Ana Pauker was a minister in the infamous Stalinist administration that ruled Romania in the early fifties; she was noted for her unswerving devotion to the orthodoxies of Moscow. A favourite anecdote of the day described how she was seen in the streets of Bucharest on a warm spring day under a large umbrella: why? 'Because', she said, 'it's raining in Moscow.'

Conviction is not always the purely mental business we sometimes assume it to be. Ana Pauker not only had ideas about where the truth lay; she acted as if she were somewhere else, somewhere where this truth was more clear and more tangible. And states have always been suspicious of those sorts of conviction which appear to demand this kind of belonging elsewhere, this additional loyalty. Christians in the eastern Roman Empire of the early fourth century were executed because they were assumed to be agents of the Christian kingdom of Armenia; not much later, they were executed in Persia as suspected agents of the now Christian Roman Empire. Pope Pius V's decree of excommunication against Elizabeth I condemned hundreds of Roman Catholics in the century that followed to persecution and death as agents of a foreign power. Communists in the West, from the twenties to the fifties, were widely assumed to have deeper loyalties to the Soviet Union than to their own land. Muslims today in the West and Christians in Iraq are victims of the assumption that they are foot soldiers in the new global conflict between the just society and its enemies.

But one of the interesting features of modernity is a suspicion that is related to this, but significantly different. What I've so far been talking about is a situation in which one 'sacred' order, a nation or an empire with a clear sense of its legitimacy and its right to demand loyalty, detects in some of its citizens a loyalty to some rival order that also claims to be legitimate or in some sense sacred. But increasingly what we see, in the actual policies of some states and in the rhetoric of the political classes in other states, is a presumption that the rational secular state is menaced by the public or communal expression of religious loyalty. It is not a matter of one sacred order (empire or nation state or religio-political unit) facing a rival, but of a sense that the public space of society is necessarily secular – that is, necessarily a place in which no local or sectional symbolic activity is permissible. Conviction is free – that is a foundational principle of modern liberal society; but visible and corporate loyalty to the marks of such conviction is to be strongly discouraged. It puts in question the neutrality of the public space and can be read as a sort of aggression against other convictions or against the programmatic absence of convictions that the state assumes for public purposes. For statutory authority to collude with, let alone actively support these loyalties fatally compromises the very basis of legitimate liberal society.

It is the argument of the French government, and of those who, for example, resist the spread of religiously based schools in our own country. It is not trivial or frivolous, but it is important to try and understand what it takes for granted before we accept it as axiomatic. And what it assumes is in fact an aspect of modern Christian rhetoric which has come to speak of religious conviction as essentially a matter of individual option – a rather shopworn version of the Reformation concern with justification by faith and the Reformation critique of unaccountable Church authority. The modern state presents itself as a guarantor of public order and peace as against religious communities that demand allegiance and come to blows over the terms of that allegiance. In the textbook version of modernity, the Wars of Religion in the early modern age produce a kind of disgust with the claims of public faith, religious commitment that is active in the arena of public life. And those aspects of the Reformation which stress interior devotion and are cautious about the Church as a visible political entity are drawn in to reinforce the argument for a non-confessional state, in which private conviction is free, but public loyalty is exclusively claimed by the state itself. In various ways and in varying degrees, the Christian confessions of the modern western world accepted this programme over the succeeding centuries.

But the story is not after all so simple. As the American theologian William Cavanaugh has recently argued, it would be much more accurate to say that the Wars of Religion arose from the use of religious controversy by political agents to justify military adventures aimed at national consolidation or expansion, which in turn consolidated the new centralised sovereign authorities within nation states. 'Liberal theorists', he writes, 'would have us believe that the state stepped in like a scolding schoolteacher on the playground of doctrinal dispute to put fanatical religionists in their proper place' (Theopolitical Imagination, p.43). But in fact, the development of the belief that the state enjoyed a monopoly of legitimate coercion meant that defining the exclusive boundaries of sovereign territory was far more important than before: the state needs to know whom it can confidently command, whose allegiance it can safely enforce, whom it can mobilise in its defence. Thus the modern state is not in origin a pacific force; it is a system committed to violence as 'the primary mechanism for achieving social integration in a society with no shared ends' (p.46).

