A Crisis of Faith or "Bait and Switch"?
But now, the shift in emphasis is unmistakeable. Kev, Orac and others have blogged about specific attempts at historical revisionism among the chelationistas, but I'd like to assess the "bigger picture" and what it means to the hypothesis-formerly-known-as-mercury-causes-autism.
For those who are not intimately familiar with the daily ins and outs of the chelationistas' arguments (you lucky people!), the Central Dogma of the chelationista faith has been - since the infamous Bernard et al paper, in which a group of people who had never seen mercury poisoning compared it to autism - that mercury causes autism. Period. Full stop. The end.
Apparently, that is no longer true.
In a move reminiscent of George Orwell's 1984, the chelationistas have changed their minds about mercury being the cause of autism, but they are claiming that they haven't actually changed their minds - they're simply "clarifying their position".
That's newspeak for "the data don't support our position, so we're doing our darndest to keep from having to admit we were wrong."
So, what does cause autism, according to the chelationistas?
- "Heavy metals" (does this include Def Leppard?)
- Aluminium (not a "heavy metal")
- Live viruses
- "Maternal toxic load" (perhaps from watching Oprah?)
The advantage of this "clarification" (translation: "obfuscation") is to make it impossible (or at least highly impractical) to ever prove them wrong. Let me demonstrate.
In order to show that the mercury-causes-autism hypothesis was wrong, we just needed to show that there was no correlation between mercury exposure (i.e. thimerosal exposure, environmental mercury, etc.) and autism. While correlation does not equal causation, a lack of correlation is essentially equal to a lack of causation.
That was relatively easy to do, although many of the chelationistas are still struggling with that unpleasant lump of reality.
However, in order to definitively squash the new "heavy metals"-aluminium-live viruses-bacteria-etc.-cause-autism hypothesis, you would need to explore all possible combinations of these "exposures". If you don't, it is a sure bet that the next "hypothesis" will be that the one combination you didn't check is the real cause of autism.
Just like mercury was... until it wasn't.
So, how many combinations would that be? It seems simple enough - there are only six items on the list. Except that some of the items are a bit broad.
You could lump all "heavy metals" together (if you can get a ruling from the chelationistas on exactly which metals are included in this group), which would leave you with a manageable number, right?
The "live viruses" could just be the ones in vaccines, or they could be any live virus that the chelationistas care to claim (without having to provide data) are in vaccines. Or they could be any live virus in "the environment". It is a number with a great potential for flexibility.
"Bacteria" is another ambiguous category. Are they referring to the killed bacteria in vaccines, or will they want to expand that into all environmental bacteria (a number in the millions, if not billions)? Or will they want to reserve the right to claim that vaccines - or any other product with a manufacturer they can sue - are contaminated with bacteria that only the DAN! doctors and their faithful laboratories can detect?
Of course, the "maternal toxic load" is another extremely broad (one might even say deliberately vague) category. We'd need to get a ruling from the chelationista high court on which "toxins" they have in mind. And we'd better get that in writing, in case they try to "clarify their position" later. I suspect that this is another "trapdoor" they've left in their new "hypothesis" in order to allow them to escape again.
Simply defining the problem is impossible, given the vague and fuzzy nature of the new "causes of autism". And even if we were to get the problem defined - and get it to stay defined - there is another factor to consider.
If we were to assume (naively) that there were only six variables to examine, how many combinations would that make? If we have n items ("n" is a variable) and we group k ("k" is another variable) of them at a time, the number of possible combinations (where order doesn't matter) is: [see here for a tutorial]
n! / (n - k)! k!(the "!" stands for "factorial")
But, we have to count all possible combinations, including groups from one to six of the possible causes. This gives us a grand total of sixty-three (63) possible combinations.
And that's what we get by considering "heavy metals" and "maternal toxins" as one category. What happens if we expand "heavy metals" to the "big five" of toxic metals (mercury, lead, arsenic, cadmium, antimony)?
Total combinations = 1023And if we expand "maternal toxins" to include the five toxic metals?
Total combinations = 16,383Mind you, we haven't even begun to exhaust all of the potential "maternal toxins", "live viruses" or "bacteria" - the potential number of those could be in the thousands, if not the millions.
So now you see the shrewd cunning of the chelationistas. By making their "position" more vague, they have ensured that they will never again find themselves in the bind they are now. Never again will science be able to say - with any certainty - that the chelationistas' "hypothesis" about the cause(s) of autism is unfounded.
That's because it is practically untestable. It is not absolutely untestable - if we were to commit the entire scientific resources of the world to searching various combinations, it would probably take slightly less than an infinite amount of time to check them all. But it will never happen.
And that's exactly what the chelationistas want.