Science: Do we believe or do we accept?

[dreadgeek]

Originally Posted by iamkeri1

[B][FONT=Comic Sans MS][SIZE=3][COLOR=magenta]My Thanksgiving when I was 10 years old was affected by the announcement just prior to the holiday that cranberries contained carginogens and should be removed from the holiday menu (we ate it anyway.) Today cranberries are touted as a healthy choice and a product that supports kidney/bladder function. Science flip flopped you might say. OR, as I believe, they jumped the gun before having all the evidence at hand. OR scientist A found one thing and scientist B found another. OR (as it is entirely possible), business wanted to increase cranberry sales, so they quashed the carcinogenic aspect of the fruit.

Let me suggest that there's another interpretation. Scientist A was wrong but did not realize that she was wrong and neither did anyone else. On better evidence, which was gained by scientist B, the error was discovered.

The late Steven Jay Gould, in a brief he helped write to the Supreme Court once stated that all scientific discoveries should come with the following codicil: "this is provisionally true, to the best of our knowledge, subject to revision upon better data". I would add to that that nothing is ever proven in science. I cannot prove to you, once and for all, that an atom of hydrogen has a single electron and a single proton. It can't be done. Even though earlier today I stated that it was diagnostic (I think I used the word definitional) of a hydrogen atom that it has a single proton and single electron, I still cannot prove it to you once and for all. I would fall down dead if we found a hydrogen atom that did not conform to that configuration and I think we could search the Universe for any length of time you care to mention and never find an exception but I still cannot prove it to you.

It is the black swan problem. There is NO observation you can ever make that would prove the statement "all swans are white". However, there is a *single* observation you can make to disconfirm (falsify) the statement "all swans are white". If I present you with a black swan then the whole white swan hypothesis falls apart. This is a subtle but nontrivial difference and one of the hardest things to grasp about how science actually works.

What looks like flip-flopping isn't actually flip-flopping, it's having better data upon which to make a conclusion that is less likely to be wrong.

Not knowing the details I'll hazard only the most tentative guess--chances are that there was enough separation between the first finding and the second that either technology or methodology enabled a more accurate conclusion. So when someone went back and tried to confirm the first study with better tools, they got a better result.

Cheers
Aj