When Scots get together to celebrate their national identity, they do so in ways steeped in tradition. Men wear the kilt, with each clan having its own tartan - and their ceremonials are accompanied by the wail of the bagpipes. By means of these symbols, they show their loyalty to ancient rituals - rituals whose origins go far back into antiquity.

Except for the fact that they don't. Along with most other symbols of Scottishness, all these are quite recent creations. The short kilt seems to have been invented by an English industrialist from Lancashire, Thomas Rawlinson, in the early 18th Century. He set out to alter the existing dress of highlanders to make it convenient for workmen.

Kilts were a product of the industrial revolution. The aim was not to preserve time-honoured customs, but the opposite - to bring the highlanders out of the heather and into the factory. The kilt didn't start life as the national dress of Scotland. The lowlanders, who made up the large majority of Scots, saw highland dress as a barbaric form of clothing, which most looked on with some contempt. Similarly, many of the clan tartans worn now were devised during the Victorian period, by enterprising tailors who correctly saw a market in them.

Much of what we think of as traditional, and steeped in the mists of time, is actually a product at most of the last couple of centuries, and is often much more recent than that. The case of the Scottish kilt comes from a celebrated volume by the historians Eric Hobsbawm and Terence Ranger, called The Invention of Tradition. They give examples of invented traditions from a variety of different countries, including colonial India.

The British set up an archaeological survey in the 1860s to identify the great monuments of India and to preserve Indian 'heritage'. Believing local arts and crafts to be in decline, they collected together artefacts to put in museums. Before 1860, for example, Indian soldiers and the British both wore Western-style uniforms. But in the eyes of the British, Indians had to look like Indians. The dress uniforms were modified to include turbans, sashes and tunics regarded as 'authentic'. Some of the traditions they invented, or half invented, continue on in the country today, although of course others were later rejected.

Tradition and custom - these have been the stuff of most people's lives for most of human history. Yet it is remarkable how little interest scholars and thinkers tend to show in them. There are endless discussions of modernisation and what it means to be modern, but few indeed about tradition. When I was researching for this lecture, I came across dozens of academic books in English with 'modernity' in the title. Indeed, I have written a few myself - but I could only discover a couple of books specifically about tradition.

It was the 18th Century Enlightenment in Europe that gave tradition a bad name. One of its major figures, the Baron D'Holbach, put things this way. I quote: 'Instructors have long enough fixed men's eyes upon heaven, let them now turn them upon earth. Fatigued with an inconceivable theology, ridiculous fables, impenetrable mysteries, puerile ceremonies, let the human mind apply itself to the study of nature, to intelligible objects, sensible truths, and useful knowledge. Let the vain chimeras of men be removed, and reasonable opinions will soon come of themselves, into those heads which were thought to be forever destined to error'.

It is clear that D'Holbach never intended a serious engagement with tradition and its role in society. Tradition here is merely the shadow side of modernity, an implausible construct that can be easily brushed aside. If we are really to get to grips with tradition, we can't treat it merely as folly. The linguistic roots of the word tradition are old. The English word has its origins in the Latin term tradere, which meant to transmit, or give something to another for safekeeping. Tradere was originally used in the context of Roman Law, where it referred to the laws of inheritance. Property that passed from one generation to another was supposed to be given in trust - the inheritor had obligations to protect and nurture it.

It might seem that the notion of tradition, unlike kilts and bagpipes, has been around for many centuries. Once more, appearances are deceptive. The term 'tradition' as it is used it today is actually a product of the last 200 years in Europe. Just like the concept of risk, which I talked about in my last lecture, in mediaeval times there was no generic notion of tradition. There was no call for such a word, precisely because tradition and custom were everywhere.

The idea of tradition, then, is itself a creation of modernity. That doesn't mean that one shouldn't use it in relation to pre-modern or non-Western societies, but it does imply that we should approach the discussion of tradition with some care. By identifying tradition with dogma and ignorance, the Enlightenment thinkers sought to justify their absorption with the new.

