Ockham's Razor

#1
A term often cited. But most who cite it rarely mean what William of Ockham did - "Plurality must never be posited without necessity." It is commonly used more in the meaning that Bertrand Russell invented for it - "Whenever possible, substitute constructions out of known entities for inferences to unknown entities."
 

Paul C. Anagnostopoulos

Nap, interrupted.
Member
#2
A term often cited. But most who cite it rarely mean what William of Ockham did - "Plurality must never be posited without necessity." It is commonly used more in the meaning that Bertrand Russell invented for it - "Whenever possible, substitute constructions out of known entities for inferences to unknown entities."
Isn't inventing unknown entities making a plurality?

The important issue with Occam is that we need two competing hypotheses that both explain the observed data. Otherwise we are comparing apples and unicorns.

~~ Paul
 
#3
Isn't inventing unknown entities making a plurality?

The important issue with Occam is that we need two competing hypotheses that both explain the observed data. Otherwise we are comparing apples and unicorns.

~~ Paul
Yes. And here's another startling fact - all dogs are animals. I'll assume you get the gist of that.

You're statement is either wonky or poorly explained. Perhaps both? That is not the "important issue" and certainly not with Ockham's Razor. It may be something that has import to you. Even then the "apples to unicorns" is silly at best.
 
#4
I wish life was that simple as
Ockham's Razor claims it is .... lol

Try it on ...women, for example . lol

Simple can easily turn into simplistic or totally wrong sometimes , that is .

That razor has been useful though , especially in science...but then again, human science does deal only with the lowest level of reality through the lowest level of consciousness : the gross perceptual sensory one.

Let's hope mankind will be able to develop a higher form of science through higher levels of consciousness /awareness...
 
#5
Isn't inventing unknown entities making a plurality?

The important issue with Occam is that we need two competing hypotheses that both explain the observed data. Otherwise we are comparing apples and unicorns.

~~ Paul
I think I disagree with that. A hypothesis has explanatory power, or not, regardless of whether or not there is another hypothesis with similar explanatory power. And realistically, I don't see any particular reason to choose between them on the basis of entities.

However, what I mostly notice is that Occam's razor seems to come up whenever an idea lacks explanatory power in the first place. The focus on "excess entities" distracts from what it means to "explain the observed data". A made up entity which has as its effect "produces the observed data" does not have explanatory power.

Linda
 

Paul C. Anagnostopoulos

Nap, interrupted.
Member
#6
You're statement is either wonky or poorly explained. Perhaps both? That is not the "important issue" and certainly not with Ockham's Razor. It may be something that has import to you. Even then the "apples to unicorns" is silly at best.
So you think it's legitimate to worry about Occam when you don't even have two hypotheses that explain the data?

~~ Paul
 

Paul C. Anagnostopoulos

Nap, interrupted.
Member
#7
I think I disagree with that. A hypothesis has explanatory power, or not, regardless of whether or not there is another hypothesis with similar explanatory power. And realistically, I don't see any particular reason to choose between them on the basis of entities.
I'm not sure what the point of invoking Occam would be if there is only one hypothesis. I suppose the one hypothesis could include an entirely unnecessary component, but how would you know if there isn't anything to compare it to?

Given two or more hypotheses that both explain the data, why not tend toward rejecting the one with the extra entities? Of course it's logically possible that you could be wrong in doing so.

However, what I mostly notice is that Occam's razor seems to come up whenever an idea lacks explanatory power in the first place. The focus on "excess entities" distracts from what it means to "explain the observed data". A made up entity which has as its effect "produces the observed data" does not have explanatory power.
Agreed.

~~ Paul
 
#9
I'm not sure what the point of invoking Occam would be if there is only one hypothesis. I suppose the one hypothesis could include an entirely unnecessary component, but how would you know if there isn't anything to compare it to?
I was thinking that would be the point of experimentation - a more direct test of whether the components are necessary.

Given two or more hypotheses that both explain the data, why not tend toward rejecting the one with the extra entities?
Because I can't think of any reason why "the fewest number of entities" would be a necessary characteristic of useful ideas.

I agree that an idea which depends upon a series of ad hoc entities seems less satisfactory than an idea with one entity which accomplishes multiple effects. And I can be seduced by the elegance of an idea as much as anybody else.

Linda
 

Paul C. Anagnostopoulos

Nap, interrupted.
Member
#10
I was thinking that would be the point of experimentation - a more direct test of whether the components are necessary.
Yes, that would be a good thing. But if we can do that, then we don't need Occam.

Because I can't think of any reason why "the fewest number of entities" would be a necessary characteristic of useful ideas.
Until the multiple hypotheses can be sorted out through experimentation, it just seems reasonable to reject the one that has more baggage to explain the same data. Granted that rejection might turn out to be wrong.

Note that I'm not someone who often invokes Occam, so I'm not pushing it as a great tool.

~~ Paul
 
Top