Oversimplifications in HtN -- Philosophy and biology

[Eric S. Raymond]

Richard Stallman <rms at gnu.org>:
> Judging theories against reality often involves testing predictions,
> but that is not the whole of it.  Deutschmann, in "The Fabric of
> Reality", shows near the beginning that what we want from a theory is
> not mere prediction of facts, but explanation of facts.

Of course he's correct.  Technically, Deutschmann talking about what 
epistemologists call "retrodiction" -- the use of a theory to construct
accounts of already observed (historical) phenomena that have generative
power, in that they suggest connections with demains of knowledge otber
than the one in which the retrodictive theory itself lives.

In terms of the criteria I wrote up, this is an aspect of consilience
("choose the partial theory which has the most in common with other
well-confirmed partial theories in adjacent domains").  It's a way
of choosing the best among competing, equivalently predictive theories.

> You seem to be arguing that we should try to understand certain
> actions of hackers without using the rest of what we understand about
> people generally

Hardly.  My understanding draws widely on neoclassical economics, legal
history, anthropology, sociology, primate ethology, and game theory.
Not by accident, either; only such a broad-based account of behavior
*can* be not only predictive but explanatory.

> By using our knowledge that hackers are people, and of the many kinds
> of motivations people can have, we can understand them a lot more.

That could be the epigraph for HtN.
-- 
		<a href="http://www.tuxedo.org/~esr">Eric S. Raymond</a>

Love your country, but never trust its government.
	-- Robert A. Heinlein.