Disentangling ourselves from the prejudices of the Enlightenment, how should we understand 'tradition'? We can make a good start by going back to invented traditions. Invented traditions and customs, Hobsbawm and Ranger suggest, aren't genuine ones. They are contrived, rather than growing up spontaneously; they are used as a means of power; and they haven't existed since time immemorial. Whatever continuity they imply with the long-term past is largely false.

I would turn their argument on its head. All traditions, I would say, are invented traditions. No traditional societies were wholly traditional, and traditions and customs have been invented for a diversity of reasons. We shouldn't suppose that the conscious construction of tradition is found only in the modern period. Moreover, tradition always incorporates power, whether they are constructed in a deliberate way or not. Kings, emperors, priests and others have long invented traditions to suit themselves and to legitimate their rule.

It is a myth to think of traditions as impervious to change. Traditions evolve over time, but also can be quite suddenly altered or transformed. If I can put it this way, they are invented and reinvented.

Some traditions, of course, such as those associated with the great religions, have lasted for hundreds of years. There are core prescriptions of Islam, for instance, that nearly all Muslim believers would hold to, and which have remained recognisably the same over a very long period. Yet whatever continuity there is in such doctrines goes along with many changes, even revolutionary changes, in how they are interpreted and acted upon. There is no such thing as a completely pure tradition. Like all the other world religions, Islam drew upon a dazzling variety of cultural resources - that is, other traditions. The same was true of the Ottoman empire more generally, which incorporated Arab, Persian, Greek, Roman, Berber, Turkish and Indian influences, among others, across the years.

But it is simply wrong to suppose that for a given set of symbols or practices to be traditional, they must have existed for centuries. The Christmas address by the Queen, which is broadcast every year in Britain, has become a tradition. Yet it only started in 1932. Endurance over time is not the key defining feature of tradition, or of its more diffuse cousin, custom. The distinguishing characteristics of tradition are ritual and repetition. Traditions are always properties of groups, communities or collectivities. Individuals may follow traditions and customs, but traditions are not a quality of individual behaviour in the way habits are.

What is distinctive about tradition is that it defines a kind of truth. For someone following a traditional practice, questions don't have to be asked about alternatives. However much it may change, tradition provides a framework for action that can go largely unquestioned. Traditions usually have guardians - wise men, priests, sages. Guardians are not the same as experts. They get their position and power from the fact that only they are capable of interpreting tradition's ritual truth. Only they can decipher the real meanings of the sacred texts or the other symbols involved in the communal rituals.

The Enlightenment set out to destroy the authority of tradition. It only partially succeeded. Traditions remained strong for a long while in most of modern Europe and even more firmly entrenched across most of the rest of the world. Many traditions were reinvented and others were newly instituted. There was a concerted attempt from some sectors of society to protect or adapt the old traditions. After all, this is basically what conservative philosophies have been, and are, all about. Tradition is perhaps the most basic concept of conservatism, since conservatives believe that it contains stored up wisdom.

A further reason for the persistence of tradition in the industrial countries was that the institutional changes signalled by modernity were largely confined to public institutions - especially government and the economy. Traditional ways of doing things tended to persist, or be re-established, in many other areas of life, including everyday life. Once could even say there was a sort of symbiosis between modernity and tradition. In most countries, for example, the family, sexuality, and the divisions between the sexes remained heavily saturated with tradition and custom.

Two basic changes are happening today under the impact of globalisation. In the Western countries, not just public institutions but everyday life is becoming opened up from the hold of tradition. And other societies across the world that remained more traditional are becoming detraditionalised. I take it this is at the core of the emerging global cosmopolitan society I have spoken of in previous lectures.

This is a society, I argued last week, living after the end of nature. Few aspects of the physical world, in other words, are any longer just natural - unaffected by human intervention. It is also a society living after the end of tradition. The end of tradition doesn't mean that tradition disappears, as the Enlightenment thinkers wanted. On the contrary, in different versions, it continues to flourish everywhere. But less and less - if I can put it this fashion - is it tradition lived in the traditional way. The traditional way means defending traditional activities through their own ritual and symbolism - defending tradition through its internal claims to truth.

