clarity, opinion, falsification and correcting mistakes
That blog is much easier to follow than the continuing on-line debate. One characteristic I've noted about Dr. Wesseley's contributions is that little becomes clearer once he is involved.
I'm tempted to branch out from the CFS debate, and launch a probe of his home territory. Does anyone know what the term "purely psychological illness" means, aside from being in DSM as the result of opinions held by professional psychiatrists? (How is this different from opinions held by professional astrologers or numerologists, who also have their own publications? Stars, planets and numbers have real objective existence.) If you do know, can you tell me how such hypotheses can be falsified? There is a real danger here of dealing with hypotheses which are immune to falsification.
The record of his profession in discarding mistaken hypotheses is not good. Science is not about avoiding mistakes, but about finding them, and learning. A vague statement that contradictory results must be the result of contamination is not enough. I've honestly tried to make such hypotheses work. Every time I make an assumption about contamination causing one result, it contradicts an assumption required to get another result from the same group.
Simply saying results which agree with your preconceptions must be right, and those which violate those preconceptions wrong, is a classic recipe for 'cherry picking' data to support a favored hypothesis. When I find five studies use a single method to discredit a study which used four distinct methods to find a virus, I do begin to wonder about motivations. The original discovery of the virus, through matches with sequence abnormalities in prostate cancer and CFS patients could be counted as a fifth method.
Whether or not the XMRV/CFS hypothesis is correct, it has produced significant results in the form of demonstration that XMRV can reproduce in human beings without prostate cancer. This is a major result independent of CFS. Researchers should never give up any line of research this productive.
Myra McClure's suggestion about drawing fresh blood from patients identified as infected, and sending this to labs with opposing views, would, in fact, resolve the central question in this debate. (I would insist the samples be blinded to guarantee that no one on the receiving end could deduce which are supposed to be positive without a valid test.) The big question in my mind now is this: why didn't those with negative results initiate this before they published? It looks like there was insufficient effort made to validate the ability to detect virus from known infected individuals. The WPI did exchange samples with labs at the NCI and CC before publication, in part because reviewers insisted. What happened during the review process of those other publications?
