The EU has made military conflict in Western Europe inconceivable. The question is, after the Cold War, can it find a fresh purpose?

Imagine a parallel Teddy Taylor-approved universe in which the European Union did not exist. Would we now be pondering how to invent a Europe-wide European Union? Would we be trying to erect a huge apparatus to heal the Cold War scars of Europe, to organise economic policy, monetary policy, immigration policy, foreign policy, security policy to cover the entire Continent (Russia apart) from Romania to Ireland via Finland and Malta?

The answer is almost certainly not. Most countries would think the effort too great; the problems too grinding; the challenges to national prerogatives too great.

But that is not the same thing as saying (returning to the real world, the one not approved by Teddy Taylor) that we should cheerfully allow the EU to fail; that we should allow the infuriating, ambitious 40-year experiment in Brussels to be crushed under the weight of new members to the east and south; by changing circumstances and confusing new challenges.

The European Union - uncertain, divided, unloved - is not in particularly good shape to make economic and political sense of the post-Iron Curtain Europe. Too bad; the task falls to the EU by default.

The Warsaw Pact and Comecon are deceased. Efta is a mere rump. The Council of Europe is worthwhile but weak. Nato is a military alliance without an obvious enemy (and Europe's new problems are not primarily military). Other institutions may have a part to play but the EU is the main player left in the game. A 21st century Europe without the EU would have the same problems but no obvious starting point for solving them.

This is the real challenge for the rolling constitutional negotiations on the future of the EU - the Inter-Governmental Conference - that will be launched by heads of government in Turin next week. The two big questions facing the European Union in 1996 are economic and monetary union (EMU), and how to expand from the present 15 to maybe 27 nations. Neither issue (typically enough) is formally on the agenda of the IGC. And yet, like cosmic black holes, both will exert huge influence on the proceedings (which could last well into next year).

Is the EU capable of meeting these challenges? Is its structure, designed by six member states to meet a different, and simpler set of problems, capable of such an evolutionary leap? Many of the original premises on which the EU was built are no longer true: Germany is no longer divided; Europe is no longer the fault line of a global, ideological chess game. Some argue that the importance of the European single market itself has been diminished by progress towards global free trade.

Since the questions have changed, why stick to the same answers? Is the whole concept of a supra-national, centralised EU - founded primarily on the preservation of a single market - outmoded? Should the concept of the aquis communautaire (in-for-one-thing, in-for-everything) be abandoned in favour of a more flexible system, in which countries sign up for some policies and not others?

We can dispose of the United States of Europe pretty quickly. Such a monolith was probably unattainable with 10 or 12 or 15 countries; with 27 it is inconceivable. But it would be equally disastrous to return to 27 or more competing nationalisms in Europe; or to allow many of the former east bloc countries to slip back into totalitarianism, or into a new Russian empire. Even mild Euro-sceptics argue that, to embrace the eastern and southern supplicant 12, we should abandon, or seriously weaken, the present EU, founded on central institutions and supra-national law. This is a dangerous delusion. Some reform and adjustment is essential; some degree of flexibility inevitable; but the EU must retain a solid institutional and legal core.

The British White Paper on the IGC published last week recognises this principle with surprising warmth (to the fury of the anti-Brussels diehards). At the same time, it places dozens of obstacles in the way of the kinds of reforms that will be needed to allow the EU institutions succesfully to admit Malta, Estonia, Poland, Hungary and the rest.

In the modern world purely national interests - in trade, research, the environment, currency management, immigration, drugs, terrorism - are becoming difficult to identify. In broad outline, the EU has been a successful attempt to go beyond the limitations and dangers of the nation state - something now being copied around the globe, but nowhere with as much ambition as in Europe.

The existence of a common body of law, and common institutions, makes the EU far more tenacious and more robust than other multi-national bodies in which nations merely "co-operate" (the Euro-sceptic watchword). At the crudest level, European law, and European institutions, give national governments a cover for unpopular decisions. They give a secure framework for deal-making, in which one interest can be traded against another. The single market - to which 60 per cent of British exports now go - could never have been built on "co-operation" alone.

All of this is far short of Jean Monnet's dream of economic integration forcing greater and greater political integration. No matter. The Monnet vision has, in a sense, been a victim of its own success. It has reached a sort of halfway house to federalism but, in doing so, it has prematurely achieved its principle aim: to make war, or any dire economic conflict, between the western European nations inconceivable. There is no real need to push on into the political Passchendaele of a fully federal Europe. On the other hand, when absorbing eastern Europe, why abandon an engine for the promotion of peace and prosperity that has been broadly successful for western Europe?

