Roman Archaeology in an Epoch of Neoliberalism and Imperialist War

Neil Faulkner

A world on fire Britain’s largest demonstration against the Vietnam War took place in October 1968. It was 100,000 strong. The contingents from the major universities each numbered around a thousand. The London School of Economics was occupied for the weekend to provide accommodation; its refectory was turned into a workshop for producing posters, its basement into an emergency first-aid centre (Harman 1988: 159–160).

The anti-war movement today is much bigger. There have been several demonstrations of 100,000, one of 400,000, and one of two million (Murray and German 2005: passim). But where are the universities? Hundreds of academics and thousands of students have been active in the movement, but organised activity within the universities has been limited. Forty years ago the universities were at the centre of mass struggle against imperialism (Harman 1988: passim). Today they seem ideologically and politically inert, the musings of their scholars irrelevant to the millions protesting on the streets.

Universities have changed, of course. Their priorities are now shaped by big business, their managements guided by neoliberal ideology, those who work in them ground down by excessive workloads and inadequate resourcing (Callinicos 2006). To protest has become harder in the grind of modern campus life. But I argue that there is another reason for the academy’s failure to engage with the movement: the ideological pre-eminence of ‘postmodernist’ theories in contemporary discussions of imperialism.

A revisionist challenge There is an interesting contrast here with the new revisionism – a right-wing tendency within the academy that is explicitly supportive of imperialism. Francis Fukuyama, for instance, found himself jettisoned to fame and fortune when he portentously announced ‘The End of History’ in an article published in the U.S. magazine National Interest in 1989 – ‘the End’ being that heralded by the East European revolutions and the apparent global triumph of liberal capitalism. Fukuyama was rewarded for his ‘prescience’ with book contracts, lecture invitations, and a university professorship (Wheen 2004: 66–67).

The British historian Niall Ferguson has also been fêted by the neoliberal establishment. His academic work on western imperialism and the world wars correlates closely with neoliberal and imperial policy. Central to his argument is the notion that the only real choices in history are those between different great powers and their empires, with some playing ‘progressive’ roles promoting peace, free markets, liberal democracy, the rule of law, and the concept of individualism, while others are reactionary ‘rogue states’ that have to be confronted and defeated in the interests of humanity as a whole (Ferguson 2003: passim). In Ferguson’s work, the prime exemplar of the former is the British Empire, which he offers as a model for the American Empire today, urging the latter’s rulers and its people to accept the reality of their imperial role and mission. ‘The American Secretary of State Dean Ascherson famously said that Britain had lost an empire but failed to find a role,’ recalls Ferguson. ‘Perhaps the reality is that the Americans have taken our [sic] old role without yet facing the fact that an empire comes with it. The technology of overseas rule may have changed – the Dreadnoughts may have given way to F-15s. But like it or not, and deny it who will, empire is as much a reality today as it was throughout the three hundred years when Britain ruled, and made, the modern world’ (2003: 381).

This is not the place to take issue with this argument. The bloody history of the British Empire has been well documented elsewhere (e.g. Newsinger 2006), as has that of the American Empire (e.g. Neale 2001). My point is rather to illustrate that neoliberal and imperial policy is being underwritten by the revisionist theories of high-profile academics, and that their ideas have become mainstream. Martin Kettle, for example, wrote in his regular Guardian column recently that ‘the story of empire is not one of unalloyed shame’ and ‘even the horror, shame and persistence of slavery itself can’t shake my view that the building of the 21st century Americas – and above all the building of the modern United States itself, a society that after much struggle was eventually a pioneer of law, democracy and freedom – has proved to be the single greatest collective human achievement of the past four centuries’ (31 March 2007).

How does this relate to Roman archaeology? Like the British Empire, the Roman Empire has many admirers. Though an explicitly pro-imperialist orientation is no longer fashionable, naïve and simplistic perceptions of the Roman Empire as essentially benevolent and progressive endure. In one leading study, for example, first published in 1967 but still in print today in its third edition, we are told of Roman Britain that ‘the result was a synthesis, intended by Rome, and welcomed by the British people as they came to realise the advantages of peace and wealth conferred by membership of the empire’ (Frere 1987: 295). The evidence for the British people’s enthusiasm is not cited. Presumably it is the abundance of Romanised material culture found on sites of the period, as if one can simply read off political allegiance from a heap of pots. More sophisticated ‘processualist’ approaches to Roman Britain may display greater awareness of interpretive pitfalls, but they still view the enterprise as one in which elite-driven and trickledown processes of ‘Romanisation’, combined with proto-capitalist market economics, produced what was essentially an integrated, developing, and largely uncontested social order (e.g. Millett 1990).

