著名科学知识社会学学者、英国卡迪夫大学社会科学学院的Harry Collins教授在2006年12月9日出版的英国《新科学家》杂志撰文，How we know what we know（我们是如何了解我们所了解的事物的），文章说：
Andrew Wakefield曾在《柳叶刀》医学杂志发表论文，说预防麻疹、痄腮和风疹的三联针疫苗之广泛接种，与自闭症发病率增加有关联。他这样说，是因为人们已知有一种与麻疹相关的病毒粒子会导致自闭症。但是，从来没人将这些病毒粒子与三联针疫苗联系起来。他还在一次新闻发布会上向公众说出了他的这一猜测，引起轩然大波。
Science has always struggled to sift crackpot ideas from genuine maverick genius. If it were just a matter of combining unambiguous data with flawless theories, the task would be quite simple. Unfortunately, says Harry Collins, science is an all-too-human activity, and heroes and villains come in every possible guise.
I was at a closed conference with friends and acquaintances from the gravitational wave world, a group I have studied and written about for 30 years. The dress code was informal - T-shirts and jeans, or open-necked shirts and sports jackets for the older scientists. There at the podium was a medical doctor with a halo of white hair, in a smart grey suit and red bow tie, spouting management-speak. As far as we could tell, he seemed to be telling the group that it did not understand its $250 million interferometers, and that his microscale experiments showed they had missed something vital about the interaction of the mirrors. I swapped smirks with those around me. But we were wrong. A couple of stubborn scientists extracted the important bit of sense from the packaging, and the design was changed.
If science were a matter of combining unambiguous data from perfectly conducted experiments with flawless theories, assessing the claims of "outsider" scientists and their maverick ideas would not be that hard. But the logic of science is not so far removed from the logic of ordinary life (though admittedly ordinary life lived among extraordinary ideas and amazing machines) and so fallible human judgement still determines what happens at the heart of even the hardest science.
Since the 1960s, the tension between the canonical model of science (experimentation and theory) and the everyday practice of science (pottering around and hunch) has been explored by historians, philosophers and especially sociologists. One way to get a sense of how that tension compounds the problem of dealing with outsiders and their ideas is by comparing funding policies.
Consider, for example, that the US military spends around $1 million per year on anti-gravity research. This is best understood by analogy with the philosopher Blaise Pascal's famous wager. Everyone, he argued, should believe in God because the cost of believing was small while the cost of not believing could be an eternity in hell. For the military, the cost of missing a technological opportunity, particularly if your enemy finds it first, is a trip to the hell of defeat. Thus goes the logic of the military investigating anti-gravity.
The same rationale led to the funding of some unlikely experiments by University of Maryland physicist Joe Weber. Weber founded gravitational-wave research, but failed to find the waves. By 1975, he had lost most of his credibility, yet in the 1980s, the US military paid to test an implication of a theory he had invented in his struggle to regain his former glory. The implication was that it was possible to detect neutrinos emitted by the reactor of a nuclear submarine, using only a crystal you could hold in your hand rather than vast underground tanks. If the US navy ignored Weber's idea but the Russians used it to build technology that could detect US nuclear subs, the undersea deterrent would no longer deter.
The same logic drives companies such as Pirelli, which is still pursuing Weber's idea in the hope of transmitting modulated neutrino signals through the Earth to carry messages without cables (New Scientist, 17 April 2004, p 36), and Canon and Toyota, which funded cold fusion long after the research councils would not touch it.
The logic of state funding agencies such as the US's National Science Foundation or the UK's Engineering and Physical Science Research Council is quite the opposite. Here the pressures of accountability can make them too conservative. There are a few new agencies aiming to fill the ground between the wild logic of the military and certain companies, and the worthy work of "official" science. One such is Donald Braben's Venture Research, which plans to back projects by picking promising researchers rather than wading through research applications (see his book Pioneering Research, Wiley, 2004).
But even multiple funding styles cannot address the problem that it is impossible to explore every new scientific idea to the standard set by science: there are just too many. Many scientists in the public eye are deluged by self-styled pioneers claiming to have found the fundamental flaw in relativity or a new energy-free method of transport. Discriminating between them requires a mixture of the "green-ink and no margins" test, and more sophisticated short-cuts, such as knowing that papers in Physical Review that describe new physics tend to be taken more seriously than articles in New Scientist doing the same, and that scientists from major institutions are trusted more than scientists from obscure colleges. No one has time to track everything to the bitter end.
