Re: Occam's Razor

On 25 Jan 2013, at 03:13, hibbsa <[email protected]> wrote:

> 
> Would the following arrangement violate Occam's Razor:
> 
> - A single methodology that initializes with the assumption the 'answer'
> will be complex not simple (maybe assumes some sort of maximum or
> infinite-in-the-limit complexity), but progressively approximates to the
> simplest possible explanation?
> 
> The way I see it, this setup is legitimate in Occam terms, if the
> legitimacy of an arbitrary occam-correct explanation is unaffected by
> the origin of that explanation. For example, it could come by a random
> guess, or a creative process....or indeed some sort of 'guiding' method.
> 
> Opinions appreciated.

Occam's razor is too vague for there to be a well-defined answer to that question of whether that variant of it violates it or not.

But I think the variant has the same flaws as the original. One of them is that the 'simplicity' or 'complexity' of an explanation itself depends on how one judges explanations (i.e. aside from by their 'simplicity'). For instance, a devoutly religious person might claim that their world view conforms to Occam's razor because "God did it" is a very simple explanation of the origin of things like the distinction between right and wrong, or the laws of physics. An atheist would take the view that the 'simplicity' of the God hypothesis is purely formal and that the hypothesis by itself isn't an explanation of those things at all. So in the atheist's view, the religious person is using complex, unacknowledged explanations and just *calling* them a simple name, and so is violating Occam's razor. 

Similarly, the religious person might also claim that their world view conforms to the variant methodology, since although God is infinitely simple, 'The Answer' to the mystery of His nature becomes ever more complex as one understands it more, but is at each stage the simplest explanation. And then the atheist might go on to say that the religious person's increasingly complex ideas about 'The Answer' are merely more and more complicated ways of evading the flaws in the original hypothesis, and therefore are neither truly complex nor truly explanations. And therefore that the religious person isn't conforming to the variant methodology either. 

This disagreement between the religious person and the atheist about what does or doesn't violate Occam's razor is not being conducted according to the rule of simplicity, nor according to the variant methodology, but by different criteria. And in fact there is no substance to their dispute about what explanation to adopt, other than those additional criteria. Simplicity isn't playing a role at all.

-- David Deutsch