-structuralist theory. This is evident in some of the writing and publishing of the later 1980s and 1990s in particular, and in some respects has now been incorporated into our ‘ways of seeing’ the Byzantine world.Footnote ² In particular issues of intertextuality, of authorial intention, of reception, and of the relativizing of cultural interpretive possibilities (in respect of our own perspective) have become part and parcel of scholarly discourse, thus greatly enriching our discipline.Footnote ³ Represented by more recent work in literary studies and art history especially, I believe this shift also facilitated a much greater degree of cross-disciplinary reading, comparative thinking, and in respect of historical context and setting, a generally more open approach to the medieval west and the Islamic world in terms of both material and method.Footnote ⁴

The result is that there are some fascinating discussions currently regarding both how we define our field, where it stands in relation to other historical studies,Footnote ⁵ how we should define our subject matter (is it appropriate to talk about Byzantium at all, or should we speak rather of medieval eastern Romans, and so forth)Footnote ⁶, and what it means to describe oneself as a Byzantinist, even whether that is a useful term at all. One question that has recently exercised some scholars is that of the chronological terms of reference of the subject, in particular, when does ‘our period’ begin? The rise of Late Antiquity has complicated matters (as Peter Brown has wryly said of the period that he has done most to define, ‘Late Antiquity - it's always later than you think!’), since we now have a different set of parameters entailing both the nature of historical transitions and transformation combined with a different appreciation of states in both the medieval west as well as in the Islamic world after the middle of the seventh century. Some argue for a new periodization that better reflects cultural historical changes and developments, one that begins in the third century CE and stretches into the later tenth or eleventh century.Footnote ⁷

Two issues in particular have drawn my interest in recent years, that of how the people we have become accustomed to calling ‘the Byzantines’ identified themselves, on the one hand, and on the other the nature of Byzantine governance, of the state and its relationship to the political ideas and praxis of its inhabitants. Most recently, we have seen an argument for a strong degree of continuity in terms of what are essentially Roman republican notions and practice of government, coupled with the idea that there was such a thing as a ‘national’ identity within the eastern Roman world from the seventh century.Footnote ⁸ These issues are clearly closely related, so in what follows I want to outline my own brief critique of the ways in which these issues have been handled and to offer some suggestions for moving ahead.

There are several ways to approach the issue of state and governance. We can look at east Roman views of the political arrangements of their world and interrogate their own understanding of what we would call ‘the state‘, ‘rulership‘, or ‘government‘. Or we can look at the ways in which the east Roman state was structured in terms of power-relationships and how these were articulated among different groups and individuals at any given time, extrapolating from this material a more general model of the nature of the state. These are two sides of the same coin, of course: to understand how the Roman state worked and evolved we need a model, a heuristic framework within which to explain why it develops in one direction and not another, or why east Romans react or respond in one way rather than another to particular developments. But we cannot construct a model without appeal to the evidence of how east Romans practised their politics, managed their state and thought about what they were doing. The two cannot logically be separated, although most attention has been focused on the first approach, and there are a number of substantive studies of Byzantine political theory and practice, among the best-known and most influential being Hans-Georg Beck's Res publica romana, published in 1970.Footnote ⁹

As Anthony Kaldellis has recently pointed out in a refreshing and stimulating restatement of the problem, Beck's views met with only a limited response and have only occasionally been incorporated into the debate, although more recently they have also been nuanced and developed, as several of the contributions to a recent volume on ‘power and subversion’ demonstrate.Footnote ¹⁰ There has also been a strong intervention on the part of a number of German scholars of Late Antiquity, especially in the field of political discourse and political action and in the identity of the various groups that could act and affect imperial politics, summed up in the term ‘Akzeptanztheorie‘.Footnote ¹¹

Most recently we owe thanks to Kaldellis for his provocative challenge to many aspects of the established assumptions about these matters. In his most recent book, The Byzantine Republic, he argues that the eastern Roman empire should not be understood as a theocracy, but rather as a ‘republican monarchy’ in which the people was politically sovereign, and in which the imperial ideology – the official rhetoric of the court – should be seen as an attempt to give support to a fragile monarchy (in the sense of the ruler of the moment rather than the institution itself), rather than a universally-held set of beliefs, and that it was not necessarily reflected in the actual political practice of most Romans. He further argues that the role of Christianity occupied a less significant place both ideologically and in practice in terms of political action, and that the continuity of what he sees as fundamentally republican ideas and ideals at all levels of society is the key to understanding medieval east Roman attitudes to rulership and to their polity (a term he prefers to use to ‘state’ since it is less confining in respect of either institutional arrangements or territorial identity). Thus the people as a totality embodies a set of republican ideals that they put into practice through autonomous political opposition whenever imperial policy is regarded as inimical to their interests (the interests of the moment, however, rather than any systemically rooted set of values). In consequence, emperors were subject to, and often deferred to, ‘the people’ (which appears under various guises depending on which oppositional moment is chosen), who exercised a normative extra-legal political activity legitimated in a republican discourse.