Ironically, even exploring an idea to the bitter end may prove impossible. For example, after a hundred years, no one has absolutely proved the non-existence of extrasensory perception. If anything, the findings run very slightly in its favour. Weber's claims that his early detectors caught high fluxes of gravitational waves and that his crystals could detect neutrinos have also never been disproved to the standards of logic. The reason that the US military finally abandoned both strands of research was sociological: Weber had had a good run and it was time to move on. There is no logic that says that Pirelli is completely crazy for taking another look.
The irreducible tension in science is to maintain enough "social control" over new ideas and spending to ensure science isn't engulfed by seas of possibilities, while leaving room for new ideas. Tossed on the waves of these possibilities are people such as Martin Fleischmann (cold fusion), Eric Laithwaite (anomalous gyroscopes), Albert Einstein (relativity), Linus Pauling (vitamin C), Alfred Wegener (plate tectonics), Thomas Gold (origin of oil), David Duesberg (non-viral causes of AIDS), and Subrahmanyan Chandrasekhar (black holes). The list goes on and on. Such ideas eventually wash up on one shore or the other, but only 20:20 foresight will tell you which one.
However, sometimes what looks like "outsider science" has content of quite a different sort. Take the measles, mumps and rubella (MMR) vaccine and autism affair in the UK. Andrew Wakefield, the doctor behind the furore, published some evidence in The Lancet suggesting a link between autism and measles-related virus particles in the gut. But these particles were never linked to MMR vaccine. There was word-of-mouth testimony from some parents, but no link between MMR and autism has ever been proved. Wakefield simply speculated about a relationship at a press conference - and no one has ever gone further than to hypothesise about it.
This case was presented to the public as a genuine scientific controversy, and, to my discomfort, the MMR story as told by many social scientists is one of struggle between wise parents and uncomprehending, authoritarian medical authorities. There have been (and will be) many real struggles of that sort, but this was not one of them. The only usable scientific evidence was epidemiological - and that pointed to the safety of MMR. Because it is so hard to prove a negative, none of this shows that there is not a hidden link between MMR and autism lurking below the statistics. But there is no evidence to show there is.
The energy it took to deal with Wakefield's claims, and to persuade parents to vaccinate their children at all after the scare, could have been much better spent. Wakefield was not behaving as a scientific outsider: he was simply not providing scientific evidence at that press conference.
In addition to the difficulty of proving a negative, scientists are also very unwilling to face up to the social and financial logic that drives their choices. A tentative claim about, say, telepathy, can provoke a sort of fundamentalist zeal among some scientists refuting the claim, which in turn undermines their claims for science as an exemplar in a divided world. They should say merely this: "Well, it's not inconceivable, I can't absolutely prove you wrong, but my time is better spent doing things I judge to have more potential."
Scientists, then, are not always their own best friends when it comes to helping others navigate the loss of absolute certainty about our world. I am also not sure how it helps if they assume omnipotence in the name of science, as Richard Dawkins did recently when he insisted that scientists must be atheists. And Stephen Hawking has been turned into a new kind of religious icon, with his books taking the place of the incomprehensible Latin Bible in our homes.
Here science becomes "revealed truth", obscuring the long hours of tedious work, the experiments open to reinterpretation (and failure), and theories with their infinities and arbitrary variables that can never quite be tamed. The Dawkinses and the Hawkings threaten to make the hard-won victory of science over religion a pyrrhic victory by replacing old faiths with new.
If science is essentially ordinary life albeit conducted in extraordinary circumstances, it must contradict literal interpretations of texts that clash with its findings, but it should not claim the right to address deeper questions of existence.
The biggest danger for science is that in missing its footing on the tightrope of certainty, it crashes to the ground. In the social sciences, this danger is best represented by the romantic value today placed on the instincts of the general public: the folk are said to be as wise, or wiser, than experts. It is a political necessity and responsibility in a democratic society to take account of the technological "preferences" of the people, but this should never be confused with technological or scientific "wisdom". That road leads to a society none of us would want to inhabit.
There is no easy and sure scientific way to sift every claim, but there are good and bad judgements. That is the safety net protecting us against scientific populism. This populism is a way of evading the hard search for the grounds of knowledge by giving equal weight to everyone's frame of reference. We must keep hold of the idea that judgement, though never perfect, is generally done better by those who know what they are talking about.
