While some scientists are excited by the latest research into the causes of schizophrenia, others are beginning to wonder whether it exists at all. Aisling Irwin reports. Mary Boyle, principal lecturer in psychology at the University of East London, has been doing some contentious research. She has studied a wide group of people who all share the same delusion.

They share it despite the fact that a 100-year search for proof of what they believe has, she thinks, produced virtually nothing to support their convictions. Her guinea pigs are mainly psychiatrists and neuroscientists. Their delusion is their belief in the existence of schizophrenia. Boyle claims that some of them are even dishonest in order to try to convince others .

Reactions to Boyle's research have been cramming the letters pages of The Psychologist. They are responses to an article she wrote that likened schizophrenia research to the "degenerating research programmes" described by the philosopher of science Imre Lakatos.

The responses have been mainly sympathetic: schizophrenia has "the epistemological status of Father Christmas", wrote David Pilgrim, senior research fellow at the Health and Community Care Research Unit in Liverpool. "There is still enough muddle and difference of opinion . . . to give Boyle good reason to question whether research in the area, as presently defined, is getting anywhere," wrote Gordon Claridge, of Magdalen College Oxford.

ÖйúAƬ

Schizophrenia as a label was invented 99 years ago. Theories of its cause are for many the symbols of ideological positions - it is a political football as well as a collection of debilitating symptoms. If it is not really an illness, for example, then patients avoid being stigmatised and there is justification for caring for them in the community, says Marjorie Wallace, chief executive of the charity Sane. If it is socially caused then a patient's family immediately feels pinned with a badge of blame.

Boyle argues, firstly, that there is no evidence to support taking a cluster of diverse symptoms (hallucinations, thought disorders, negative symptoms such as social withdrawal) and packaging them under a single heading.

ÖйúAƬ

In this she is supported by Richard Bentall, professor of clinical psychology at Liverpool University: "The concept is completely hopeless," he says. "It has been seriously misleading for the past 100 years." He argues that the category "schizophrenia" should be abandoned and the individual symptoms should be investigated instead.

There are nine definitions of schizophrenia, Bentall says. One researcher found that the number of schizophrenia patients he had varied from four to 200 depending on which definition he used. Each definition is reliable: arm a selection of scientists with one of the definitions and they will each pick out the same patients from a group. But it is not enough for a definition to be reliable. "I could tell you about Bentall's disease, in which sufferers have red hair, long fingernails and more than four Pink Floyd records," says Richard Bentall. "That's a reliable definition - but it's not valid." The term schizophrenia will never be valid because thought disorder is not connected with hallucinations and delusions; and negative symptoms do not correlate with either of them, he says.

A second layer of criticism is that there is no evidence that schizophrenia symptoms have a physical cause. Boyle says: "We don't have this assumption of a physical basis with most psychological problems. We look at social worlds, family or other psychological factors." She says the search for a physical basis has as much justification as a search for "fairies at the bottom of my garden".

Thirdly, Boyle questions why we call some psychological symptoms defects while others escape criticism or even receive praise. "The assumptions we make about disturbing behaviour and what we consider to be important questions to ask about it, can have profound effects on people's lives," she writes in The Psychologist.

Paradoxically, the very research world that Boyle is criticising is buzzing with excitement that new insights into schizophrenia may be just around the corner. Sane has funded a research centre in Oxford, which opened this week. Elsewhere scientists have sharpened their tools for probing the underlying genetics; researchers are investing in high-tech brain imaging machines; and new drugs are less unpleasant. Surely, say the researchers, this optimism can only spring from success in studying a very real disease?

One area of fascination is high-tech brain imaging, which is booming in London. Chris Frith, senior scientist in the clinical neuroscience section of the Medical Research Council Cyclotron Unit, says that scanning the brain can combat a perpetual problem faced by experimental psychologists - that of investigating subjective events. "These techniques are exquisitely sensitive to mental activity occurring in the absence of overt behaviour," he writes in The Psychologist.

Imaging has been used to investigate the inner voices that some schizophrenia sufferers hear. The theory is that most of us have a mechanism in our brains which tells us whether a sentence is our own thought or has come from a voice outside us. This mechanism may be faulty in shizophrenics.