As an account of how the vocabulary and concept of government as transmitted through legal texts and longstanding literary tradition were wielded and manipulated, and of how those who understood these could deploy them; and as a theory of the imperial system as it was historically understood and imagined by such people, this is an important intervention that sets up a number of effective challenges to modern misperceptions of the attitudes and values prevalent in the medieval east Roman world. As a demonstration also of the fact that, regardless of their understanding or awareness of political concepts and theories, some people (but which ones?) in the medieval east Roman world were fully aware of their extra-legal ‘rights’ to oppose a regime or its decisions and policies, it carries likewise an important message. No-one would challenge the notion that the particular set of political-theoretical concepts and practices inherited through the Principate and Dominate constituted both this ‘republican’ background as well as ideas of monarchy and imperial rule. It framed people's ideas of what political activity entailed, permitted or demanded, and it was this that offered the metaphorical space within which to ‘think’ opposition, and within which practical action could take place. If the argument is to explicate and illustrate how this complex set of concepts and practices was articulated and given voice, then — with reservations about which groups or individuals are meant — it makes the point well.Footnote ¹² That is to say, people in the eastern Roman empire rebelled when they wanted to (for example) because they knew that this course of action was one of the traditional ways through which opposition could be expressed and could succeed.

But the argument is pushed further than this, and I am not sure that my characterization reflects its real direction and tenor. Much of the argument is in fact devoted to a demonstration of the reality of ‘the people’ intervening autonomously in the politics of the imperial government, and as a demonstration of the considerable numbers of occasions on which the populace of Constantinople was mobilized to oppose or challenge imperial policy, to accept or reject a candidate for the throne or to object to the appointment of one or more officials, it illuminates. Again, no-one can doubt that there were frequent and violent demonstrations of popular opposition to government in Constantinople, or that there was likewise provincial opposition and hostility to the court or to a particular ruler. Yet in spite of the contention that these were genuinely popular demonstrations that reflected a bottom-up demonstration of the will of the people, I think the case is difficult to maintain.

First, the question as to whether such political action reflected a conscious awareness of ‘republican’ ideas in all but a tiny educated elite is not really addressed, and that will, no doubt, be one of the issues that those who disagree with Kaldellis’ perspective will no doubt pursue. Just as importantly, the role of a wide range of vested sectoral interests (of individuals, cliques, families and clans, office-holders, imperial siblings and relatives, ecclesiastics, theologians and so forth) is to a degree subsumed within the notion of a republican consciousness, which means that the motivations that dominated each specific example of opposition tend to be lost.Footnote ¹³

One of the problems in some of the literature is the often casual use of terms such as ‘the Romans/Byzantines‘, ‘the people‘, and so forth, thus ignoring the fact that we cannot possibly speak of any single unitary body without extreme caution, at least not at all levels of social and political action, as the analyses of Flaig and Meier have demonstrated.Footnote ¹⁴ Yet most historians would argue that usurpations or attempted usurpations were not a consequence of some widely-recognised extra-legal authority held by ‘the Roman people’ to set aside or dethrone rulers or officials they did not favour. They were rather the result of a political system in which, first, inherited cultural assumptions included the idea that rulers could be challenged, and second, that a successful usurpation could be legitimated ideologically after the event (and, by the same token, de-legitimized by later commentators). The grounds for the challenge could be personal or dynastic rivalry or any other ground through which imperial policies or acts, whether in the secular or the religious sphere, were perceived as deleterious to the interests of the challengers. And of course, for those who understood the genre, expressing such opposition in the language of a traditional literary-political discourse was natural.Footnote ¹⁵ There is no doubt that members of the educated elite exploited and manipulated the language and the concepts of their past and deployed it in rhetoric, historiography and in politics to explain their own history, both contemporaneously and retrospectively. Kaldellis’ analysis demonstrates that this included tropes and terminology reflecting ‘republican’ sentiment that have been marginalized or ignored by modern commentators. This is an important point: the language of republican thought deployed by such people – whether Michael Psellos in the eleventh century or Procopius of Caesarea in the sixth – reflected both the cultural expectations of their readers as well as they themselves as writers, in addition to offering a ready-to-hand vocabulary for describing and accounting for political events in an acceptably elite form (Kaldellis is thus absolutely right, for example, to challenge the obsolete notion of an uncontested theocratic despotism and autocratic imperial rule).Footnote ¹⁶ I do not think many would disagree with his basic argument about the contested nature of imperial power. But imperial power there certainly was, and it was exercised in a clearly autocratic manner much of the time. That there was ever ‘a Roman people’ in day-to-day reality that could function as an autonomous political actor seems questionable.Footnote ¹⁷