A world where modernisation is not confined to one geographical area, but makes itself felt globally, has a number of consequences for tradition. Tradition and science sometimes mingle in odd and interesting ways. Consider, for instance, the much-discussed episode that happened in India in 1995, when the deities in some Hindu shrines appeared to drink milk. On the same day, several million people, not only in India but throughout the world, tried to offer milk to a divine image. Denis Vidal, an anthropologist who has written about this phenomenon, remarks that I quote again: 'by manifesting themselves simultaneously in every country of the world inhabited by Indians, the Hindu deities may have succeeded in performing the first ever miracle in tune with an era haunted by the slogan of globalisation'. Just as interestingly, it was widely felt - by believers as well as non-believers - that scientific experiments were needed to authenticate the miracle. Science was enlisted in the service of faith.

Tradition in such an example isn't only still alive, it is resurgent. Yet traditions also often succumb to modernity, and are doing so in some situations all over the world. Tradition that is drained of its content, and commercialised, becomes either heritage or kitsch - the trinkets bought in the airport store. As developed by the heritage industry, heritage is tradition repackaged as spectacle. The refurbished buildings at tourist sites may look splendid, and the refurbishment may even be authentic down to the last detail. But the heritage that is thereby protected is severed from the lifeblood of tradition, which is its connection with the experience of everyday life.

In my view, it is entirely rational to recognise that traditions are needed in society. We shouldn't accept the Enlightenment idea that the world should rid itself of tradition altogether. Traditions are needed, and will always persist, because they give continuity and form to life. Take academic life, as an example. Everyone in the academic world works within traditions. Even academic disciplines as a whole, like economics, sociology or philosophy, have traditions. The reason is that no one could work in a wholly eclectic fashion. Without intellectual traditions, ideas would have no focus or direction.

However, it is part of academic life continually to explore the limits of such traditions, and foster active interchange between them. Tradition can perfectly well be defended in a non-traditional way - and that should be its future. Ritual, ceremonial and repetition have an important social role, something understood and acted upon by most organisations, including governments. Traditions will continue to be sustained insofar as they can effectively be justified - not in terms of their own internal rituals, but as compared to other traditions or ways of doing things.

This is true even of religious traditions. Religion is normally associated with the idea of faith, a sort of emotional leap into belief. Yet in a cosmopolitan world, more people than ever before are regularly in contact with others who think differently from them. They are required to justify their beliefs, in an implicit way at least, both to themselves and others. There cannot but be a large dollop of rationality in the persistence of religious rituals and observances in a detraditionalising society. And this is exactly as it should be.

As tradition changes its role, however, new dynamics are introduced into our lives. These can be summarised as a push and pull between autonomy of action and compulsiveness on the one hand, and between cosmopolitanism and fundamentalism on the other. Where tradition has retreated, we are forced to live in a more open and reflective way. Autonomy and freedom can replace the hidden power of tradition with more open discussion and dialogue. But these freedoms bring other problems in their wake. A society living on the other side of nature and tradition - as nearly all Western countries now do - is one that calls for decision-making, in everyday life as elsewhere. The dark side of decision-making is the rise of addictions and compulsions. Something really intriguing, but also disturbing, is going on here. It is mostly confined to the developed countries, but is becoming seen among more middle class groups elsewhere too. What I am speaking about is the spread of the idea and the reality of addiction. The notion of addiction was originally applied exclusively to alcoholism and drug-taking. But now any area of activity can become invaded by it. One can be addicted to work, exercise, food, sex - or even love. The reason is that these activities, and other parts of life too, are much less structured by tradition and custom than once they were.

Like tradition, addiction is about the influence of the past upon the present; and as in the case of tradition, repetition has a key role. The past in question is individual rather than collective, and the repetition is driven by anxiety. I would see addiction as frozen autonomy. Every context of detraditionalisation offers the possibility of greater freedom of action than existed before. We are talking here about human emancipation from the constraints of the past. Addiction comes into play when choice, which should be driven by autonomy, is subverted by anxiety. In tradition, the past structures the present through shared collective beliefs and sentiments. The addict is also in thrall to the past - but because he or she cannot break away from what were originally freely chosen lifestyle habits.

