[Vision2020] Ripping the Easter Bunny's Head Off : Was:Of, By &

[Paul Rumelhart]

One problem that a lot of people run into is the conflation of everyday 
terms with scientific ones.  I've run across the idea many times (and 
subscribed to it at one point) that a "theory gets enough evidence and 
becomes a law".   This is incorrect, at least as far as I understand it.

A "theory", I think, should be renamed to be a "framework" or a 
"model".  Theories are given worth more by how useful they are as 
opposed to how completely correct they are.  In other words, do they 
describe reality well enough to be useful in predicting things?  For 
example, the theory of special relativity replaces much of Newtonian 
physics, yet we don't throw Newtonian physics out because it is a highly 
useful approximation at velocities that are not a significant fraction 
of the speed of light.

What most people call a "theory" in everyday language is closer to being 
a "hypothesis" than a "theory".  A hypothesis can be falsified, and is 
pretty much defined by that fact.  A theory is judged upon its merits as 
a useful tool for making predictions - one anomalous result will not 
necessarily sink an entire theory.  Later, another theory may be devised 
that handles the discrepancies with the first one. 

A "law" is more of a description of some basic, universal principle that 
appears to be true based on multiple observations over many years.  The 
constancy of the speed of light and the law of inertia are two 
examples.  It's not a framework, it's more of a consistently observed fact.

Anyway, that's my limited understanding of the subject.

Paul



Dave wrote:
> This is why it is impossible to actually "prove" anything in science.
>
> Dave
>
>
> Joe Campbell wrote:
>   
>> For any one theory there 
>> are an infinite number of alternatives. Showing that all of them are false 
>> would be impossible. 
>>     
>
>