Shared ends cannot be taken for granted; so particular kinds of social integration are problematic. The state takes each citizen as an agent in need of integration with others, and proposes a concordat to secure agreed boundaries between the interests of these agents. Thus the modern state assumes a direct relation with its citizens, not mediated through other communities –a relation whose contractual quality is already in evidence early on, though it becomes far more evident in the market societies of late modernity. And part of this relation is the state's assurance of individual rights, claims that can be enforced in its courts; among these is the right to freedom of conscience as regards personal belief. However, modern states have not been eager to recognise any priority for personal conviction when this may affect the performance of public duties – as is clear from the varying attitudes to conscientious objection in wartime. While there is usually recognition of a belief system that declares all war illicit, it is far harder to make any case for the person who regards this or that particular war as illicit on grounds of religious belief. Such an attitude suggests that the person's religious commitment involves both an additional level of social belonging, a membership in some other nexus of relations than that of the state, and a formation in critical questioning of the state's decisions, a reluctance to take for granted the legitimacy of these decisions without some further scrutiny.

This whole cluster of issues has become more immediate and practical with the current complexities over the modern state's relation to Muslim identity. Liberal commentators properly concerned to combat anti-Muslim prejudice (Nick Cohen, for example, in an article in the New Statesman of October 1st) persist in assuming that Islam is a set of convictions in the mode of much modern Christianity. To suggest that the Muslim owes an overriding loyalty to the international Muslim community, the Umma, is worrying; it is a factor in Muslim identity (say the liberal commentators) that intensifies suspicion towards the Muslim community in a quite unnecessary way. What is desirable is thus for Muslims to make clear that their loyalty is straightforward modern political loyalty to the nation state, unaffected by the private convictions that individual Muslim believers happen to hold in common.

This means accepting – among other things – the view that the rights conferred by the modern state are essentially rights accorded to individuals in respect of their private consciences. And here is where one of the major conflicts in modern legal thinking comes into focus. The liberal state has repeatedly had to make accommodation with minority communities, not simply with individuals – ethnic minorities whose identities have been damaged by state centralism, religious groups, even those making a specific choice of lifestyle (the fiercely controverted politics of gender and sexual orientation is commonly cast in terms drawn from the struggles of ethnic and religious minorities. The liberal state has made numerous concessions to a multicultural model which is in fact at odds with classical liberal thinking. And the British state is unusually and interestingly hybrid in this respect, combining vestiges of a confessional polity (the establishment of the Church of England) with a generally liberal political culture and an increasing pragmatic acceptance of multiculturalism.

Maleiha Malik, a professional jurist of Muslim allegiance, has recently written at length on this conflict, arguing that, since we cannot just go back to conservative nationalism, and since the interests of minority groups are not adequately safeguarded by classical liberal principles of individual entitlement and non-discrimination, we need a more sophisticated model of the relation between the state and its minorities, which in turn requires some rethinking of the original picture of the state contracting with a mass of atomised individuals. Her conclusion is that the state cannot avoid legislation about the rights of communities; but that there then needs to be a pathway for minority communities to find new ways of identifying with public processes and social institutions, so that (instead of making the main form of protection against discrimination the guarantees provided by the courts, and so consuming immense energy in litigation) specific groups may play a positive role in framing policy before legislation is finalised. And this needs more developed representative institutions of consultation. The issue of 'rights' for a minority religious community thus comes to be allied with a wider set of questions about local democracy and the weaknesses of an 'elective dictatorship' model of parliamentary rule (M.Malik, 'Muslims and Participatory Democracy', in British Muslims:Loyalty and Belonging, ed. Mohammad Siddique Seddon, Dilwar Hussain and Nadeem Malik, pp.69-85).

