Loading ...
Sorry, an error occurred while loading the content.

Re: [hackers-il] The "Science can't explain this" fallacy

Expand Messages
  • Arik Baratz
    ... Personally if it works and it s a placebo I don t need any of it, it s an indication that the problem can be taken care of by other means. ... Well, more
    Message 1 of 8 , Jul 3, 2006
    • 0 Attachment
      On 7/3/06, Nadav Har'El <nyh@...> wrote:
      > On Mon, Jul 03, 2006, Arik Baratz wrote about "Re: [hackers-il] The "Science can't explain this" fallacy":
      > > So? What's your point? Penicillin is better than placebo. So are the
      > > other cures you mentioned.
      >
      > Indeed. And "better than placebo" is the only thing that matters what it
      > comes to quacks selling you stuff. Why should I buy 100 shekel "medicine"
      > when a 1 shekel "medicine" will work just as well (both will be placebos)?

      Personally if it works and it's a placebo I don't need any of it, it's
      an indication that the problem can be taken care of by other means.

      > Who said nobody did this test? Maybe the manufacturer did it, and didn't
      > like the results. Maybe the manufacturer didn't want to do the test, because
      > he knew that he just invented this remedy (e.g., "let's take juice from some
      > fruits I have here on the table, and sell this as a cure to XYZ") and there
      > is no way in the world that the test will show it to work?

      Well, more likely than not the mfg didn't do the tests.

      And that's my point. It sells without doing the tests, BECAUSE
      scientists don't do these tests. Since scientists don't do these tests
      all the quacks can sell it without doing tests because people are used
      to the fact that alt. medicine doesn't go through tests and it is
      effective to some degree, so they just buy it without the tests.

      Perhaps if enough people use something someone will take it seriously
      enough to test it. Echinacea is widely used today, and I've seen some
      research about it, some of it claims that it's entirely ineffective.
      That's a good start, when you have contrary opinions.

      > And even if nobody tried to test it, it doesn't give this product temporary
      > validity. It just proves the seller is a quack, as all decent sellers (not
      > just in the field medicine) make sure that their product works before selling
      > it.

      Why would they test it? What incentive do they have to test it? People
      buy it anyway.

      > > Conspiracy theorists will say the large corporates don't want these
      > > things to be tested because they don't make money from unpatented
      > > mixtures of herbs.
      >
      > That's like saying that any mathematical theorem is true, it's just that
      > the large universities conspired to call only the math theorems they invent
      > "true".
      >
      > Most people don't care about the corporate's wishes. See, for example, free
      > software. People use it because it actually works - even if the large
      > corporates might wish that it never existed because they don't make money
      > from it.

      I don't like your analogy to software. Software either works or
      doesn't or is buggy to greater or lesser extent. You either can
      complete a task with some software or you can't. A math theorem can be
      verified by another mathematician. Medicine tests are subjective and
      you need a large body of statistical data to show a trend. If I give
      you the herb and you heal yourself with it, you can't claim anything
      except that, for you, it worked.

      > > Oh, do be skeptical, and work to disprove what is claimed, but do the
      > > work and don't just ignore it.
      >
      > The burden of proof is the other way around. The more outlandish your
      > results are, the harder *you* should work to convince others that you're
      > right. If your result is surprising *and* nobody can understand your
      > reasoning, don't be surprised if nobody believes you.

      You left the incentive to prove it outside this equasion.

      > The problem is that it's much easier to invent false theorems, than correct
      > ones. Scientists don't - and shouldn't - spend their time discrediting all
      > the possible false theorems, but rather finding correct ones.

      Well, if there's anecdotal evidence that doing some mental exercise 3
      times a day cures cancer, I think it's worth a glance. How do you find
      correct ones? Go back to the scientific process I described. The part
      of it where you have to be creative, where you have to think up
      something that has no basis so you can test it - you don't know it's
      correct until you verified it.

      > > I can't explain it rationally, except that I think scientists just shy
      > > away from these topics, believing them in the realm of religion or
      > > new-age or whatnot.
      >
      > I think you are confusing "scientists" with "scientific method". Ok, then
      > "scientists" - such as tenured professors and employees of billion dollar
      > drug companies - shy away from such research. Ok. But who cares? Whoever
      > does care about these subjects - such as the people who actually run such
      > workshops (for example) - can do the research themselves, using the scientific
      > method.

      But they don't need to because again they have no incentive.

      > Why can't they say "1,000 people graduated from my workshop in the last 10
      > years; Let's see if the percent of them that got sick is smaller than that
      > in the general population"? They can. And if they want to convince me that
      > there's any truth to their method, this is what they need to do.

      Nope. It has to be reproducible and has to be reproduced. Remember cold fusion?

      And regarding the workshop, that's just bad statistics, because there
      might be something in common to the people who go to the workshop;
      they chose to go there and they weren't selected in random, you did
      not isolate the random variable you're testing, no control group, no
      double-blind test. It proves nothing. It is, still, anecdotal
      evidence. Very good one at that.

      -- Arik
    Your message has been successfully submitted and would be delivered to recipients shortly.