From issue 2581 of New Scientist magazine, 09 December 2006, page 46-48
Harry Collins is Distinguished Research Professor at the school of social sciences of Cardiff University, UK. His books include: The Golem: What everyone should know about science (with Trevor Pinch, published by Cambridge University Press); Gravity's Shadow, and Dr Golem (with Trevor Pinch), published by University of Chicago Press. His next book (with Rob Evans) will be Rethinking Expertise.
Clear evidence of falsification of data should now close the door on this damaging vaccine scare
“Science is at once the most questioning and . . . sceptical of activities and also the most trusting,” said Arnold Relman, former editor of the New England Journal of Medicine, in 1989. “It is intensely sceptical about the possibility of error, but totally trusting about the possibility of fraud.”1 Never has this been truer than of the 1998 Lancet paper that implied a link between the measles, mumps, and rubella (MMR) vaccine and a “new syndrome” of autism and bowel disease.⇓
Authored by Andrew Wakefield and 12 others, the paper’s scientific limitations were clear when it appeared in 1998.23 As the ensuing vaccine scare took off, critics quickly pointed out that the paper was a small case series with no controls, linked three common conditions, and relied on parental recall and beliefs.4 Over the following decade, epidemiological studies consistently found no evidence of a link between the MMR vaccine and autism.5678 By the time the paper was finally retracted 12 years later,9 after forensic dissection at the General Medical Council’s (GMC) longest ever fitness to practise hearing,10 few people could deny that it was fatally flawed both scientifically and ethically. But it has taken the diligent scepticism of one man, standing outside medicine and science, to show that the paper was in fact an elaborate fraud.
In a series of articles starting this week, and seven years after first looking into the MMR scare, journalist Brian Deer now shows the extent of Wakefield’s fraud and how it was perpetrated (doi:10.1136/bmj.c5347). Drawing on interviews, documents, and data made public at the GMC hearings, Deer shows how Wakefield altered numerous facts about the patients’ medical histories in order to support his claim to have identified a new syndrome; how his institution, the Royal Free Hospital and Medical School in London, supported him as he sought to exploit the ensuing MMR scare for financial gain; and how key players failed to investigate thoroughly in the public interest when Deer first raised his concerns.11
Deer published his first investigation into Wakefield’s paper in 2004.12 This uncovered the possibility of research fraud, unethical treatment of children, and Wakefield’s conflict of interest through his involvement with a lawsuit against manufacturers of the MMR vaccine. Building on these findings, the GMC launched its own proceedings that focused on whether the research was ethical. But while the disciplinary panel was examining the children’s medical records in public, Deer compared them with what was published in the Lancet. His focus was now on whether the research was true.
The Office of Research Integrity in the United States defines fraud as fabrication, falsification, or plagiarism.13 Deer unearthed clear evidence of falsification. He found that not one of the 12 cases reported in the 1998 Lancet paper was free of misrepresentation or undisclosed alteration, and that in no single case could the medical records be fully reconciled with the descriptions, diagnoses, or histories published in the journal.
Who perpetrated this fraud? There is no doubt that it was Wakefield. Is it possible that he was wrong, but not dishonest: that he was so incompetent that he was unable to fairly describe the project, or to report even one of the 12 children’s cases accurately? No. A great deal of thought and effort must have gone into drafting the paper to achieve the results he wanted: the discrepancies all led in one direction; misreporting was gross. Moreover, although the scale of the GMC’s 217 day hearing precluded additional charges focused directly on the fraud, the panel found him guilty of dishonesty concerning the study’s admissions criteria, its funding by the Legal Aid Board, and his statements about it afterwards.14
Furthermore, Wakefield has been given ample opportunity either to replicate the paper’s findings, or to say he was mistaken. He has declined to do either. He refused to join 10 of his coauthors in retracting the paper’s interpretation in 2004,15 and has repeatedly denied doing anything wrong at all. Instead, although now disgraced and stripped of his clinical and academic credentials, he continues to push his views.16
Meanwhile the damage to public health continues, fuelled by unbalanced media reporting and an ineffective response from government, researchers, journals, and the medical profession.1718 Although vaccination rates in the United Kingdom have recovered slightly from their 80% low in 2003-4,19 they are still below the 95% level recommended by the World Health Organization to ensure herd immunity. In 2008, for the first time in 14 years, measles was declared endemic in England and Wales.20 Hundreds of thousands of children in the UK are currently unprotected as a result of the scare, and the battle to restore parents’ trust in the vaccine is ongoing.