What this implies is in fact a subtle reframing of the issue of loyalty. Loyalty to a sovereign authority is replaced by or recast as identification with a public process or set of public processes; the simple question about loyalty, 'Are you with us or against us?' becomes a question about adequate and confident participation in a law-governed social complex. We are taken beyond a polarised picture of exclusive loyalty to the state menaced by mysterious fifth column-ish affiliations elsewhere. Loyalty to the Umma is not necessarily in competition with dependable citizenship in the state if the state's practices of consultation and acknowledgement of communal identities remove the threat of a total and terminal privatising of religious conviction – the individual believer facing the state which tolerates his or her belief system but cannot deal with them as they concretely are, embedded in the specific relationships that in fact constitute the believing identity itself.

In the same book as Maleiha Malik's essay there is also a piece by Tim Winter, one of the foremost expositors of Islamic thought in the UK, on 'Muslim Loyalty and Belonging', in which he argues that the danger faced by Muslim communities from within is a primarily negative model of Islamic identity, a model lacking 'cultural embeddedness'. Alienated equally from familial tradition (usually regarded as hopelessly compromised), from a living tradition of reflective exegesis and from the environment fostered by global capitalism, the Muslim radical recognises the essence of Muslim faith and law only in what totally negates all these; and this, says Winter, is 'a religion of the gaps, a kind of void...a list of denials, of wrenchings from disturbing memories' (ibid., p.9). But the implication of this is that a secular state that fails to address Malik's agenda about participation for minority communities is one of the factors that intensifies pressures towards such a negative definition of religious loyalty; more seriously, it may intensify the sense that there is indeed a competition between two comparable loyalties.

Now there are certainly features in classical Islamic political thought that lend themselves to the notion that the 'House of Islam' is always and necessarily a clearly defined positive political unit; but there is also a long tradition of reflection on what it is to be a minority in a non-Islamic setting. Part of the problem in all this is precisely the modern assumption that political belonging itself is always unitary and clearly defined. The Islamic model is perhaps best thought about not as creating a parallel state, but as positing a loyalty to divine law, variously incarnated in different concrete political units, but always having a claim even when imperfectly realised (see Winter, art.cit.,p.20, for a very helpful brief discussion of this. Such a loyalty is indeed about the possibility of critical participation in the institutions of a secular state; but it requires the state to think clearly about what it understands by religious belonging and not to reduce it to the level of a private voluntary association of the like-minded.

This particular discussion ought to sharpen the agenda of Christian theologians, and to send them back to some foundational texts. Early Christianity, as we have seen, is a communal phenomenon proclaiming an allegiance that is deeply threatening to the unitary and sacred identity of the ancient city and the ancient empire. I have argued elsewhere that what we find in some of the records of the martyrs is in fact a surprisingly novel account of political loyalty: the accused refuse to treat the emperor as divine, but they accept the duty of paying taxes and praying for the public good. Thus they see themselves as participating in a public process, not as rebels against existing order; but they will not regard their loyalty to the state as a matter of exclusive and absolute obligation, religious obligation. They are, it seems, trying to clarify the sense in which political loyalty and religious loyalty are not in direct competition – though they may be in conflict if the former makes claims appropriate only to the latter. Similar points are made in the famous second century Letter to Diognetus, in which we find a clear affirmation of the Church as a distinctive community with different moral horizons, but also a denial that this means a visible separation from public life. The lapidary statement in the Epistle to the Philippians that the politeia of Christians is in heaven is evidently taken to imply both that Christian belief is a form of participatory citizenship and that this exists in parallel with and sometimes in tension with the forms of citizenship that exist in any given social order.

