John Grehan jgrehan at SCIENCEBUFF.ORG
Wed Jun 16 11:12:52 CDT 2004

I was probably a bit out of line in posting my statement that
definitions don't matter without bothering to give some context.

My assertion that definitions don't matter is that definitions do not
generate solutions to scientific problems, rather definitions represent
the 'solution' as the end product of a research activity. Definitions
therefore do not tell one anything more than one already understands.
For example, if I were to say that a panbiogeographic node is a "center
of incongruence" this would be meaningless to anyone practicing
biogeography outside a knowledge and understanding of panbiogeography,
no matter how many dictionaries were consulted. Definitions come with a
background knowledge, and it is this context that may be more important
than the authority of any single meaning imposed upon a term. 

In an ironic way it seems that definitions can even create or foster
confusion (I recall the past and ongoing debates about defining species
for example) or even obscure the fact that one one really knows what
they are talking about in a subject (the Gothean definition of the
carpal might be a good example).

> From: Richard Jensen [mailto:rjensen at]
> If you're statement were true, then we would have no need for
> dictionaries.

I think my views above indicate otherwise.

> It is clear from the postings you have made on this you have not been
> using
> unambiguous definitions of characters, phenetics, and cladistics, 

Possibly, although I doubt there are any unambiguous definitions of

> you
> have persisted in implying that characters may be innately phenetic or
> cladistic, despite the best efforts of others to point out that this
> wrong-headed thinking.  

Or my effort to suggest that this thinking is not necessarily

According to your reply, it does no good for
> anyone
> to provide a definition because you already know the definition.  

Not correct

But (and
> I suspect most readers of this thread will agree with me) "what you
> know" does not conform with what the rest of us know (by reference to
> accepted definitions).

Possibly - although definitions may not be universally 'accepted'.

> I assume (and you will correct me if I'm wrong) that your comment
> certain practices not conforming refers to analyses of molecular data.
> someone compiles a matrix of putative molecular homologies, and
applies a
> maximum parsimony algorithm, then that is not, no matter what some
> might tell you, a phenetic analysis.  It simply does not conform to
> phenetics is.

It does if the characters represent overall similarity - that the
homologies used to select the characters for analysis are not limited to

John Grehan

More information about the Taxacom mailing list