![]() | |
| Writings Discussion Authors Help Search Home | |
Simplicity By: ScottN on 8/23/2000; 6:13 PM I'm gonna bring a subject over from Nitcentral, because I haven't received a satisfactory response over there. To wit: Occam's Razor states that "Entities shall not be multiplied without reason", but is usually interpreted to mean that "the simplest solution is best". The question then becomes, "What is the simplest solution?" Or, more succinctly, "What is Simplicity?" Ah, there's the rub. I'll bring an example that I've heard of... Caveman Joe invents the classical square wheel. He doesn't like the ride -- too bumpy, surprise, surprise -- and thinks... He comes to the conclusion that there are too many corners on his wheel, and gets a brainstorm! He creates a "simpler" triangular wheel with only 3 corners! So, the quest for simplicity is not clear. From what I've heard, most people in the scientific community use the "I know it when I see it" approach. I'd like to take a crack at a better definition... Are you guys up for it?
RE: Simplicity By: Mark Morgan on 8/23/2000; 7:07 PM Scott sneers at my poor definition of simplicity, so I'll point to Occam's Razor over at the Skeptic's Dictionary as an introduction to the idea. And I still say, I know a simpler explanation when I see it. Mostly.
Untitled By: Brian Carnell on 8/28/2000; 8:32 PM The rambling Skeptic's Dictionary article was an interesting partial history of the parsimony principle (Occam's Razor) but it doesn't really get into the meat of the problem -- namely how do we falsify hypotheses and theories. Here's the lie I was taught in high school about the Earth not being the center of the universe: a long time ago people in the West believed that the Earth was the center of the universe. As astronomical observations increased, however, observations were made that falsified the ptolemaic model. This is at best a half truth. In reality what happened was that making the ptolemaic system working became exceedingly complex for the technology of the times, where movement of celestial bodies were printed in tables in bound books. In fact with today's computers you or I (or anyone with enough knowledge) could create a ptolemaic model of the universe that produces results accurate enough for pretty much everything we would need (the typical example is that you could easily launch and bring back the space shuttle based entirely on data derived from a ptolemaic model which is usually considered to be "wrong.") What the parsimony priniciple really hits at is what exactly science *is*. Does science accurately describe an objectively existing world or rather are all of our models "false" in a very deep sense (and if we are simply choosing between models why favor simplicity *aside* from the fact that it tends to minimize the computing power necessary to solve problems?) (BTW, it is odd for the SD entry to claim legal realism as an example of Occam's Razor since the legal realist position actually complicates legal matters by positing that most of the common sense ideas about legal systems and justice are irrelevant. The best parallel I could think of would be an astronomer or physicist saying that trying to explain the gravitational force is nothing but mysticism -- bodies do what bodies do.).
RE: Simplicity By: ScottN on 9/19/2000; 11:21 PM We all know about Occam's Razor, but the problem is how to tell which entities are multiplying... In Special Relativity, time and distance change with your relative velocity (the Lorentz Transform). In three dimensions, this can be worked out fairly easily (I was going to say simply, but...). However, if you jump to four dimensional space-time, it can be seen that the Lorentz transform is merely a rotation in four-dimensional space-time, and the object itself is invariant. Here we have multiplied one of our entities (dimensions), but with reason. I guess an alternative question is that, if entities shall not be multiplied without reason, what is an acceptable reason?
|
|
||||||||||||||||||||
| Copyright Notice | Privacy Policy | Contact | |
![]() |
|