Sunday, August 26, 2007

Criticizing Science

My recent article on IQ and the Wealth of Nations at produced some interesting commentary. I want to note one error that I think all of my harshest critics made, because I think it is a fairly common one. Basically, all of them said either or both of:
1) You did not positively disprove the authors' thesis; and/or
2) You did not offer a better, alternative theory of your own.

Folks, neither 1) nor 2) are necessary in a valid critique of a scientific theory. Someone putting forward a new scientific theory is in the position of the prosecution in a criminal case: They must present compelling evidence that the accussed (causative factor) is guilty. Their critics are like the defense: all we have to do is try to find holes in their case. In no way are we responsible for proving the accussed could not have done the crime (although, of course, if we can, that's great), nor suggesting who else might have done the crime. (Although Perry Mason was always able to both get his client off and practically convict someone else in the process.)

Note that this is almost the opposite of the Popperian paradigm, in which any theory is on equal footing with any other until it has been falsified. That suggestion, if taken seriously by scientists on the level of practice (which it never has been), would simply bring science to a grinding halt under the weight of the immensity of not-yet-falsified theories that could be devised.


  1. You are most likely familiar with the following:

    To establish that all crows are black, the corvologist observes as many crows as possible and triumphantly finds them without exception black.

    But "All crows are black" is logically equivalent to "All non-black things are non-crows," so the scientist may just as well observe as many non-black things as possible and will no doubt find them without exception to be non-crows.

  2. You used this principle to handle a tech support problem at MECA: The customer said Managing Your Money would not run on his Apple IIe. We at MECA couldn't locate an Apple IIe (perhaps relevant to why it wouldn't run?!), but you advised us that it was unnecessary: all we need to do was show the customer all non-MYM-running things and demonstrate that they are non-Apple-IIes.


Current review queue

Pearce: British Journal for the History of Philosophy Deneen: The American Conservative Chao-Reiss: Computing Reviews