It is a less clear-cut picture than we find in classical Islam – or perhaps we should say, a more sceptical one, in which hopes for the realisation of an actual state embodying the law of God are far more muted. But we should not therefore conclude that the difference between the two traditions of faith is between a public and a private or a corporate and a personal ideal. The modern Christian tendency has indeed been to stress – as Reinhold Niebuhr so eloquently did – the way in which faith can purify motivation and reinforce the standards of moral commitment that prevail in a liberal society; but an increasing number of Christian moralists have observed that this leaves no role for the Church as a body with its own integrity and peculiarity, as a community whose end is to form human lives according to some model (and by some power) other than that of the individual agent of the liberal ideal. Stanley Hauerwas has recently characterised Niebuhr's view as effectively 'a complex humanism disguised in the language of the Christian faith' (With the Grain of the Universe, p.131); the liberal Christian approach assumes that the business of Christian commitment is not to produce lives that participate in the holiness of Christ so much as lives that can be lived with a fairly easy conscience within the arrangements of the modern state, motivated by a rather unspecific inspiration.

This takes us back to the sort of case made by Cavanaugh in the work referred to earlier. He identifies with some sharpness the way in which the modern state generates what he calls a 'state-society complex', defining what is culturally acceptable or dominant. 'What is permissible as public discourse increasingly obeys the logic of accumulation; state-funded school lunch programmes are defended n terms of increasing students' performance and thus enhancing the country's position in the global economy vis-à-vis the Japanese' (p.71). This cultural hegemony is deeply hostile to the practice of communities that do not begin from the same standards of instrumental rationality; it insists upon the naturalness and indisputability of this sort of rationality, and increasingly affects the concrete practice and the rhetoric of religious communities themselves. Cavanaugh, like Hauerwas, denies that the 'public space' of contemporary culture is in any meaningful sense neutral or free, and insists that the Christian community must be distinguished by the telling and enacting of a story that is different from that propagated by the modern state. This, of course, involves exposing the fact that the modern state does in fact tell a story – that is, that it is not the embodiment of a timeless rationality. In the face of the narrative of modernity, the Church declares that its account of human nature and calling and possibility is constituted by a set of highly specific events, figured forth in sacramental action. And the community that gathers around these actions is one in which human beings are deliberately shaped in the likeness of one particular form of life, the form that is defined in the story, the identity, of Jesus Christ.

Thus Cavanaugh goes on to say that the activity of the Church in the public sphere should not and indeed cannot be confined to 'lobbying'. Helpful as this is in some circumstances, the main task is to create 'spaces' for an alternative story – to challenge the self-evidence of the narrative of secular modernity. This is in no way a bid for religious takeover or confessional monopoly, but it does make a claim for the visibility of Christian community – and other religious communities. It assumes that the state does not have the right to demand a sort of secrecy on the part of religious groups, a contract by which no sign of particular religious loyalty is admitted in the public space. But it means that the religious community needs to be clear about its primary responsibility as a place where people are formed in moral vision by shared practice; and such practice, says Cavanaugh, is both the ritual practice by which the basic story is learned and repeated and stories of alternative human behaviour shaped by this. Christian involvement in the public sphere is visible celebration of the sacramental reality by which believers live, and the devising and implementing of usually small-scale projects suggesting possibilities for human beings different from those assumed by contractual and acquisitive stories. Cavanaugh's book ends with a moving narrative bringing these two elements together: following the murder in 1977 of a Salvadorean priest who had resisted government injustice, Archbishop Oscar Romero had decided that the requiem for this priest should be the only mass celebrated in the diocese on the Sunday following, so as to 'collapse the spatial barriers separating the rich and the poor' (p.122). The reader will recall Romero's own death not long afterwards, in the context of another eucharistic celebration: drawing rich and poor together in this way is also to draw near to real danger in some political contexts.

