On Ockham's Razor

Jan 06, 2009 20:24

Maybe I misunderstand the principle -- but this in part is because of its (mis)use -- but I am not a fan of Ockham's razor ( Read more... )

logical flaws, philosophy, science

Leave a comment

Comments 5

themadcow January 7 2009, 03:24:11 UTC
VERY poignant observation. I am annoyed by the same behaviors.

Your entry reminded me of an excerpt from C.S. Lewis (I think it was Mere Christianity):

It is no good asking for a simple religion. After all, real things are not simple. They look simple, but they are not. The table I am sitting at looks simple: but ask a scientist to tell you what it is really made of--all about the atoms and how the light waves rebound from them and hit my eye and what they do to the optic nerve and what it does to my brain--and, of course, you find what we call 'seeing a table' lands you in mysteries you can hardly get to the end of ( ... )

Reply


aefenglommung January 7 2009, 05:24:58 UTC
Aristotle also refers to this principle. It's called sozein phainomena, "to save the appearances ( ... )

Reply


pastorscott_o January 7 2009, 16:01:23 UTC
I think what you're frustrated with, is the unbalanced use of a general principle which was designed (ironically) to counter-balance the tendency of human nature to complicate many things which are best understood within the context of simplicity ( ... )

Reply

lhynard January 7 2009, 23:04:54 UTC
(Nice to see you have a user name. I just set it up so you can read more of my entries than just these public ones.)

For what it is worth, I wasn't so much claiming that the principle itself was stupid as saying that the principle as I understand it and as it is commonly used -- and I know that I must understand it incorrectly and that the common usage of it must be wrong -- seems stupid, so something must be missing.

Reply


jeltzz January 8 2009, 09:02:46 UTC
You are quite correct about it's common usage.

numquam ponenda est pluralitas sine necessitate
- never is a plurality to be asserted without necessity.

or else:

entia non sunt multiplicanda praeter necessitatem
- entities should not be multiplied beyond necessity.

Understood rightly, Ockham is laying forth a principle that theories should not have extraneous components. A more robust theory has less 'parts' and yet equally accounts for all complexities.

Reply


Leave a comment

Up