There are two problematic concepts or notions at the heart of much of this discussion, the ‘state’ and ‘ideology‘. To be useful heuristically, both must describe more than just the specific society and culture under review, whether the eastern Roman empire, medieval China or some other culture. So I will conclude with a few words on my own suggestions for how best to deploy these terms in as heuristically useful and constructive a way as possible. ‘Ideology’ is a term that bears many meanings for many people, and Kaldellis opens a useful discussion of some of the issues associated with its use.Footnote ¹⁸ He is absolutely correct to challenge the ‘top-down’ idea of ideology as rhetoric or as a set of doctrines, but we need to define such terms to show how such beliefs or sets of ideas were grounded in the socio-cultural realities and relationships through which they were given expression. The debate is already an old one, of course, yet rarely addressed by historians of the Byzantine world, and Kaldellis’ motives in raising the issue in the first place cannot be flawed.

In everyday speech ‘ideology’ is generally used to mean a particular set of ideas representing the interests of a particular party — an interest group, a social class, a political party, or a government, for example, although it can also mean simply ‘what people believe’ as well as sets of ideas which are, to the outside observer, demonstrably false or one-sided in their account of the world. For the most part scholars of the later Roman and Byzantine world use the term to refer both to the generality of what people believed about their world — ‘matrices of meaning and normativity’ in Kaldellis’ words — as well as particular sets of ideas, both of which impact on social practice and which operate at different levels, and his discussion offers a sensible and flexible notion of how he will use the word. But we do not always have a clear idea of exactly how ‘ideology’ functions, even if at the end of his discussion (p. 3–4) he comes close to suggesting that a broader set of conceptual tools is required. In his account beliefs and ideology are tied into their cultural conditions of existence only rather loosely in respect of how social, economic, cultural and political conditions generated specific sets of ideas. It would be helpful, therefore, if we could agree on how exactly we want to deploy the term and within what sort of framework. To do that, we need first of all to understand where cognition and social praxis meet and how they are mutually constitutive, to think about the relationship between social structure on the one hand and human cognition on the other.Footnote ¹⁹

In the past I have used the notion of ‘symbolic universe’ to refer to the totality of cultural knowledge and practice in a social formation, within which and through which regular everyday life is carried on. Not all cultural knowledge is explicitly available to everyone, much of it remains tacit and unstated, much of it is specific to particular groups, all of it is embodied in the natural rhythms of social and biological reproduction. The relationship between consciousness and practice must be understood as a dialectic through which individuals receive their subjective awareness of self and their personal environment. This process also generates the conceptual apparatus through which they can in turn express what they know about the world, act back upon it; yet at the same time it sets limits to what they can know and how they can know, limits within which what we might call ‘the culturally possible’ can be thought. Contingently the symbolic universe is itself a product of social practice within a specific spatial and cultural environment, through which it is continuously reproduced. The activities carried on by individuals actively engaged in socially reproducing themselves, and hence in reproducing the social relations of production and reproduction of their particular cultural system, along with its roles and social institutions, have the material cultural effect of constantly regenerating the structural forms within which the same individuals are inscribed. And according to the socio-economic and cultural situation in which they find themselves, individuals and groups maintain particular roles and identities, drawing on different strands or narrative threads depending upon context. People can thus draw on a wide range of concepts and ideas in order to situate themselves with regard to others and the world around them, narratives or discourses that permit them to make sense of their place in society and in relation to the divinity, bundles of ideas and beliefs about the world extrapolated from the broader symbolic universe. This is for me a more useful way of thinking about the ways in which beliefs interpellate individuals (in Althusser's sense), because it retains a stronger emphasis on the individual's constitutive function in a social-cultural context. It allows us to situate specific strands or threads of cultural knowledge within a whole, to explain why there can be tensions and contradictions between different narratives and the forms of social praxis they represent and reflect, and to locate the historical moments and social contexts at which some of these threads or narratives become more, or less, active, and how they respond to shifts in their conditions of existence, i.e. how people deal with situations for which they have no prior model or experience. In this respect I would prefer to restrict the term ‘ideology’ to a specific and particular programmatic set of values and assumptions, a set of notions that has evolved in order to legitimate and justify a particular order of things — usually a political order.Footnote ²⁰ I think if Kaldellis had theorized more explicitly his notion of ideology and how it fits into a broader pattern of human cognitive activity it would have considerably enhanced and supported some key aspects of his discussion. It would also, perhaps, have permitted the avoidance of the — for me, at least — problematic overly-emphatic rejection of the Christian elements of late Roman and Byzantine beliefs and symbolic universe, without losing sight of his key point about the existence — and neglect by historians – of the secular cultural space).