This is a dramatic example – too dramatic almost to be helpful. Drama is always attractive for a variety of questionable reasons, and it is important to remember the caveat long ago entered about a particular kind of socially-conscious British theology or spirituality – that it borrows the language and imagery of more tragic and serious situations to glamorise its well-meaning reformism. Yet we need some stories of alternative possibility, even in the less apocalyptic atmosphere of modern Britain: perhaps the church-based mass campaigns around international debt, and, currently, trade justice provide something of this, especially since they have not been by any means confined to 'lobbying' but have been regularly linked with concrete policies and actions by local churches (commitment to the use of Fair Trade goods, for example). When all the cautions have been entered, the point still stands: the Church as a political agent has to be a community capable of telling its own story and its own stories, visible as a social body and thus making claims upon human loyalty. While not a simple rival to the secular state, it will inevitably raise questions about how the secular state thinks of loyalty and indeed of social unity or cohesion. To this degree, it is not in a different case from the Muslim Umma.

But the overall picture begins now to look very complicated. The defender of public secularity might well say, 'This is precisely what we need to worry about: social cohesion is threatened by highly visible religious groups. This is not an issue about some sort of sinister totalitarian uniformity, only of the conditions for public stability, dependable order.' The concern is a serious one. The apologist for a robust account of religious visibility in the public sphere has to do some work to clarify the role of the state in such an argument, and the character of the stability or 'social integration' that might be hoped for. It will not do, surely, to rest content with a sort of Hobbesean situation in which strong, community-based ethical frames of reference simply fight it out between themselves.

In what remains of this lecture, I want to try and suggest some ways of responding to this concern. The first thing to say is that no one begins with a clean slate in this context: we are not seeking to develop a polity out of nothing. The modern state has developed for reasons that represent irreversible trends in social history, trends to do with vastly expanded population levels, the universal administration of welfare and the exigencies of defence. As has been argued at length in some recent writing, the shift from 'nation state' to 'market state' represents an intensifying of the contractual element in modern polity, to the degree that national government is both pressed to increase its responsiveness to consumer demand and eager to devolve some of its duties to private contractors. Its actual sovereignty is also deeply compromised by global economic forces. Yet it remains, even in 'market' form, a system for guaranteeing at least two things: a structure that determines the economic standing of every person within defined geographical boundaries and thus provides a clearing-house of comprehensive information (and often, in consequence, a comprehensive welfare scheme); and a monopoly of legitimate force, as expressed in a legal system and a defence budget. In a world in which populations are mobile as never before, and in which there is therefore bound to be wider ethnic and religious diversity in any area than hitherto, there is no possibility of dismantling either of these. Government needs to know whom to tax – and in some circumstances whom to conscript – and who has a claim to be defended. The apparatus of the modern state is essential for a law-governed society in which diverse communities exist side by side.

The modern state thus realises – however unsatisfactorily – one central element in any corporate religious vision: it militates against simply forgetting any person, treating individuals as dispensable. Since every religious system considers that the existence of each person is non-contingent, an expression of positive divine will, this has to be a basic religious concern. In this respect, modern polity and the religious communities we are familiar with stand together against – for example – ancient slave societies. And, as hinted earlier, this allows us to see political loyalty as bound up with the claims of law rather than mere geography or ethnic history (the emotional loyalty felt towards place and history is another matter, of course, and not our primary focus here): the state apparatus is a means to secure certain rights and liberties, which, whatever exactly we might want to say about them, at least represent a statutory seriousness about each human subject as citizen.

The state thus implicitly proposes a kind of minimum level for political virtue – the keeping of the law as a form of acknowledging the basic claim of other agents to the same stability or security as you desire for yourself. It is not quite the same as the all-important liberal principle of toleration; more a bare recognition of mutual limits between agents and of a certain conceptual space occupied by other agents in respect of an order not determined by either you or them in isolation. Our problems arise when this bottom line of political virtue is fleshed out by some more definite philosophy – often derived from specifically non-religious axioms (Cavanaugh's 'state-society complex'). And the religious apologist ought to be able to say something like this. 'I accept the universality of the rule of law in this specific contingent setting, and the monopoly of force, internal and external, which it entails. I do so because of the irreducible commitment of my faith to the non-contingent quality of each agent. But this is still mostly a procedural or formal acceptance. As a religious believer, I have some matters of content to propose here as representing what my tradition sees as appropriate virtues, or even as making for holiness. And my task becomes one of negotiating how much of that content a diverse population can own for itself.'

