Ancestors and descendants

[NJPharris@aol.com]

Ken Kinman asked, a while ago, whether one were more closely related to one's 
parents and siblings, or to one's descendants a thousand generations removed.

That is an interesting question, and I feel it is best resolved by turning it 
around: are my descendants a thousand generations removed more closely 
related to me, or to my parents and siblings?  The answer, as I think is 
fairly obvious, is that they are more closely related to me, and that at some 
level, those descendants possess--or their ancestors possessed--certain 
features that link them to me, to the exclusion of the rest of my family.  
Thus, if I were creating a classification from my family tree, I would be 
justified in erecting a group containing myself and all of my descendants, to 
the exclusion of my parents and siblings.

Likewise, I might examine a collection of living and fossil organisms, decide 
they forma a natural group, and call them "birds".  I might then happen upon 
some specimens of _Deinonychus_ and _Ceratosaurus_.  I would fairly quickly 
notice that _Deinonychus_ and "birds" share a number of features not present 
in _Ceratosaurus_, and I would therefore erect a group that included 
_Deinonychus_ and birds, but not _Ceratosaurus_.  But _Ceratosaurus_ shares 
many features with the first two taxa that are not found in _Plateosaurus_, 
so I group the first three together to the exclusion of the fourth.  And so 
on, and so on, until (ideally) I have classified every individual organism 
accessible to scientific study.

My question to you, Dr. Kinman, is this:  what is it, in your eyes, that 
would make this "bottom-up" classificatory system less useful than one which 
groups _Deinonychus_, _Ceratosaurus_, and _Plateosaurus_ together to the 
exclusion of the birds?

Nick Pharris