The Grad Student and the Dropout

[Jeff Bone]

On Tuesday, Apr 29, 2003, at 11:36 US/Central, Russell Turpin wrote:

> I've seen this tension elsewhere in discussions
> of Ockham's razor. There are two fundamentally
> different interpretations of Ockham's principle.
> One says don't make your theory more complex than
> necessary. The other says don't multiply the
> assumed objects more than necessary. These often
> run in contrary direction, when axioms are
> introduced to get rid of spurious objects. The
> first variant can be justified in much the same
> way as hypothetico-deductive science: if a weaker
> theory gives rise to the same predictions, then
> the "excess" in the stronger theory is not
> justified by the evidence at hand. It is fluff.
> The second variation has always seemed rather
> naive to me, since it is essentially a supposition
> in favor of a smaller universe/meta-verse.

Seems that there's a class / instance distinction lurking in there.  In 
the "few objects" formulation, does that refer to distinct sorts of 
objects, or to instances?  If the latter, I agree that this is silly.  
If the former, that might actually be a *definition* of simpler theory 
--- fewer *kinds* of objects in a theory ~= simpler theory.

> I think there's a pop-sci book waiting to be
> written here. Do you think there's any money in
> it?

One day many years ago in an alternate universe called Austin, TX ("90 
square miles surrounded by reality") in a mythical time called The 
Eighties, a wise grad student and a disillusioned dropout are having 
tabooley and dolmas for lunch.  The unlucky dropout happened to be 
carrying a copy of _The Tao of Physics_, and got an hourlong earful of 
how shitty that book is and how bad pop-sci books are in general as 
learning tools.  The dropout never finished said book, in fact stopped 
reading pop-sci almost entirely from that point, having gained an 
appreciation (some might call it a serious jones) for going to source 
material directly and gleaning what he could from it.  Horses' mouths 
and all that.

Point being, I didn't think you liked pop-sci, Russell. ;-) :-)

Regardless, usually ideas don't go straight to pop-sci --- usually 
there's a corpus of "real" lit that gets built beforehand, before some 
science journalist picks up the meme or some scientist decides it'll 
help with grants if the get a little PR and pop-cred.  Can we actually 
short-circuit this process?  Why would you want to?

jb