måndag 23 juli 2012

Feynman's Logical Fallacy

This clip shows that even Richard Feynman has fallen into the trap of the logical fallacy of confirming a hypothesis by observing consequences of the hypothesis. Or rather the other way around: Feynman in the role of the great scientist takes a firm grip of the audience by the trivial information that if a consequence of the hypothesis is at variance with observation, then something must be wrong with the hypothesis.  But by lifting this triviality to a deep insight by a great scientist, Feynman opens to the fallacy of confirming a hypothesis by observing consequences.

Feynman thrills the audience by revealing that a physicist starts out by simply guessing a law/hypothesis (which makes the audience laugh) and then seeks confirmation by observing consequences. Feynman does not say that a physicist starts by giving some direct rational reason why the hypothesis should hold. Simply guessing is the physicists method. No wonder that modern physics is so strange.

A longer clip is here. Feynman continues with a discussion about preferring hypotheses which are more likely, that is hypotheses with some rationality and not just wild guesses as he started out with. Feynman thus blurs a most essential aspect of science and causes confusing rather than enlightenment.

12 kommentarer:

  1. You are mistaken. Since there is no claim for absolute truth there is no logical fallacy.

    Physics is about observing and describing nature. It is also the practice of making predictions about phenomena not yet observed. As I remember it, this is thoroughly explained in Feynmans lecture (it's a couple of years since I saw the lecture).

    Pure physics doesn't deal with metaphysical bs.

  2. SuperNova,

    Yet modern science trades in Dark Energy, Dark Matter, Black Holes, and Evolution. Unprovable metaphysical BS.

  3. Well Anonym (Eller vad du nu heter), that is an opinion you have and you are certainly entitled to have one. Good luck with that.


    did we see the same clip (the longer one you are linking to)

    You wrote,

    Feynman opens to the fallacy of confirming a hypothesis by observing consequences

    Feynman is very clear that this is not how the scientific methods are used.

    What are your trying to accomplish? Public ridicule?

  4. No, Feynman starts out by saying that first a guess is made and then consequences are tested against observations. Implicit is that if consequences are not contradicted by observation then the guess may be turned into a law.

  5. Don't pull things out of it's context just to fit your claims, it's extremely unscientific.

    This contradicts your claim,

    Feynman says that you can always prove any definite theory wrong

    And directly after,

    Feynman says notice that we never prove it right

    Hence your claim of what is said just isn't true.

  6. The trouble with Feynman's approach is that the physical law is supposed to be valid as long as consequences are not contradicted by observation, thus without any requirement of direct justification of the very law itself. This has led to string theory with the basic laws both unknown and impossible to directly justify. Physics based on verifiable laws would be much healthier than physics based on wild guesses which cannot be directly tested.

  7. Your reasoning becomes stranger and stranger.

    If a law can be used to repeatedly predict the outcome of a certain experiment, then the physical law of course is valid for that application. It can of course never prove any general validity which Feynman clearly states. So what's the fuzz?

    If you look at the piece from the lecture once more you will also see that Feynman claims that there isn't any wild guessing going on. You seem to be strangely biased towards not seeing what is said both lucid and clear.

    So I have two questions for you.

    1. How on earth would you execute a verification of a physical law?

    2. In your opinion, what is the purpose of expressing a physical law in the first place?

  8. You miss the point: There is no reason to use "physical laws" which cannot be directly tested and thus verified, more or less. But modern physics is not reasonable in this sense, and commonly uses "laws" or assumptions which cannot be directly tested such as the assumption of "molecular chaos" of statistical mechanics or symmetry/anti-symmetry of the wave function of quantum mechanics. Relying on assumptions which cannot be tested may lead to empty science, and thus should be avoided.

  9. SuperNova,

    it is not my opinion. Any "proof' you present of the above items will rest heavily on interpretations of observations based on ASSUMPTIONS. I welcome your lesson in how misguided I am.

    Miles Mathis also takes issue with modern science for similar reasons. When there is only Math that has been adjusted for decades to match observations without underlying physical mechanisms you do not have science. You have engineering.

  10. Dear Anonym (or whatever your name is),

    do take the time to read and ponder

    Warren Seigels - Are you a quack?

    It could be educational for you.

    And further notice that Nova in my alias is not a reference to the astrophysical phenomenon. It is my name. I don't care particularly about astrophysics.

    The Super is just because I'm both awesome and hot.


    do take the time to read and ponder

    Warren Seigels - Are you a quack?

    This is not a joke. It connects strongly to your original post about Feynman and further says a lot about the scientific method. And it also got some good humor in it. That is win, win, win.

  11. It is not so clear who is the quack. In any case the discussion is over.

  12. You miss the point.

    I'm not saying that you are a quack.

    I first wrote down the reference to Seigels page as a respond to the person who refers fourth Miles Mathis.

    I then realized that the webpage has a lot to say about science and the scientific method and if you haven't read the page it should be a good read for you. Or are you hindered by pride?