Extraordinary claims
There appears to be this idea out there that "extraordinary claims require extraordinary proof [or evidence]" is a description of the scientific method and the usual burden of proof. That is, I've encountered people who seem to think that any claim that differs from the null hypothesis is the "extraordinary claim," and assign the burden of proof accordingly. This is annoying error in terminology, but nothing more.
Less benign are those that think that any claim that differs from the status quo is the "extraordinary claim," and that the burden of proof is upon supporters of that particular claim or hypothesis. It's often used in the sense of which of a competing pair (or more) of ideas is the one that requires proof. A good example are bike helmet debates; I've seen someone insist before that the claim that bike helmets increase the risk of head injury is "the extraordinary claim," and requires evidence, while the claim that bike helmets reduce the risk of head injury is obvious and does not require special support - i.e., it is self-evident that helmets give protection.
Of course, that's not true. The null hypothesis - that is, our a priori assumption about effect of bike helmets on head injury risk - is that they do nothing. Claims that they affect the risk of head injury in a way that differs from that of random chance require evidentiary support.*
"Extraordinary claims require extraordinary proof," is a phrase that was coined by Marcello Truzzi, a sociologist who was an outspoken critic and tester of pseudoscientific claims. Well, actually, what he said was this:
And when such claims are extraordinary, that is, revolutionary in their implications for established scientific generalizations already accumulated and verified, we must demand extraordinary proof.
This statement is usually paraphrased as above, and popularized by Carl Sagan as "Extraordinary claims require extraordinary evidence." It's nothing more than a warning that claims that contradict or otherwise have serious implications about the veracity of evidence already accumulated require close examination. Someone claiming that cancer can be cured with meditation and a raw foods diet (yes, there are people who say that this can be done) has a lot of things to demonstrate before they've made a convincing case!
The final word is that ALL claims require evidence. Calling a claim "extraordinary" is not a matter of methodological terminology, but a statement about its place in the contemporary library of accumulated knowledge.
* As for bike helmets, if you want to know the end of that story: there is very little evidence that helmets increase head injury risk (motorists may pass slightly closer, that's about it) or worsen injury prognosis. There is also no solid evidence that they reduce the incidence of head injury, but good evidence that there is a minor correlation between helmet use and patient prognosis. That is, you're no less likely to be injured, but you'll probably do better if you're injured while wearing a helmet. That said, the risk of injury is pretty small anyway; not wearing one is perfectly safe. Personally, I make a kind of Pascal's Wager for head injury, and wear it for road riding and usually for commuting, but I also tool around without one without any concern.
Less benign are those that think that any claim that differs from the status quo is the "extraordinary claim," and that the burden of proof is upon supporters of that particular claim or hypothesis. It's often used in the sense of which of a competing pair (or more) of ideas is the one that requires proof. A good example are bike helmet debates; I've seen someone insist before that the claim that bike helmets increase the risk of head injury is "the extraordinary claim," and requires evidence, while the claim that bike helmets reduce the risk of head injury is obvious and does not require special support - i.e., it is self-evident that helmets give protection.
Of course, that's not true. The null hypothesis - that is, our a priori assumption about effect of bike helmets on head injury risk - is that they do nothing. Claims that they affect the risk of head injury in a way that differs from that of random chance require evidentiary support.*
"Extraordinary claims require extraordinary proof," is a phrase that was coined by Marcello Truzzi, a sociologist who was an outspoken critic and tester of pseudoscientific claims. Well, actually, what he said was this:
And when such claims are extraordinary, that is, revolutionary in their implications for established scientific generalizations already accumulated and verified, we must demand extraordinary proof.
This statement is usually paraphrased as above, and popularized by Carl Sagan as "Extraordinary claims require extraordinary evidence." It's nothing more than a warning that claims that contradict or otherwise have serious implications about the veracity of evidence already accumulated require close examination. Someone claiming that cancer can be cured with meditation and a raw foods diet (yes, there are people who say that this can be done) has a lot of things to demonstrate before they've made a convincing case!
The final word is that ALL claims require evidence. Calling a claim "extraordinary" is not a matter of methodological terminology, but a statement about its place in the contemporary library of accumulated knowledge.
* As for bike helmets, if you want to know the end of that story: there is very little evidence that helmets increase head injury risk (motorists may pass slightly closer, that's about it) or worsen injury prognosis. There is also no solid evidence that they reduce the incidence of head injury, but good evidence that there is a minor correlation between helmet use and patient prognosis. That is, you're no less likely to be injured, but you'll probably do better if you're injured while wearing a helmet. That said, the risk of injury is pretty small anyway; not wearing one is perfectly safe. Personally, I make a kind of Pascal's Wager for head injury, and wear it for road riding and usually for commuting, but I also tool around without one without any concern.