Simon Baron-Cohen: Ali G's smarter cousin and Britain's leading expert on autism

George Osborne, the Shadow Chancellor, once amused journalists at a Conservative fringe meeting by regaling them with tales of his ability to retain odd facts. Perhaps, one of them asked, he was "faintly autistic"? Referring in turn to a much-touted rumour that had been doing the Westminster rounds, Osborne replied: "We're not getting on to Gordon Brown yet."

Osborne was blasted from all sides because he seized on the word "autistic" to deliver a cheap jibe. He tried to argue that he was merely caught on the hop because someone had just suggested that he showed autistic traits himself. But no dice. Nick Hornby, the writer whose first son was diagnosed with the brain development disorder, spoke for many when he declared: "George Osborne doesn't seem to have noticed that most people over the age of eight no longer use serious and distressing disabilities as a way of taunting people."

Yet though the joke may have been in poor taste, the set-up was mundane. The truth is that "autistic" has entered the international vocabulary, and people generally feel quite comfortable in speculating that anyone who displays either the mildest of social difficulties, or the most gently unusual of mental capacities, may be "a bit autistic".

This widespread awareness of the condition's existence, and its commonplace inclusion in everyday speech, has been a major generational sea-change. Twenty-five years ago, when Professor Simon Baron-Cohen first started working with autistic children, people used to routinely mishear him when he talked about his work, and assume that the children he was teaching were "artistic".

As a young graduate in human sciences, Baron-Cohen had a job in a small unit called Family Tree, which employed six dedicated teachers to educate its six demanding pupils. Baron-Cohen found the experience intense and fascinating, and ever since he has worked in the field of "autistic spectrum disorders" – which also includes autism's "high-functioning" cousin Asperger's syndrome – conducting many influential studies himself. Married with three children – and first cousin to the comedian Sacha Baron Cohen (who has somehow lost his hyphen) – he is presently head of Cambridge University's Autism Research Centre.

Ensconced in his small, cosy office in a 1930s block tucked away behind the ancient splendour of Trinity College, Cambridge, Baron-Cohen talks with quiet, measured passion about the trouble that imprecise talk about autism causes. "If the term is used too loosely," he explains, "it can lose its meaning as a medical diagnostic term. I think it should be reserved for people who are suffering, because that's when you get the diagnosis. When the features of autism are interfering with your life to such an extent that you're suffering, and you're seeking help and you take yourself off to a clinic and get a diagnosis, that's the only time when the term should be used."

Nevertheless, the idea that lots of people are wandering around unaware that they are on the autistic spectrum prevails. There are good reasons for this. Diagnosis is made far more frequently now than it was when Baron-Cohen first started out (Asperger's syndrome was virtually unheard of in Britain in the 1980s), and that is put mainly down to better detection. Yet some experts believe that autistic spectrum disorders, which affect about six in every thousand people, remain under-diagnosed. As for the overwhelming evidence that the disorder is more common in males, about which Baron-Cohen has written with intelligence and insight himself, even that is now being questioned, as further research suggests that girls are particularly vulnerable to failure of diagnosis.

Like most of those working in autism research and treatment, Baron-Cohen hopes for a breakthrough in the ability to diagnose autism using "biomarkers" (biological indicators of a particular condition) and closely follows developments, frequently reported, which may be moving towards achieving a genuine critical mass. However, it is unlikely that a "smoking gun" will ever be discovered.

One piece of new research points to flaws in proteins called "cell adhesion molecules", and suggests this could account for up to 15 per cent of cases. Another study, led by the same researcher, Dr Hakon Hakonarson, has identified deletions or duplications of DNA that increase the risk of autism. But significant as these findings are, they all point to autistic symptoms being a complex accumulation of a number of different factors that are hard to disentangle.

"All the biomedical research that is going on into autism – at the level of genetics, proteins, hormones – involves the search for so-called biomarkers, because at the moment you diagnose autism on the basis of behaviour not biology. You interview the child or the adult, you interview their family, you observe them, but it's all behavioural and observational. It is argued that biomarkers might make diagnosis more precise.

"Anything to improve our diagnosis would be good because at the moment diagnosis is based on clinical judgement, which is ultimately a bit subjective. Although people have tried to use more standardised methods to make it more objective, there is still room for ambiguity: you might get fairly good agreement between two independent physicians but in some cases there is still room for doubt. Whereas for other medical conditions, where it's based on a biological marker, as is the case for Down's syndrome, if you have the extra chromosome, then it's much more black or white."

Yet even important research breakthroughs, or the way they are reported at least, can bring problems of their own. At the beginning of the year, Baron-Cohen's own centre published research based on the study of 235 typically developing children from birth until the age of eight. It found that high levels of testosterone in the amniotic fluid of pregnant women was linked to autistic traits in the children at follow-up. Those findings were misconstrued as a signal that in-utero testing for autism spectrum disorders might be a plausible near-future development. The assumption was mistaken, but it still upset and alarmed many of those living with the diagnosis, their families – and Baron-Cohen himself.

"If we get to a point in the future where we know enough about the biology to be able to say, 'Yes this person has it', or 'No they don't', then it might be important to have the ethical debate about what we would do with that kind of knowledge when we have it. Could it be misused? If there was a pre-natal screening test, would parents use it to make a decision about termination at some point?"

What worries Baron-Cohen is that talk of genetic markers sounds cut and dried. Yet, as he says, "even the hard-line genetic theorists recognise that the environment is important. Genes are thought to contribute a certain amount to the cause of autism but it's not 100 per cent. It might be about 60 per cent genetic. So there are going to be environmental factors that mediate the impact of autism."

