RE: Peering at review

[Jeff Hecht <jeff@jeffhecht.com>]

At 11:02 PM +0000 10/22/02, Matthew Bonnan wrote:

> However, as scientists, we should use our imaginations to devise 
> testable hypotheses inspired by the data we have at hand.  Using 
> your example of ceratopsian horn length, how might we figure out how 
> to determine within reasonable limits the "true" length of 
> ceratopsian horns?  Does mammalian horn growth tell us anything 
> about reptilian patterns?  What sort of test or experiment could we 
> devise to figure this out?  Is there a consistent ratio?  A 
> morphological signal or character we've overlooked?  Are dinosaurs 
> doing something different with horn growth, and, if so, how would we 
> look for that?

Testable hypotheses are important in evaluating theories, but 
sometimes there are two steps involved. First comes the theory -- 
e.g., Jack Horner's proposal at SVP that ceratopsian skulls were 
sheathed in keratin. It's based on some evidence, but doesn't make 
specific predictions that can be tested. The testable hypothesis to 
assess that theory may come later from a different person. Especially 
with fossils, some theories are more testable than others, and it may 
take a while to figure out the tests. (Short of luckily stumbling 
upon the right fossil of, say, a feathered dinosaur.)

-- Jeff Hecht