Re: Sociological/Psychological Phenomenon

From: Douglas A. Gwyn (DAGwyn_at_null.net)
Date: 06/01/05 

Date: Wed, 01 Jun 2005 00:53:30 -0400

Pubkeybreaker wrote:
> What is it about math and crypto that makes an amateur
> believe that he/she can actually come up with a new theorem
> or algorithm that has somehow been miracuously missed by
> large numbers of previous (or current) mathematicians???
> I am sure that these same people would not post their ideas
> for (say) a new neuro-surgical technique. They do recognize
> that they have no competence.

Obviously there are genuine psychological factors
involved in any quackery or humbuggery, such as
lack of realistic self-assessment, need for peer
approval, etc., etc. Perhaps the reason one sees
more of it in math is that math in some form is
taught to nearly everyone, so some degree of
achievement really is possible, and the way the
subject is usually taught doesn't very well
indicate the much greater depths of technicality
that are mastered by math professionals; i.e. the
misguided soul may think that there isn't much
more to the subject than he has already learned,
thus further progress is a matter of who comes up
with the "aha" insight first.

To a lesser extent the same factors may be at work
among crypto novices: some degree of success is
readily achievable, and it can require very deep
knowledge to tell the difference between a worthy
cryptosystem and a worthless one. (The ciphertext
appears just as unintelligible either way.)

> We also see the same thing in sci.physics-- Amateurs
> (and cranks) who post their own (mostly mistaken) 'ideas'.

Possible another contributing factor in physics is
that general universal models of behavior are
called "theories", and it would appear that anyone
can have a theory. There is also great appeal,
shared by legitimate researchers, in the idea of
finding a universal explanation for "everything".

There actually are cases of cranks in the medical
and other professions, although they might not be
so apparent. Note also that much of what once
was believed by "the establishment" is now
generally believed to have been mistaken (or at
best, woefully incomplete), so by extrapolation
much of what is currently believed may well be
"quackery" that has not yet been corrected.

One of the problems with arguing against cranks
is that it often takes a lot of work to build a
complete, convincing logical demonstration of
their folly, which usually the crank cannot or
will not understand. Often the crank's response
to a demonstration that he does understand is
merely to tweak his theory slightly so that the
particular counterargument no longer applies.
A couple of rounds of this and the counterarguer
gives up, which makes the crank think that his
latest version has no counterargument.

Relevant Pages

  • Re: JSH: Oh, and another thing
    ... members of that community just dismiss all facts they don't like. ... You're arguing math, not ... in mathematics) even suspects you have anything close to a legitimate ... Virtually universal characteristics of cranks include: ...
    (sci.skeptic)
  • Re: Book on Pre-MATH for cryptography and cryptanalysis. Reply
    ... discrete math is quite important in cryptography. ... What's needed more is sort ... math and want to learn crypto theory? ... Actually, if you want to study crypto theory, there's another book that ...
    (sci.crypt)
  • Re: Wilsonian EM Control Frames.
    ... accuracy to detect anomalous precession using "brute force" ... ...and "using math that they can't understand" means anything beyond ... Yeah, we're both cranks, or so you believe. ... and I'm sure I speak for Sue as well. ...
    (sci.physics.relativity)
  • Re: Confused
    ... IMO, if math scares you, then you are probably wasting your time. ... I mean no disrespect, but you have no business trying to do crypto ... You can't learn cryptography without at least a minimal background in ...
    (sci.crypt)
  • Math for crypto?
    ... I'm an extreme crypto newbie beyond the security engineering level usage ... I'm currently doing Calc II in college.. ... Probability and Statistics I and II, and CS Discrete Math and the usual ...
    (sci.crypt)