Both traditional and processual approaches are essentially affirmative. They do not offer an anti-imperialist perspective. They do not challenge the new revisionism. For an alternative, many look to post-processualism. How does it meet the challenge of helping us to understand the past so that we can engage critically and effectively with the present?

The postmodernist paradigm: a non-critique Postmodernists are keen to stress how culturally rooted all ‘discourses’ are. They might do well to examine the cultural roots of their own. Postmodernism became an academically fashionable paradigm around 1980 (Callinicos 1989). This period was characterised by two related developments. First, there was a rejection of Keynesian economic management, state control of industry, redistributive taxation, and the welfare state: instead, there were to be tax cuts, spending cuts, privatisation, and free markets. Second, after the heyday of protest and radicalism in the late 60s and early 70s, there was a collapse of the Left, both in terms of numbers and influence, and in terms of ideas and optimism. After heady hopes of radical change, not only had the world failed to change for the better, it was looking worse all the time. The gloom was deepened by the illusion of many on the Left that Stalinism had represented a political alternative; the collapse of state capitalist dictatorships in Russia and Eastern Europe, therefore, appeared to them to be ‘the end of communism’. This was the context for the rise of postmodernism in the academy.

Postmodernism amounts to a token radicalism of despair. It proclaims ‘alternative voices’, but denies them any potential to change the world. In fact, it goes further: it denies them the analytical tools even to try. For, among the central claims of postmodernism are: a) that there is no single world that we can get hold of and attempt to change; and b) that certain knowledge of whatever it is that does exist – the certain knowledge that we need in order to act with confidence and effect – is unachievable. Postmodernism is nothing less than an attack on the Enlightenment, on reason, on science, on the very idea that human beings can understand their world and act collectively to control and change it. It is for this reason that Francis Wheen, in his acerbic study of modern intellectual fads, calls the postmodernists ‘the demolition merchants of reality’ (2004: 4). They are, to be blunt, engaged in a fundamentally reactionary project. It has been argued that postmodernism – or what tends to be called ‘post-processualism’ in the context of academic archaeology – is more a series of loosely related ideas than a coherent paradigm (Johnson 1999: 101). I suspect that one important reason for this is that it has become so deeply embedded in the academy that many are unaware of its intellectual roots, its theoretical implications, and the pervasiveness of its influence: it has become the unquestioned ‘taken-for-granted’ of academic discourse in general. A good example is provided by Matthew Johnson in his Archaeological Theory: an introduction – a book which, despite the title, is in fact a postmodernist polemic. Johnson is explicit about his identity as a post-processualist, yet also declares: ‘There are better and worse interpretations of the past. I do not know of a single archaeologist of whatever theoretical stripe who would disagree. … The criticisms of an unrestrained relativism are many, various and irrefutable’ (1999: 170–172). So there we have it: we all know there are good theories and bad theories. But how do we know? There is only one possible answer: something must exist ‘outside the text’ – a single material world able to generate concrete evidence against which we can test our theories. QED. Johnson has rendered his own claim to be a ‘post-processualist’ meaningless – yet apparently without blinking an eyelid.

Alex Callinicos has been among those proclaiming that the emperor has no clothes. He sees the roots of postmodernism in three related cultural developments: a change in the arts (broadly defined) from ‘modernist’ to ‘postmodernist’ forms and themes; a current in French philosophy known as ‘post-structuralism’; and the essentially socio-economic observation that we have moved from an ‘industrial’ to a ‘postindustrial’ age (Callinicos 1989: 2–3). These intellectual developments achieved critical mass in the context of wider political and ideological changes – essentially the shift from social-democratic reformism to neoliberalism referred to above – around 1980. For Callinicos, moreover, postmodernism’s key social base is a class of disillusioned and ageing former leftists – ‘the 1968 generation’ – who now have relatively well-paid jobs in academia, media, and the arts (1989: 7). Do any of our self-proclaimed ‘postprocessualists’ recognise themselves in the following?