Thus, the state will formally be bound to provide for the nurture and formation of children. The religious person will want to argue for this to take the particular form, say, of state support for monogamous marriage. For the believer, this is something enjoined by basic beliefs – that human fidelity is an imaging of divine covenant with the community, for example. But the public argument has to be that this might be a desirable framework for maximal stability for the growing child; and there would need to be statistical work to support this as a negotiating position. Likewise, the state provides protection against homicide. The religious person sees this as grounded in the divine relation to each person, from conception to death, a relation that sets up certain indestructible inter-human relations also; and is therefore extremely suspicious of the limitations on protection involved in the legalisation of abortion or euthanasia. But the public argument will be about whether these limitations involve so deep a contradiction that the actual principle is undermined.

The point is that the religious 'negotiator' in such cases (which are, as you will recognise, far from academic at the moment) accepts that much of the argument will be pragmatic – but acts and argues from a position that is not pragmatic, and that is formed by specific narratives and practices. S/he does not seek the right to impose those narratives, but, by engaging in public dispute on these matters, denies that there is a simple and neutral resolution. Even as arguments of contingent evidence are deployed, the believer's own position is not finally dependent on them, and the motivation for pursuing a critical debate arises from loyalties that are bound up with narratives of God's commitment to the believing community. It is thus equipped to survive particular defeats within the system; it insists that majority votes do not specify what is true, even if they determine what is, descriptively, at any moment lawful. And this is a significant critical element in any society, which, if it is not to be at the mercy of pure legal positivism, needs some vantage point from which it is possible to discuss moral questions in terms that go beyond the constraints of a majority vote alone.

This is to fill out somewhat an argument sketched by Raymond Plant. Specific communities of religious commitment can go a fair way to accepting the common rule of law as enshrining something central for them; but they will inevitably be seeking to 'thicken the texture' of law in the continuing process of public argument over political and social virtue. A society that is rich in settings for such public argument is one in which the participation of diverse religious groups will not threaten the basic acceptance of law, but will seek to persuade a hybrid public of the possibilities of putting fuller content into its common life – at least to accept or endorse policies that keep open such possibilities, even for those who do not share the beliefs and practices that ground them. We return to Maleiha Malik's case for a constant strengthening in a mature modern democracy of the openings for substantive debate, involving communities of commitment – not only individual bearers of legal rights. 'Institutional identification is more likely where substantive issues concerning the common good are discussed' (art. cit. p.80).

A final word: I do not intend to comment in detail on current proposals about legislation against 'religious hatred'; but I note that their critics are mostly secular liberals and conservative Christians. Secular liberals fear the restriction of critique and challenge, including satire, conservative Christians fear the restriction of seeking conversion or open debate with members of other faiths. And while I am not wholly sure of the rightness of the legislative proposals in question, all I have said in this lecture is meant to question the presuppositions of such critics. The secular liberal may need convincing that a stronger, more publicly recognised community will be less rather than more likely to fear public discussion, even satire. The conservative Christian may need convincing that the law is capable of recognising specific and active difference and disagreement without assuming that it will always issue in violent or prejudicial action. But this will not happen for either of these unless there is a general climate that understands the public character of religious loyalty, even in the secular state – and that is prepared to work with the visibility of such commitment without panic. The signs in our own political milieu are not wholly discouraging by any means, but we need more thought about the basis of co-operation, if 'faith communities' are to be more than a pool of cheap labour for projects of social integration. But ultimately we do not have to be bound by the mythology of purely private conviction and public neutrality; and, if my general argument is right, the future of religious communities in modern society should show us some ways forward that do not deliver us either into theocracy or into an entirely naked public space.

© Rowan Williams 2004