A second question raised is that of the state. Whether or not the people we are studying had their own concept of a ‘state‘, and whether or not the modern word ‘state’ should be used in situations where no such concept existed is, it seems to me, to miss the point entirely.Footnote ²¹ We need conceptual tools that fulfill a heuristic function, and the fact that other cultures, contemporary or in the past, did not use a particular term should not disallow the application of such tools, if adequately defined. Indeed, we would not be able to analyse the political structures of the east Roman empire if we did not use such tools. It is certainly correct to point out that few scholars of the eastern Roman empire bother themselves with adequately defining their terms of reference, that there are contested meanings for the word ‘state‘, and that if we are to use the word, then it needs flexibly to account for the different ways in which other cultures, in this case the eastern Roman empire, articulate their political and power relations. Kaldellis’ conclusion is in itself not new: that the east Roman political system represented a form of state organization in which the relative stability of its institutions is vitiated by the patrimonial nature of patterns of social power. But there is no contradiction between the notion of patrimonial and kinship relations determining key aspects of the transmission of power and prestige within a bureaucratic system, on the one hand, and the fact of its institutional stability (in terms of continuity of administrative practice and technique, for example), indeed I think this is where a false dichotomy has been suggested, according to which we accept (or have accepted) either one or the other of these two approaches. This is simply not the case: indeed, I have consistently made it clear that the notion of the state being from one perspective ‘a web of personal relations that emanated from the emperor’ is entirely compatible with its institutional stability – the question is to establish how these relationships are articulated with the sets of administrative-bureaucratic practices that make up what we would call ‘the state‘.Footnote ²² As I have said elsewhere, ‘One of the most important points to bear in mind is the close identity between the institutional and systemic evolution of the state and its systems or apparatuses, and the individuals and groups from whom they were recruited or who comprised them, and whose local and vested interests determined to a large extent how such arrangements actually functioned in practical terms‘.Footnote ²³ As the social anthropologist Radcliffe Brown put it, the state is the product of social and economic relations and must, therefore, not be reified or personified in the process of analysis, even if it is also the case that the state has an identity as a field of action, as a role-constituting site of power and practices which can be independent, under certain preconditions, of the economic and political interests of those who dominate it and which can continue to function independently of significant disruptions.Footnote ²⁴

Part of the problem here lies in a confusion between an analytical description of the east Roman state on the one hand, and the perceptions and views of those who constituted it on the other. The way a state works reflects socio-economic relationships, cultural inheritance (in respect of configuring social praxis) and administrative-bureaucratic possibilities (in terms of communications and the management of distance and time, record-keeping, and so forth). What contemporaries thought about these arrangements and how they rationalized them through their own perceptions and understanding, as expressed through the cultural vocabulary available to them in their symbolic universe, is a different matter. Of course, in practice, in the way individuals and groups behaved in day-to-day respects, the two are barely distinguishable and are causally connected. But we need to distinguish them analytically in order to establish what these causal associations were and in order to explain change. And this brings us to a key point in Kaldellis’ discussion, namely ‘what the state is good for’.Footnote ²⁵