‘The discourse of postmodernism is best seen as the product of a socially mobile intelligentsia in a climate dominated by the retreat of the western labour movement and the ‘overconsumptionist’ dynamic of capitalism in the Reagan-Thatcher [we might now say Bush-Blair] era. From this perspective the term ‘postmodern’ would seem to be a floating signifier by means of which this intelligentsia has sought to articulate its political disillusionment and its aspiration to a consumption-oriented lifestyle. The difficulties involved in identifying a referent for this term are therefore beside the point, since talk about postmodernism turns out to be less about the world than the expression of a particular generation’s sense of an ending.’ (Callinicos 1989: 170–171)

Callinicos identifies the key philosophical underpinnings of the new paradigm as follows: ‘Despite many disagreements, all three [Gilles Deleuze, Jacques Derrida and Michel Foucault] stressed the fragmentary, heterogeneous and plural character of reality, denied human thought the ability to arrive at any objective account of that reality, and reduced the bearer of this thought, the subject, to an incoherent welter of sub- and trans-individual drives and desires’ (1989: 2)

In relation to the study of the past, a postmodernist perspective seems to involve the following essential claims:

1. That there is no single past. Rather, there are diverse experiences and discrepant perceptions, and these constitute a series of different pasts.

2. That the past is a construct of negotiations between different historical actors, each bringing their diverse experiences and discrepant perceptions to the negotiation.

3. That there is no single ‘truth’ about the past knowable in the present. This is partly because the past was formed of diverse experiences and discrepant perceptions, such that no single ‘truth’ ever existed. It is also because academic discourses reflect contemporary ideologies, leaving us trapped in our own thought-world with no reliable access to the thought-worlds of the past. Moreover, in the view of some, it is even because no external material reality capable of being known actually exists; there are those for whom there is, quite literally, nothing ‘outside the text’.

Such claims have been the target of withering criticism. The historian Eric Hobsbawm, for example, has decried:

‘the rise of ‘postmodernist’ intellectual fashions in western universities, particularly in departments of literature and anthropology, which imply that all ‘facts’ claiming objective existence are simply intellectual constructions. In short, that there is no clear difference between fact and fiction. But there is, and for historians, even for the most militantly antipositivist ones among us, the ability to distinguish the two is absolutely fundamental.’ (quoted in Sokal 1996: 2)

These obvious criticisms seem to me unanswerable. There is usually, in any case, no attempt to answer them: instead, confronted by them, we have what might be called ‘slippage’, whereby postmodernists simply deny that what they are being criticised for is what they actually believe, ignoring the fact that the very foundations of their paradigm are being knocked away. Despite the problems, these ideas have achieved the widest currency in archaeological discourse. Let me illustrate some of them as they appear in recent academic writing on the Roman Empire. John Barrett’s work represents an extreme case, in which all sense of structure in the past is dissolved in an endless sequence of negotiations and discourses.

‘… upon what grounds do we believe something called the Roman Empire actually existed? … The Roman Empire, as a product of discourse (the product of the numerous material and historical conditions under which such discourse was practised and the different qualities of power which addressed different subjects within that empire), was never a single reality.’ (1997: 59)

The prose is obscure, and that is no accident. This is a contradictory, and therefore confusing, statement, because in the parentheses power inequalities are implied (I presume that is what is meant by ‘different qualities of power’), and these have to be part of a single, concrete, actually existing reality. You can be powerful in relation to me only in so far as I am powerless in relation to you: the two conditions are inherently part of a single and very real social relationship. Yet for Barrett, ‘different qualities of power’ have no sooner been mentioned than they have dissolved into nothing – along with the Roman Empire, which we have no good grounds for believing ‘actually existed’ – except of course as ‘a product of discourse’.

