Not so gay in Paris or so wonderful inCopenhagen

In Glasgow, 15-year-old Alan takes an overdose of his father's antidepressants and dies in the ambulance on the way to hospital. In a suburb of Paris, 16-year-old Nicole has become secretive and is refusing to eat. In Dulwich, in south London, Karen, age 12, cheerful, popular, a "highflyer", is brought home by the police after being caught stealing cosmetics from Boots. In Rome, Emilio, aged 17, is depressed. In Copenhagen, 14-year-old Anton's parents are summoned to his school - "it's something to do with drugs", his Mum says.

Such are the situations with which this report commissioned by the multidisciplinary European thinktank Academia Europaea is concerned. It asks if there are more Alans, Nicoles, Karens, Emilios and Antons in Europe today than there were 50 years ago; if so, why; and what can be done about it.

Its editors, Michael Rutter (professor of child and adolescent psychiatry at the Institute of Psychiatry in London) and David J. Smith (at the time of the report a senior fellow at the Policy Studies Institute in London and now professor of criminology at Edinburgh University), are uniquely well placed to tackle these questions. Together with a widely based team of psychologists, sociologists and psychiatrists, they look in depth at the changes over time (the "time trends" of the title) among young people in (mainly western) Europe, in a number of "target disorders" - crime, drug abuse, suicide, eating disorders and depression. They offer comprehensive overviews, including an authoritative chapter by Rutter on causal inference and detailed conclusions. In all, an impressive tour de force: some 800 pages, packed with data, argument and analysis.

What do they find? As for the first question, whether the target disorders are on the increase among young people in Europe, the short answer is yes. There are many problems even with this relatively straightforward issue, problems of definition, of reporting rates and so forth: and there are exceptions (West Germany, for instance, is an exception to the Europe-wide rise in suicide rates in young men): but taken overall the conclusion is inescapable, that young people are, in the sense of this report, more disorderly than they were.

The second question, the "why" question, is trickier. For a start, there is plenty of scope for complex causal interactions, for disorders causing each other, acting singly or together, directly or indirectly, and in straight or branching cause-effect chains.

But even without this complication, there are few firm conclusions to be drawn. Is the rise in target problems due to unemployment? No (well, overall: though unemployment may certainly affect individuals). Is it due to poverty? Again, no (at least in general, for poverty may contribute to crime and alcohol abuse; and a factor of wider significance could be the increasing gap between young people's material expectations and their fulfilment). Are we, or are our young at any rate, in moral decline? Certainly not: in some respects, indeed, to the contrary - young people, for example, are more tolerant than their elders. What about that old chestnut, the generation gap? Closing, actually. Television, then? Too complicated to say. Sex? Could be - earlier puberty and longer education equals more Romeos and Juliets. Nothing conclusive, then. There is, certainly, an important negative message here, in that many widely held beliefs about these disorders, from boot camps to alienation, simply do not hold up under detailed scrutiny. This, as Rutter and Smith make clear, is a useful preliminary, part of the ground-clearing for future research. But as for the positive message, as to just why these disorders have been on the increase, this remains, as they say, "a project for the future".

We should expect nothing less than a cautious conclusion from the Rutter research stable. Meticulous gathering of data, careful stepwise development of causal theories, testing and retesting of imaginative hypotheses are the hallmarks of his research. His work, and the work of his many research teams, has been the paradigm for psychiatry of what the philosopher Thomas Kuhn called "normal science".

Yet if the conclusions are cautious, there is perhaps a somewhat less cautious assumption behind the report as a whole, namely that research of this kind is, in and of itself, and given enough time and money, sufficient to answer the questions with which it is concerned.

Let us be clear about this. Rutter is the pastmaster of research of just this kind. His work on the causes of a wide range of disorders of childhood has transformed the fields of child and adolescent psychiatry. And in other areas, too, causal factors of the sort considered here have been linked with a number of adult disorders, as in the sociologist George Brown's work on the social origins of depression. Past profits are no guarantee of profits in the future. But on past performance, it is overwhelmingly likely that research of this kind will contribute in this area as well.