Any effect of the scare on the incidence of mumps remains in question. In epidemics in the UK, the US, and the Netherlands, peak prevalence was in 18-24 year olds, of whom 70-88% had been immunised with at least one dose of the MMR vaccine.2122 Any consequence of a fall in uptake after 1998 may not become apparent until the cohorts of children affected reach adolescence. One clue comes from an outbreak in a school in Essen, Germany, attended by children whose parents were opposed to vaccinations. Of the 71 children infected with mumps, 68 had not been immunised.23
But perhaps as important as the scare’s effect on infectious disease is the energy, emotion, and money that have been diverted away from efforts to understand the real causes of autism and how to help children and families who live with it.24
There are hard lessons for many in this highly damaging saga. Firstly, for the coauthors. The GMC panel was clear that it was Wakefield alone who wrote the final version of the paper. His coauthors seem to have been unaware of what he was doing under the cover of their names and reputations. As the GMC panel heard, they did not even know which child was which in the paper’s patient anonymised text and tables. However, this does not absolve them. Although only two (John Walker-Smith and Simon Murch) were charged by the GMC, and only one, the paper’s senior author Walker-Smith, was found guilty of misconduct, they all failed in their duties as authors. The satisfaction of adding to one’s CV must never detract from the responsibility to ensure that one has been neither party to nor duped by a fraud. This means that coauthors will have to check the source data of studies more thoroughly than many do at present—or alternatively describe in a contributor’s statement precisely which bits of the source data they take responsibility for.
Secondly, research ethics committees should not only scrutinise proposals but have systems to check that what is done is what was permitted (with an audit trail for any changes) and work to a governance procedure that can impose sanctions where an eventual publication proves this was not the case. Finally, there are lessons for the Royal Free Hospital, the Lancet, and the wider scientific community. These will be considered in forthcoming articles.
What of Wakefield’s other publications? In light of this new information their veracity must be questioned. Past experience tells us that research misconduct is rarely isolated behaviour.25 Over the years, the BMJ and its sister journals Gut and Archives of Disease in Childhood have published a number of articles, including letters and abstracts, by Wakefield and colleagues. We have written to the vice provost of UCL, John Tooke, who now has responsibility for Wakefield’s former institution, to ask for an investigation into all of his work to decide whether any more papers should be retracted.
The Lancet paper has of course been retracted, but for far narrower misconduct than is now apparent. The retraction statement cites the GMC’s findings that the patients were not consecutively referred and the study did not have ethical approval, leaving the door open for those who want to continue to believe that the science, flawed though it always was, still stands. We hope that declaring the paper a fraud will close that door for good.
Competing interests: All authors have completed the Unified Competing Interest form at www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declare: no support from any organisation for the submitted work; no financial relationships with any organisations that might have an interest in the submitted work in the previous three years. HM chairs GMC fitness to practise panels. He had no association with the Wakefield hearings and the views expressed in this article are his own and do not represent those of the GMC.
Provenance and peer review: Commissioned; not externally peer reviewed.
Schechter AN, Wyngaarden JB, Edsall JT, Maddox J, Relman AS, Angell M, et al. Colloquium on scientific authorship: rights and responsibilities. 1989;3:209-17.;:-17.
Wakefield AJ, Murch SH, Anthony A, Linnell, Casson DM, Malik M, et al. Ileal lymphoid nodular hyperplasia, non-specific colitis, and pervasive developmental disorder in children [retracted]. 1998;351:637-41.;:-41.
Black C, Kaye JA, Jick H. Relation of childhood gastrointestinal disorders to autism: nested casecontrol study using data from the UK General Practice Research Database. 2002;325:419-21.;:-21.
Taylor B, Miller E, Lingam R, Andrews N, Simmons A, Stowe J. Measles, mumps, and rubella vaccination and bowel problems or developmental regression in children with autism: population study. 2002;324:393-6.;:-6.
Madsen KM, Hviid A, Vestergaard M, Schendel D, Wohlfahrt J, Thorsen P, et al. A population-based study of measles, mumps, and rubella vaccination and autism. 2002;347:1477-82.;:-82.
The editors of the Lancet. Retraction—Ileal-lymphoid-nodular hyperplasia, non-specific colitis, and pervasive developmental disorder in children. 2010;375:445.;:.
Transcripts of hearings of fitness to practise panel (misconduct) in the case of Wakefield, Walker-Smith, and Murch, 16 July 2007 to 24 May 2010. GMC; 2010.
Kutty PK, Kruszon-Moran DM, Dayan GH, Alexander JP, Williams NJ, Garcia PE, et al. Seroprevalence of antibody to mumps virus in the US population, 1999-2004. 2010;202:667-74.;:-74.