States may be rationally explicated as having a function, as being ‘good for’ something, and The Byzantine Republic does a good job of telling us how east Romans (or, simply, Romans) understood their politeia. Many states (but by no means all) in the pre-modern world have evolved sets of ideas about the reciprocity between ruler or rulers and subjects, between elected representatives and ‘the people‘, which typically involved a contractual obligation regarding the rights and obligations of both sides of the equation (quite apart from the differential relationship of either side to the divine and the sacred). Such ideas are almost always later rationalisations and legitimations, of course, even if people nevertheless harbour ideas about such things at all times.Footnote ²⁶ But that says nothing about the actual structure of a state and the ways in which the various elements of any political formation are articulated. To be quite clear: while states can be called into existence ostensibly in order to achieve moral (one thinks of the legitimating notions surrounding the evolution of the early Arab-Islamic state, for example) or even political and economic aims, their creation depends upon the outcome of conflicts over the control, distribution and consumption of resources, no more and no less, a point that is generally recognized in the modern literature.Footnote ²⁷

There is general consensus that states come into being only where there has evolved a division of labour between groups with differential relationship to the resources of a society (i.e. to the means of production, distribution and consumption of social wealth) and as a result of the assertion by one group – which becomes dominant – of its power over others. We can legitimately refer to these antagonistic groups as classes, in a purely economic sense, of course – whether or not this antagonism is expressed in political action, for example, or remains immanent is not the point.Footnote ²⁸ The existence of antagonistic classes in pre-capitalist social formations does not mean that the latter were not also divided institutionally and ideologically along other lines: the estates of medieval Europe, for example, or the tripartite political division of the eastern Christian world by theologians and others into producers, defenders and spiritual protectors.Footnote ²⁹ Even the simplest kind of organised polity generally has the effect of maintaining or creating a given socio-economic order, a situation that with few exceptions in the pre-modern world means that they protect the interests of those who control, by one means or another, the resources and wealth of a society, including both producers and what they produce.Footnote ³⁰ And in most states, if not all, authority and power ultimately rest on coercion or the potential for one group to impose its will on all other competing groups. Naturally the effectiveness of such authority will depend upon a series of contextual factors: geographical extent of the state, institutional forms through which power is actually exercised (for example, through a centralised and supervised central bureaucracy or through a dispersed provincial ruling elite), the nature of the legal apparatus, custom and tradition, the degree of penetration through society as a whole of various forms of ideological legitimation, and so on.

The social-economic aspect of the east Roman state is rather marginalised in The Byzantine Republic, which focuses heavily on the perceptions that Romans had of their state, but as a result tends to elevate their logic into the logic of the state itself. We thus end up with a Roman analysis of the east Roman empire that certainly helps us to see how they envisaged and understood their politeia, but does not explain how that politeia actually worked and how the social and economic relationships that underlay the polity created the ground in which different sets of ideas could grow and flourish, or not. The account of oppositional tendencies and forces, vested interests, contradictory ideological trends and narratives is largely divorced from the social-economic contexts in which they are rooted. This means that, in the end, the version of Byzantine politics that we get, in which notions of republican values and sentiment play such a key role, loses a good deal of its explanatory punch. Networks of power relationships, competition over control of resources, ideological competition over status and identity, whether or not shown to have been expressed by the east Romans themselves through the conceptual vocabulary of a ‘republican monarchy‘, these are pushed to the side or ignored. As a demonstration of how sets of ideas and concepts continued to play a key role in the self-identity of the elite; of how pre-existing (Roman, ‘republican‘) cultural-political practices and assumptions created the metaphorical space within which ideas about legitimised rebellion could be maintained and put into practice; as a demonstration of how such ideas could impact on the political activities of the elite and sometimes other groups who might belong to ‘the people‘, and as a demonstration of the inadequacy of the hitherto dominant notions of an all-pervasive ‘imperial ideology‘, the argument of The Byzantine Republic is persuasive. But socio-economic conflict or tensions, competition for control over and access to resources (whether in respect of money, manpower, land or political or ideological — including religious — influence) are left out of the picture. In the end the idea that it was ‘the Roman people’ who were the ultimate arbiters of who ruled and how is, for me, unpersuasive.

I do not intend my remarks to be seen as overly-critical of such recent stimulating interventions: quite the contrary — and there are other publications in the last few years around which a discussion might have focused in order to make some of the points above. But since some of the issues raised are of particular relevance to my own interests, I wanted simply to draw attention to some important and interesting new debates in the landscape of Byzantine (or medieval east Roman) studies, debates whose impact will reach far beyond their initial focus of interest and will certainly change the shape of our field in some interesting ways.