The point, though, is that in these earlier studies closely defined problems were investigated with well-validated measures by highly trained observers under localised conditions of time and place. Whereas the present study, as Rutter and Smith themselves emphasise, is innovative just in that it tackles, simultaneously, a range of probably not generic problems, across a number of rather different countries, and through a broad sweep of action-packed European history. The question, therefore, must be whether the scientific strategies so successfully employed in well-defined local skirmishes can be scaled up to the larger, the Napoleonic, requirements of a fullscale war.

And there are other research strategies available. Take methodology, for instance. There is a whole new technology for the analysis of pooled data made possible by the development of computerised searches of large international databases. These meta-analyses, as they are called, are already making an important contribution to medical research. But a key feature of this technology is quality control of the data by explicit definition of the standards of research design that the studies to be pooled must satisfy (they should be "double-blind", for instance).

Rutter and Smith tell us relatively little about this aspect of their methodology. Their team members were all experts, of course: but experts, as the history of medicine repeatedly shows, are no less liable than anyone else to the perpetuation of myths. Their experts' reports, they emphasise, were subject to careful peer review: but peer reviewers, like peers of the realm, are notoriously self-selected.

Then again, what about the choice of variables? Clearly any study has to be circumscribed. But beyond the broad scope and object-ives of the study we are given relatively little information about precisely why this or that disorder, or this or that causal factor, was in or out. Who really decided what to study and why?

Among causal factors, for example, why choose to have a whole chapter on the media, important as it may be, but include education (a topic on which Rutter has written a major report) only as a subtext? There is a practical point here. The media seem to be increasingly beyond the reach of political control, whereas education, in principle at least, can be changed. It may be, as Rutter and Smith say, that schools have only a modest effect on overall educational outcomes. But, as they go on to say, really good schools can make a critical difference to individual pupils and in particular to just those troubled pupils with which this report is concerned.

The sense of diktat is perhaps strongest in the choice of experts. The preface to the report emphasises the scientific approach. But the model of science adopted is predominantly positivist. Rutter and Smith acknowledge the importance of meanings, as well as causes, in social science research. Rutter makes this point expressly in his chapter on causal inference; and the initial working party, by which the study was set up, included a theologian and a philosopher. But there are few meanings in the report; there are few people, even. It is all graphs and tables, data and distributions.

Statistics are essential. But there are well-established techniques available for adding meaning to means, qualitative to quantitative data, significance to sign. The disciplines are there: anthropology, ethnography, linguistic analysis. So why not use them? And if these are thought to be too suspectly postempirical, what about history? There are historical data in the report. But in a study of the time trends in complex human variables, sponsored by an academy set up to support multidisciplinary research, should there not have been at least one historian among the scientists?

These are not criticisms of the report as such. Within the limits of its self-set terms of reference it is an invaluable contribution to the debate about these distressing disorders. The reviews are thorough; the discussions illuminating; and as a ground-clearing exercise the report's conclusions dispel many widely held myths.

But limits may be limitations. And the limitation of research of this kind could be crucial when we come to the third of the three initial questions - the question of what can be done about the rising rates of psychosocial disorders among young people. Rutter and Smith only touch on this. And legitimately so, up to a point. The main purpose of the study was, after all, to gather data (though with a view to advising policymakers): and the main conclusion to be drawn from its 843 pages is the negative one, that more research is needed.

But for Rutter and Smith, more research means, essentially, more of the same. We need, they say, more complete data, more countries, more detail, more sophisticated causal models. And we do.

But let us also have quality controls on the data; let us have a wider and explicitly justified choice of variables, including positive as well as negative factors, what works as well as what does not; and let us have a more representative choice of disciplines, bridging the humanities as well as the sciences, encompassing meanings as well as causes, as a basis for a genuinely multidisciplinary research strategy.

This last point could be especially important heuristically. Karl Jaspers, one of the founding fathers of modern scientific psychiatry, emphasised the significance of meanings as well as causes for our understanding even of conditions such as schizophrenia, for which a disease-like brain disorder may well be found. How much more, then, will this be true of the psychosocial disorders with which this report is concerned. But the point could also be important humanistically. The rise in the number of Alans, Nicoles, Karens, Emilios and Antons in Europe over the past 50 years has coincided rather closely with the hegemony, in public policy and politics as much as in social science and psychiatry, of a predominantly positivist philosophy. This is one of the time trends not examined by Rutter and Smith; and, well, it is just a coincidence, isn't it?

Bill Fulford is professor of philosophy and mental health, University of Warwick.