Perhaps it is a mistake even to engage with such empty-headed nonsense. The work of David Mattingly is a different matter. Here, contradictions between theoretical paradigm and social reality assume much greater importance – essentially because Mattingly knows perfectly well that the Roman Empire most definitely ‘actually existed’, especially for its countless victims. He has contributed usefully and powerfully to our perception of the Roman world as a place where social experience and cultural identity were variable because they were embedded in unequal power relations. The concepts ‘discrepant experience’ and ‘discrepant identity’ are a welcome riposte to the naïve and sanitised versions of the past implicit in most accounts of ‘Romanisation’. We part company on two critical matters: first, that he continues to operate within what I consider to be a broadly postmodernist framework; and second, that this framework is incapable of explaining the dynamic of Roman military imperialism.

Introducing a session entitled ‘Dialogues in Roman Imperialism: power, discourse, and discrepant experience in the Roman Empire’ at the First Roman Archaeology Conference in 1995, Mattingly argued the following:

‘Imperial ‘power over’ encounters both resistance and indigenous ‘power to’. To some extent, the victims of such power networks are those who resist actively, and those who negotiate and who seek a measure of accommodation with the dominant partner can be empowered in the process.’ (1997: 10)

‘The material record shows the result of a long-continued process of negotiation, accommodation and resistance. … the power network could be a creative force for change in society, with some part of the population empowered through its operation.’ (1997: 19) Now there is a contradiction here. On the one hand, there is an implicit invitation to embrace the notion that past social reality is merely a construct of negotiations informed by the diverse experiences and discrepant perceptions of different historical actors. On the other, ‘the power network’ is given a central significance. Nor is the latter a mere passing remark: Mattingly insists repeatedly on the importance of the power network and on locating discrepant experiences and identities within it:

‘The first priority must be to locate indigenous people in the power networks and colonial discourse that bound them to Rome, and to seek to understand the prelude, processes and results of their complex negotiations (societal and personal) with the imperial power.’ (1997: 10) The problem is that the power network is not explained – there is no theory of Roman imperialism, the Roman state, and Roman class society – so we are never able to contextualise the essentially secondary process of creating ‘identities’ which Mattingly is at such pains to describe. The absence of any such explanation takes us in dubious directions. Consider the claim, for example, that the power network was ‘a creative force’ capable of ‘empowering’ native people. In so far as you have power over me, you do not need to negotiate: you simply impose your will, and I can either conform or suffer the consequences. Negotiation does not take place. Effective choice is denied. That is because there is a structure, and it is real. Imagine, for instance, that you are a Roman tax-collector and I am a native British farmer in A.D. 100. I may wish to ‘re-negotiate’ my relationship with the empire and adopt a new cultural ‘identity’ – as a Roman citizen. The corollary, of course, is that I would not have to pay tax (a good reason for choosing this particular identity). Are you likely to agree?

The point is that there is no ‘empowerment’ for me where you have power and I do not. What Mattingly is referring to is not ‘empowerment’ at all, but the fact that certain players, by accommodating themselves to the new order, found a place within it and thereby became part of the existing power structure. This feature of Roman rule is a commonplace in academic discussion. But the fact that Agricola was Romano-Gaulish, Hadrian, Romano-Hispanic, and Augustine, Romano-African represented, at the time, a reinforcement, not an undermining, of the imperial order. This is not to deny that there were sometimes conflicts within the power structure, conflicts that may occasionally have pitted Romanised native elites in one part of the empire against central authority. The internal history of the Roman Empire is, of course, punctuated by civil wars. Nor is it to deny that ‘negotiations’ sometimes took place between the polity and those with independent power-bases. Frontier politics were dominated by such processes. But in any real negotiation, both parties must have power. As an isolated native farmer I could not refuse to pay my taxes to your tax-collection squad without dire consequences. But if I joined with ten thousand others in refusing to pay, I might thereby empower myself enough at least to negotiate a reduction. I do not make an abstract point here: there are numerous examples recorded in the ancient sources of popular revolts in which taxation was an issue (e.g. Faulkner 2002a: 43–44). But in such cases, ‘negotiation’ was secondary, being entirely contingent on the power of collective resistance. Mattingly is explicit about the reality of Roman imperialism in his recent book on Roman Britain. He describes the Roman state as ‘a ruthless military power’ (2006: 6) and suggests that ‘for every winner under Roman rule, there were a hundred losers, with the gap between the richest and poorest in society widening as never before’ (2006: 20). He also defines empire as ‘rule over very wide territories and many peoples without their consent’ (2006: 13), and argues that ‘the power dynamics of imperial rule are dramatically unequal and the globalising influence of the coloniser should not be underestimated’ (2006: 525). Despite this, the theoretical core of the study, what he calls his ‘key analytical tool’, is not, as one might expect, a theory of structure – of how power is created, maintained, and exercised – but a theory of identity (2006: 18). We learn that there are many identities in Roman Britain, and that they result from divergent experiences, perceptions and responses, and from ‘the Roman and native dialogue’ in which ‘the negotiation of power’ takes place (2006: 520–528). There is an elephant in the room – Roman imperialism – and the intelligent observer cannot help but notice it; theory, on the other hand, seems blind. Mattingly knows that Roman imperialism exists; he knows that there is a power structure that controls and constrains; but describing how people react to that reality, how they define themselves in relation to it – describing, that is, the (changing) ‘identities’ they adopt – is not the same as offering an explanation of the thing itself.