The old argument used to be between broadening (enlargement of the EU) and deepening (making more political decisions at European level). The debate has now moved on to try to accommodate both. The buzz-phrase is "variable geometry": an EU in which all pupils do not have to study the same subjects. In a sense this was invented by the Maastricht treaty. Britain won an opt-out from social policy. It was accepted that all member states would not be ready to join a single currency by the end of the century. There is nothing wrong with the principle; some form of flexibility will be essential in an EU of 27. The danger is that flexibility may lead to disintegration.

The Germans and French, backed by the Commission, say they want a bigger community; but they also want the right to push ahead with political integration alongside a small group of like-minded states. In a sense, this is nostalgia for 1957 when the original Six ignored the warnings of Britain and others and pushed ahead alone.

The difficulty is to know exactly what the Franco-German elite would DO that they are not doing already. The single currency is the obvious candidate; but that is already provided for in the Maastricht treaty. If a small group of countries do merge their currencies - probably later than the present 1999 deadline - they would inevitably form a kind of economic and political Premier League. What other issues would be a candidate for the hard core? Foreign or defence policy? The French guard their foreign affairs veto as jealously as the British. A European defence policy without Britain, which has the best armed forces, would be fatuous.

Some degree of variable geometry will inevitably occur in a 27-member EU. But, single currency apart, its importance may be exaggerated. The real problem will be to decide which areas of EU policy remain compulsory for all; and how to organise institutionscapable of serving such an extraordinary collection of countries.

What should be the irreducible minimum of the EU as it spreads like a weed over most of Europe? How should it be organised?

The bedrock must remain the single market, enforced as now by European law and institutions. Here the old Monnet model is still vital. The effort by the eastern apsirants to achieve economic integration with western Europe (with the help of regional aid from Brussels) will force them to adjust politically to the western European political pattern. They will be pushed (as Spain and Portugal have been; less so Greece) to act and think like western countries - open, pluralist, liberal, and bickering endlessly. There can be no variable geometry here.

Some argue that the development of EU foreign and security policy is now more important than the single market: that peace, rather than prosperity, will become the core business of the community. This is pitched too high; but the relationship with Nato on the one hand, and Russia on the other, will absorb more and more EU energy and haggling time in the years ahead. Here Britain and France are probably right; the old European religion of supra-national authority, majority voting and strong, central institutions cannot apply to foreign and security policy in the foreseeable future. The test will be to find some other way to produce a European foreign policy that is capable of sustained action, as well as words.

The overwhelming problem is the most mundane: how to adapt the EU's institutions to cope with 27 member nations? How to make them more intelligible - and more responsive - to up to 600 million EU citizens? Here variable geometry is entirely beside the point. You cannot have a Commission or Parliament reformed for a hard core few but not the many.

This is also the area fraught with the most difficulties for the British government. Majority voting by governments in the Council of Ministers is already the rule, rather than the exception. To avoid deadlock, will Britain be ready to accept an extension into new areas such as research, the environment and (more controversially) indirect taxation? The White Paper says no; the Labour Party says maybe. A Commission with 30 or more members - at least one from each state - would be an unwieldy abomination. Can member states accept a sensible alternative: a president approved by the member states, and ratified by Euro MPs, who chooses his own team, according to ability and regional balance?

Here is the central conundrum. Any movement towards democratic accountability - a directly elected commission president, more powers for Strasbourg, a more open and accountable Council of Ministers - would all tend to weaken the powers of member states. But some reforms along these lines are the essential price of coherence, even survival, for a much larger EU.

The IGC will tinker rather than address these issues wholeheartedly. The prospect for deadlock is depressingly high. Nonetheless, the terms of debate at the IGC can set the tone and pattern for much of what follows: a 10-year morass of intertwined negotiations covering EU enlargement, Nato enlargement, the EU budget, the place of other Moscow-approved bodies such as the Organisation for Security and Co-operation in Europe (OSCE). The meeting in Turin is the start of a marathon process that will decide how successfully Europeans live together in the next millennium.

And what will emerge? At worst, a weakened, lowest common denominator structure, a European league of nations doomed to collapse like a pack of cards at the first real challenge.At best, something chronically unsatisfactory and little loved, but capable of exhibiting a minimum degree of common purpose to secure the Continent's peace and prosperity.