Re: Induction vs Rubbish

From: Patrick Leahy <jpl.domain.name.hidden>
Date: Wed, 25 May 2005 17:53:43 +0100 (BST)

On Wed, 25 May 2005, Benjamin Udell wrote:

>
> The induction-friendly universe with so much detectable rubbish that a
> wide variety of phenomena cannot be unified into a simple theory sounds
> like a universe where induction works but surmise, or inference to the
> simplest explanation, faces grave difficulties and too often fails. In
> other words, in difficult cases, efforts toward surmise -- i.e.,
> "rambling speculations about half-formed ideas that probably won't pan
> out to anything" -- really will lead too often too far astray to be
> practicable, and cogent everyday surmises would be few and far between
> -- not everyday or quotidian at all. A greatly increased difficulty in
> the formation of explanatory hypotheses would, it seems, hamper not only
> science but SASs in general. Would intelligence and commonsense
> perception tend, on balance, to be useful in such a world? It sounds
> like a world which would allow vegetable-like systems (i.e., essentially
> mindless in the usual sense) but be severely punitive toward SASs
> inclined to try to be shrewd or clever and to try, for instance, to
> infer particular entities or events or universal laws (as opposed to
> prolonged tendencies) as explanatory reasons, or to try to play
> architect instead of subsisting on the continuation of tendencies. It
> also sounds like the evolution or "natural architecting" of even merely
> vegetable-like systems would likely be under pressure to play it a lot
> safer than it does in our world, so that the systems thus evolved would
> tend to be not only vegetable-like but also a lot more "generic" than
> those which we see. I guess I'm trying to argue (unconfidently) or
> suggest, for what it's worth, that induction-friendly but
> much-detectable rubbish universes with SASs are induction-friendly but
> surmise-unfriendly universes with SASs, and that their measure would be
> rather small.
>
> Best regards,
> Ben Udell

It's a question of degree, again. There is surely a level of "noise" which
doesn't cause the problems you mention (although I would say that surmise,
common sense etc are basically inductive reasoning from past experience,
including past experience genetically encoded by natural selection which
is one big inductive experiment). For most of history, the world has
seemed a pretty random place to people (probably still does to most
people), but they managed to survive without understanding how QM unifies
the structure of matter, Natural selection explains so much about living
things, etc. If the rubbish was there, we'd get used to it. Only
scientists would be frustrated that they couldn't make any kind of sense
of it. But they would be able to isolate the features of their world which
did show regularity, so it wouldn't prevent science, either.

Paddy
Received on Wed May 25 2005 - 13:00:58 PDT