Mattingly’s often excellent study of Roman Britain is hollow in the middle: there is no theory of Roman imperialism. He has, I suspect, reached the point where, to move forward – as he clearly wishes to – he must recognise that much of his theoretical apparatus is still rooted in a postmodernist ‘turn’ which is both deeply reactionary and intellectually vacuous. Postmodernism does not, and cannot, explain imperialism. And to the extent that archaeologists cling to elements of the postmodernist framework, to that extent they cannot explain ‘the power relations’ that frame agency.

A Marxist paradigm: uniting agency and structure One of the strengths of Marxist approaches to understanding the past is that they combine a theory of structure and a theory of agency in a single framework. Structure both constrains and enables action. It limits what we can do – preventing me from getting away with not paying my taxes, for example – but at the same time makes certain types of action possible – giving a political elite the resources (from taxes successfully collected) to make war, for example. Structure is the framework within which agency takes place. It is essential, therefore, if one wishes to understand the past, to analyse structure, and that means getting to grips with inequalities in wealth, power and opportunity, and understanding the processes of class exploitation, state power, and imperialist war that give rise to such inequalities. The study of structure is therefore necessarily a study of processes, since it is the latter that create and recreate structure over time; this in turn means that it becomes also a study of agency, since all social processes involve human action. In other words, not only does structure, by constraining and enabling, frame agency in the present; it is itself a product of agency in the past, a product of the accumulating results of human activity and social processes over time. It is a fundamental and obvious error in postmodernist thinking not to realise that ‘agency in the present’ is conditioned by ‘agency in the past’ that has been transformed into structures of power. The action of thousands of tax-collectors, supported by thousands of soldiers, in the past creates a state structure which is powerful enough to force a native British farmer to pay his taxes in A.D. 100. This is how Marx put it:

‘Men make their own history, but they do not make it just as they please; they do not make it under circumstances chosen by themselves, but under circumstances directly encountered, given and transmitted from the past. The tradition of all the dead generations weighs like a nightmare on the brain of the living.’ (quoted in Callinicos 1983: 99)

How does all of this compare with postmodernist perspectives? Let us first deal head-on with the claim that postmodernism is a critical and radical alternative to traditional approaches. This seems to hinge on two main contentions: a) that it deconstructs the self-justifying ideologies of past elites; and b) that it gives voice to, and re-evaluates the contributions of, non-elite actors. The problem here is that there is nothing remotely new about this. The idea that the ruling class promotes an ideology to justify its wealth and power is, of course, basic Marxism. ‘The ideas of the ruling class,’ wrote Marx, ‘are in every epoch the ruling ideas: that is, the class which is the prevailing material force of society is at the same time its ruling intellectual force’ (quoted in Callinicos 1983: 99). Or again, speaking more generally about the wider cultural context in which ideas develop, there is this:

‘Men, developing their material production and their material intercourse, alter, along with this their actual world, also their thinking and the products of their thinking. It is not consciousness that determines life, but life that determines consciousness.’ (quoted in Callinicos 1983: 78) As for giving a voice to the unheard, this has been central to the work of Marxist historians since at least the 1940s. What seems to be happening with regard to postmodernist claims in this respect is that two quite different things have been conflated. On the one hand, there is the theoretical claim that there is no single past, but rather many divergent experiences and discrepant perceptions adding up to any number of different pasts: this is a distinctively postmodernist argument. On the other hand, there is the attempt to reconstruct the lives, actions and thoughts of the exploited and oppressed, those often ‘hidden from history’. Now, there is nothing postmodernist about the latter. ‘History from below’ was an essential feature of the work of a group of outstanding postwar British historians that included Christopher Hill, Rodney Hilton, Eric Hobsbawm, Edward Thompson, and others. It is worth quoting from the preface of what was perhaps the greatest single study produced by this group, The Making of the English Working Class, where Thompson explains his project:

‘I am seeking to rescue the poor stockinger, the Luddite cropper, the ‘obsolete’ hand-loom weaver, the ‘utopian’ artisan, and even the deluded follower of Joanna Southcott, from the enormous condescension of posterity. Their crafts and traditions may have been dying. Their hostility to the new industrialism may have been backward-looking. Their communitarian ideals may have been fantasies. Their insurrectionary conspiracies may have been foolhardy. But they lived through these times of acute social disturbances, and we did not. Their aspirations were valid in terms of their own experience; and, if they were casualties of history, they remain, condemned in their own lives, as casualties.’ (1980: 12)

Here you have it all: the aim is to bring the lives of the exploited and oppressed into the historical limelight, and to make sense of their attitudes and actions in the context of their own troubled world. But there is no implication that the world of the hand-loom weaver was not the same as that of the mill-owner. Indeed, it was precisely because the two were part of a single world that the former faced ‘obsolescence’. Let us be quite clear: it is one thing to bring those hidden from history into prominence; it is quite another to argue that they were able to construct their own social reality. Only the latter claim is distinctively postmodernist. And reduced to this, we discover that it is not a radical claim at all. Quite the contrary: if the exploited and oppressed are not constrained by structure, if they are free to construct their own social reality through ‘discrepant perceptions’ and ‘negotiation’ with other actors, then clearly, their exploitation and oppression will soon evaporate in the hot air of ‘discourse’.

The point is that, without a theory of structure, one cannot contextualise and explain agency. The postmodernist paradigm has the effect of dissolving all structure into agency, such that agency is unconstrained, free-floating, endlessly ‘negotiating’ a social reality that is in fact no reality, because it never takes form, never becomes solid and enduring, remaining forever a mere will-o’-the-wisp. We are left with no way to explain the inequalities of class and power that are self-evident to all thoughtful observers, and if we cannot explain them, neither can we develop any theoretically grounded critique of exploitation, oppression, and war.

Notes towards a Marxist theory of ancient imperialism What might a theory of the structure of the Roman Empire consist of? My own perspective is a synthesis of many other studies (among the most important of which are Finley 1973, Hopkins 1978, Harris 1979, and de Ste Croix 1981). It foregrounds the imperialist character of the Roman state, arguing, in effect, that Rome was a dynamic system of robbery with violence. This theory has been developed in detail elsewhere (Faulkner 2000, 2002b, 2004, and 2008), but, for convenience, I will summarise it here in the following terms.

Roman conquest involved huge levels of killing, displacement, plunder, and destruction, which a) enriched and empowered the ruling class which controlled the imperial state, b) were hugely wasteful of productive resources, and c) caused incalculable human misery. It can be estimated, for example, on the evidence of his own testimony in The Gallic War, that Caesar’s eight-year campaign of conquest in Gaul in the 50s B.C., killed a million people, enslaved a million more, and torched several hundred settlements; on the other hand, it was this campaign that provided him with the resources to wage a successful domestic struggle for supreme power in the civil war of 49–45 B.C.. The scale of such transfers of wealth from conquered peoples to the Roman ruling class cannot usually be quantified, but careful analysis of the exceptions allowed William Harris to claim with confidence that plundering was the main purpose of Roman warfare (Harris 1979, passim).