As the influence of tradition and custom shrink on a world-wide level, the very basis of our self-identity - our sense of self - changes. In more traditional situations, a sense of self is sustained largely through the stability of the social positions of individuals in the community. Where tradition lapses, and life-style choice prevails, the self isn't exempt. Self-identity has to be created and recreated on a more active basis than before. This explains why therapy and counselling of all kinds have become so popular in Western countries. When he initiated modern psychotherapy, Freud thought he was establishing a scientific treatment for neurosis. What he was in effect doing was constructing a method for the renewal of self-identity, in the early stages of a detraditionalising culture.

After all, what happens in psychotherapy is that the individual revisits his or her past in order to create more autonomy for the future. Much the same is true in the self-help groups that have become so common in Western societies. At alcoholics anonymous meetings, for instance, individuals recount their life histories, and receive support from others present in stating their desire to change. They recover from their addiction essentially through rewriting the story-line of their lives.

The struggle between addiction and autonomy is at one pole of globalisation. At the other is the clash between a cosmopolitan outlook and fundamentalism. One might think that fundamentalism has always existed. This is not so - it has arisen in response to the globalising influences we see all round us. The term itself dates from the turn of the century, when it was used to refer to the beliefs of some Protestant sects in America, particularly those who rejected Darwin. Yet even in the late 1950's there was no entry for the word 'fundamentalism' in the large Oxford English dictionary. It has come into common coinage only over the past two or three decades.

Fundamentalism is not the same as either fanaticism or authoritarianism. Fundamentalists call for a return to basic scriptures or texts, supposed to be read in a literal manner, and they propose that the doctrines derived from such a reading be applied to social, economic or political life. Fundamentalism gives new vitality and importance to the guardians of tradition. Only they have access to the 'exact meaning' of the texts. The clergy or other privileged interpreters gain secular as well as religious power. They may look to take over the reins of government directly - as happened in Iran - or work in conjunction with political parties.

Fundamentalism is a controversial word, because many of those called fundamentalists by others wouldn't accept the term as applying to themselves. So can an objective meaning be given to it? I think it can, and I would define it in the following fashion. Fundamentalism is beleaguered tradition. It is tradition defended in the traditional way - by reference to ritual truth - in a globalising world that asks for reasons. Fundamentalism, therefore, has nothing to do with the context of beliefs, religious or otherwise. What matters is how the truth of beliefs is defended or asserted.

Fundamentalism isn't about what people believe but, like tradition more generally, about why they believe it and how they justify it. It isn't confined to religion. The Chinese red guards, with their devotion to Mao's little red book, were surely fundamentalists. Nor is fundamentalism primarily about the resistance of more traditional cultures to Westernisation - a rejection of Western decadence. Fundamentalism can develop on the soil of traditions of all sorts. It has no time for ambiguity, multiple interpretation or multiple identity - it is a refusal of dialogue in a world whose peace and continuity depend on it.

Fundamentalism is a child of globalisation, which it both responds to and utilises. Fundamentalist groups almost everywhere have made extensive use of new communications technologies. Before he came to power in Iran, the Ayatollah Khomeini circulated videos and cassettes of his teachings. Hindutwa militants have made extensive use of the Internet and electronic mail to create a 'feeling of Hindu identity'.

Whatever form it takes - religious, ethnic, nationalist, or directly political, I think it right to regard fundamentalism as problematic. It is edged with the possibility of violence, and it is the enemy of cosmopolitan dialogue.

Yet fundamentalism isn't just the antithesis of globalising modernity, but poses questions to it. The most basic one is this: can we live in a world where nothing is sacred? I have to say, in conclusion, that I don't think we can. Cosmopolitans, of whom I count myself one, have to make plain that tolerance and dialogue can themselves be guided by values of a universal kind.

All of us need moral commitments that stand above the petty concerns and squabbles of everyday life. We should be prepared to mount an active defence of these values wherever they are poorly developed, or threatened. None of us would have anything to live for, if we didn't have something worth dying for.

1999 Reith Lectures Home | BBC Radio 4 Home | BBC Homepage