Scientists have put sufferers in brain scanners and waited for them to hear voices. The scans show that the "inner speech" centre of the brain lights up at the very moment that a sufferer is having an auditory hallucination, says Frith.

ÖйúAƬ

Several critics have said that it would be more exciting if the inner speech centre was found not to light up during a hallucination. Boyle is unmoved by the discovery: "I don't know how to say anything about this and be polite. Finding out what happens when people hallucinate is not important. It's like studying a cough in order to understand tuberculosis. The question is why should people choose to externalise some thoughts and not others."

In a different research area there is another group of optimistic scientists. Tim Crow of Oxford University and John Done of the University of Hertfordshire have been exploiting a study of a cohort of babies born in the same week in 1958. At as early as seven, there were hints that certain children would go on to develop schizophrenia: they were more likely to be anxious and hostile; by 11 they were more likely to be withdrawn and depressed.

ÖйúAƬ

But what is exciting Crow is the number of such children who fail to become either right or left-handed. In the brains of most right-handed people the left hemisphere is dominant for speech. Could it be that if you don't develop "handedness" you have not developed sufficient brain asymmetry? "It's very early data," says Crow. "Clearly there are a lot of people in that distribution who don't develop the psychosis."

Crow directs Sane's new centre. He is convinced that the findings imply that schizophrenia is genetically caused, because they have managed to control to a large extent for family background. There is evidence that schizophrenia occurs in clusters in families. Twin studies have been used to look for clues to heredity. Verdicts differ: Robin Murray of the Institute of Psychiatry says: "There is no question that there is a genetic contribution. If one twin has it then the other will have it in 40 per cent of cases. The other 60 per cent must be due to environmental factors." Peter McGuffin, professor of psychological medicine at the University of Wales, says: "Our most recent estimate is that about 80 per cent of it is genetically caused."

Professor McGuffin's "80 per cent genetic" prediction leaves 20 per cent to the environment. But "environment" can mean a range of influences other than social ones. "It can mean a virus, a blow to the head, it could mean random ways in which the brain gets wired up," he says. The "environment" could be an internal one generated by some gene malfunction. There are many tricks that DNA can get up to. "My guess is that the remaining 20 per cent will be one of these," says McGuffin.

But critics of the genetic programme hit it right at its roots, questioning the data. Boyle says: "There is no question that if someone in your family has been diagnosed then you also run a risk of being diagnosed. But these probabilities vary by a factor of about 30 (from study to study). Studies have either produced no results or they have been methodologically very bad. The data is misrepresented when presented to the public."

Another group of scientists is excited about developments in drug therapy for schizophrenia. Clozapine is a new drug given to those for whom conventional drugs have failed. Such sufferers often spend long periods in hospital with severe hallucinations, paranoia and delusions. Clozapine improves symptoms for 60 per cent of these people: half of those recover sufficiently to be discharged from hospital.

Critics acknowledge the success of these drugs but claim that they treat individual symptoms rather than the syndrome as a whole. This is demonstrated by the fact that it is not possible to predict which patients will respond to the drugs, says Bentall.

But the critics are not simply being negative. Bentall says: "My solution to the whole problem comes from the fact that people never complain of having schizophrenia but of particular symptoms. They say : 'I hear voices' or 'I feel persecuted by my neighbours'. I think the way forward is to study the causes of these individual clusters of behaviour."

He has done an ingenious experiment with paranoid patients. These people blame all their problems on outside circumstances to an exaggerated extent, never blaming themselves and therefore apparently having high self-esteem.

Bentall has asked them sneaky questions that seem to be factual but in fact lure them into deciding who is at fault for an incident. He has discovered that the paranoids unconsciously blame themselves for everything.

This approach is also being used at Birmingham University, where it is leading to experimental therapies for hallucinations. Bentall says that his ideas are beginning to win the ear of biological researchers. When challenged, some at least do not reject his theories.

Professor Frith claims that many scientists have already "set the more achievable aim of explaining specific signs and symptoms associated with schizophrenia". McGuffin calls schizophrenia "a useful working hypothesis"; Murray prefers to call it a syndrome rather than a disease. If they are to satisfy their psychologist critics they will have to drop the word "useful" and transform their study of "specific symptoms associated with schizophrenia" into a study of phenomena that exist in their own right.

Next year the concept will be 100 years old. Odds on its survival are mixed.

ÖйúAƬ