Once established, Roman rule involved huge ongoing transfers of surplus to the imperial core (defined in a political and economic rather than a geographical sense), which a) were mostly wasted in military and other non-productive expenditures, b) despoiled and debilitated local economies, and c) imposed increased levels of exploitation and poverty on the productive classes of the empire. The Roman Empire constituted a ‘military-supply’ economy, in which state tax-collection and arms expenditure, oiled by state supply of coin (the ‘tax-pay cycle’), drained surplus from the provinces to support the military. This created a two-tier economy: a primary economy of local subsistence production continued, but it was overlain by a secondary economy of production, exchange and distribution through regional, provincial and imperial marketing networks, with the former taxed in various ways to sustain the latter.

How was this wealth used? The ancient world was divided into rival polities that were frequently at war. Each of these polities was also internally divided: there were conflicts between elite factions at the top of society, and conflicts between different social groups across society as a whole. In Rome’s case, we can define the central dynamic as a process of ‘competitive military imperialism’, where aggressive war was motivated by the drive to accumulate the resources to fund a) intra-ruling class competition for office and wealth, b) a security apparatus to contain popular resistance, and c) the frontier defences that protected imperial territory. The central state was a bureaucratic-military complex for expropriating and consuming surpluses amassed through war, conquest, taxation, and labour corvées.

The empowerment of landowning elites was a secondary but necessary feature of this system. Rome allowed such elites to retain property and power, and provided an internal security apparatus that enabled them to increase the rate of exploitation on their estates and accumulate enlarged surpluses that were then mainly wasted in monumental building and privatised luxury. This created a large class of wealthy, powerful, and locally-rooted stakeholders in the imperial system, who could be relied upon to maintain order, collect taxes, and uphold the status quo. The point here is that the Roman Empire remained a traditional agricultural society, and its two heaviest foundation-blocks were land and labour. Maintaining the infrastructures of empire and civilisation – soldiers, forts, towns, villas, ‘the world of taste’ – depended upon effective elite control over the countryside, the exploitation of peasant labour, and the accumulation of agricultural surpluses.

Much in the foregoing may be provisional and in need of modification or elaboration. I certainly would not want to claim that these notes represent a fully-fledged theory. I do say that only by deploying such ‘hard’ concepts – concepts like class exploitation, state power, and imperialist war, concepts derived from scientific observation of social reality – can the past be properly understood. Postmodernism’s ‘soft’ concepts – ‘discourse’, ‘negotiation’, ‘identity’ – merely describe (or mis-describe) aspects of agency; they do not, and cannot, explain the constraining and enabling structures that frame agency.

Critical theory: the test of praxis The world is again on fire. The neoliberal policies and imperialist wars of western rulers face global resistance and mass protest on a scale without historic precedent. Millions of people mobilised in anti-war and anti-capitalist protests are open to radical alternatives to the ideas offered by political elites and their media echo-chambers.

Postmodernism is congenitally incapable of supplying such alternatives. There is no point of connection between its mystification of reality and the radical commitment of the movements. When Blair claimed that Iraq had ‘weapons of mass destruction’, this was not an alternative ‘discourse’: it was a lie. When a young Afghan joins the Taliban because US bombers have destroyed his village, he is not ‘re-negotiating’ his ‘identity’: he is joining a war. When the tortured describe their ordeals in Abu Ghraib, these are not ‘discrepant perceptions’: they are records of atrocities. Because exploitation, oppression and violence are embedded in structures of wealth and power, no paradigm which lacks a theory of structure can have any purchase in a real clash of social forces. In a world on fire, postmodernism is neither radical nor relevant. It has failed the test of praxis. Critical theory must have the ability to inform and inspire critical practice. To do this, it must have a theory of both structure and agency, of their interrelationships, and of how those interrelationships create both processes of exploitation, oppression, and violence, and possibilities of fighting for change.

Surely the pervasive and lingering influence of postmodernism is one reason for the political quiescence of the universities today. Is it not the case that the absence of critical theory has sapped the capacity of the academy to engage with protest? Is that not why our conferences have become dull litanies of fashionable jargon instead of centres of debate in a polarised world? The academy may be resisting the return of Marx. In the outside world, on the demonstrations and in the social forums of the protest movement, Marx is back. The only question is how long it takes the academy to catch up.

Acknowledgements Thanks are due to David Mattingly (University of Leicester), Gabriel Moshenska (University College London), Steve Roskams (University of York), and Dave Wythe (University of Newcastle) for reading this article in draft and offering useful critical comment.