a super(b) paper on human evolution

[Mike_Cahill at sil.org]

Perhaps we are closer together than it appears. I certainly don't want to
ditch theories wholesale! I probably should have written "Here are some
theories ABOUT X, but we know each one has some shortcomings." I don't know
of a theory (well, at least in my subfield of phonology) that covers
everything perfectly (e.g. no phonological theory explains partial nasal
place assimilation to labial-velar stops). But different ones are useful
for different types of investigations, and theories help you integrate
facts that would otherwise be random. I do NOT think one theory is as good
as another, but they are useful only as they have explanatory power. And we
shouldn't claim a theory is "true" when we know there are defects.

My undergrad degree was in biochemistry, so I've also had the opportunity
to see how theory applies in the physical sciences. There's similarities
and differences between evidence and support for a linguistic theory and a
theory of, say, atomic structures (which would be worth someone's
dissertation), but it is interesting that theories in the physical sciences
are always claims about reality, what is "really there." Theories in
linguistics are less likely to make such claims.  As for biology, well,
maybe we've had enough discussion on what's supposed to be a linguistics
discussion forum.

There is a place for butterfly collecting, but right, it doesn't take you
very far.

Mike Cahill
======================================================


Mike Cahill wrote (8/10/05)

> The assumption here is that one needs to have a theory to hold on to,
> otherwise you (metaphorically) drown.

Sorry.  You assumed that, not me.  And indeed if to you science is just a
spectator sport, you surely can indeed say

> "Here are some theories, but we know
> that none of them is adequate."

But if you're going to do anything in science beyond butterfly collecting,
you have to have a theory.  If you think you don't, that means you have an
implicit theory but don't know what it is.  Without a theoretical framework
to fit into, facts are meaningless.  So what you have to do is pick the
best
theory and try to improve on it.  Or come up with something that explains
all that the best theory explained and then some.

So,

> could we live with
> an agnostic position that says "Natural selection isn't adequate, and I
> don't know what the mechanism is."?

I don't think so.  Spectate, sure, but not do serious science.  You have to
have a hypothesis about the mechanism, period.  Niche construction theory
is
a case in point.  It doesn't say "NS isn't working, so junk it."  NS
obviously does work--up to a point.  So keep it but introduce additional
factors that might interact with it.  Not sink into a "Well, one theory is
as good as another" quagmire, still less lreap aboard something that just
junks NS and has nothing to put in its place.

Best to all,
Derek