This, in turn, means "you could have someone who has all of the genes that denote a predisposition, but who never develops the syndrome. Diagnosis also relies on the patient or their family saying, 'I need help'."

Baron-Cohen thinks that the media concentrates too tightly on "the very talented people with autism", but agrees that the disorder can sometimes bring with it "amazing gifts, that are to be celebrated".

He talks of Daniel Tammet, who found his niche by performing memory stunts and this year published a well-received book, Embracing the Wide Sky: A Tour Across the Horizons of the Mind. Tammet, as a 13-year-old, made no eye-contact, had no friends, but as an adult has made his unusual qualities work for him, and has "outgrown" his need for a diagnosis.

Ideally, for Baron-Cohen, when a person does seek and get a diagnosis, there should also be support services in place. Baron-Cohen remains committed to developing a "good model" for helping people with autism, and emphasises the importance of helping people adjust to the environment that may not suit them as well as it suits "neurotypical" people.

Certainly, objective diagnostic markers would silence the most aggressive critics of autism, of which there are some. Michael Savage, the US shock jock that no one in Britain had heard of until the Home Secretary Jacqui Smith banned him from entering the country, last year described autism as "a fraud, a racket ... in 99 per cent of the cases, it's a brat who hasn't been told to cut the act out".

Further, an absence of detailed understanding about genetic markers for autism, and a speculative grasp of which environmental factors may be pertinent, makes autism an easy target for the sort of scaremongering that occurred during the MMR debacle. Then, a tainted piece of research by Andrew Wakefield and colleagues, hysterically publicised, led many people to refuse a triple vaccination for their baby, for fear that it could trigger the condition. This affair found expression in mainstream culture, with no less an actor than Juliet Stevenson appearing on prime-time television, playing a mother who refused to believe that her son's autism had not been caused by his injection, despite widespread scientific rejection of the theory.

As partisan as this portrayal of autism was, it joins a recent but intense media and cultural tradition that has fostered polished and sustained presentations of the issues around autism spectrum disorders, most famously in the Dustin Hoffman movie, Rain Man. All this, in turn, has fed a public awareness of and fascination with autism for which most sufferers and most of those involved in working with autistic spectrum disorders are broadly grateful.

Yet the MMR scandal is just one example of a mass of undisciplined and sometimes highly damaging speculative developments that autism has attracted. Baron-Cohen, intellectually rigorous, but also highly empathetic, is sensitive to the vulnerabilities of people with autism, and also to the stress that living with the condition sometimes causes in parents. His own professional standard is only to consider treatments that have been proven to help and that do not have unwanted side-effects. Others are not so scrupulous.

"You'll always come across these anecdotes – 'This cured my child'," says Baron-Cohen. "So doctors who are pushing something that hasn't yet been put through a proper evaluation can lead parents down a blind alley, or put children through unethical procedures for no obvious benefit.

"There was a lot of publicity around a treatment called Holding Therapy in the 1980s. You took the child and you forced them to be hugged. From the outside, it looked very upsetting because the child was pushing to get away and was crying. The practitioners said, 'Well, you've got to go through this for the child to be recovered, or saved, from their autism'. Some of these treatments seem quite extreme. That treatment quietly went away. I don't even know it if ever went through a formal evaluation.

"Another treatment appeared in the early 1990s, called Facilitated Communication, and it was very big in the States, to such an extent that every school for autism in the US was supposed to have someone who could provide this, and they were spending millions. It involved having a specially trained facilitator, who holds the child's hand over a computer keyboard, and children who were supposedly mute were alleged to start communicating by typing, even though they had low IQ and limited language.

"That treatment did go through a formal evaluation and it was found that when the facilitator didn't know what the child was wanting to request or say, the messages came out pretty meaningless. So that's the value of formal testing.

"And just a few years back, there was a drug treatment using Secretin. Parents were going abroad and spending thousands to get hold of this hormone, which was meant to be very good for dietary issues. There was a plausible link because children with autism often have gastro-intestinal problems.

"But there was also a claim that the children were developing language and developing socially as a result of these injections, so there had to be a proper trial. When a double-blind treatment trial was carried out it was found that Secretin was no better than a placebo."

Put in this context, jibes like that of George Osborne, or even ignorant refutations like that of the shock jock Michael Savage, are a distraction. The search for understanding of the causes of autism is important. But the continuing development of properly evaluated treatments and useful support is paramount.

Baron-Cohen's hope is that autism will eventually come to be understood as "part of the individual differences between people. To me, there's already a big shift from the old view. When I first started in this field there were all kinds of stereotypes about autism, as if these were children from another planet, or children who had been brought up by wolves, that they weren't part of our population and were somehow separate.

"The great benefit that comes out of talking about it more is that autism or Asperger's syndrome or social communication difficulties become a bit more ordinary, like dyslexia. When we hear that someone has dyslexia we no longer recoil in shock or think it's something scary. We just think, 'Well, we know that this is very common, and they need extra time in exams or need a different kind of reading programme when they're being taught to read.'

"I'm hoping that autism is going to get to that same point, where it becomes quite ordinary to say, 'I have autism', or 'I have Asperger's syndrome', and that there will be many more resources available to make life easier for people on the autistic spectrum. If your child has dyslexia, there are lots of resources available. Parents have choices and feel there are avenues they can try.

"What we don't know yet is whether in the right environment even people with classic autism might be able to calm down a lot, and be able to learn more than they otherwise might. Somebody with Asperger's syndrome once put it very well. He said: 'We are like freshwater fish in salt water'. So people with autistic spectrum disorders are just in the wrong environment, struggling and sometimes not managing. 'If you put us in the right environment,' he said, 'we can function very well.'" Who could possibly take issue with such a pragmatic, humane approach?