[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Subject Index][Author Index]

Re: Some thoughts on cladistics

On Thu, Nov 01, 2001 at 09:52:42PM +0000, Ken Kinman scripsit:
>     Please, give us a break.  Phylogenetic systematics certainly has no 
> exclusive patent in using phylogeny as a standard (far from it), but it 
> seems overly simplistic to use phylogeny as the overriding standard to the 
> exclusion of other views of relationship.

*What* other views of relationships?

The folks in favour of phylogenetic taxonomy have a bunch of
descriptions of the kind of relationship they're talking about, argue
about how to measure it, and generally present it as something about
which science can be done.  (If I say that rostral bone length/width
ratio is a character in ceratopsians, and someone else does a lot of
stats and says, nope, it varies allometrically, my character has been
falsified; if I say that the following eighteen characters diagnose
oviraptorids and someone else can provide examples which cause that list
to be obviously insufficent versus the definition of oviraptorid,
they've falsified that, too.)

> It does not always make sense to classify an ancestor with all its
> descendants to the exclusion of that ancestor's close relatives
> (especially siblings).

Why not?

If the idea is to catalog present organisms, and to index them in some
useful way, we don't care how they got here.  If the idea is to figure
out how they got here, the shape of the tree of life, that someone _has_
descendants and that someone else _doesn't_ is significant to the shape
of the tree of life.

> Cladisto-eclectic centrist approachs to classification (Benton's, for 
> example) ARE nested and based on phylogeny----but they go beyond just 
> nesting and phylogeny, and reflect other aspects of relationship as well 
> (exemplified by the crude analogy of Charlemagne and his siblings).  And 
> thus I think such an approach (once polished) will be more successful, more 
> utilitarian, more informative, and more stable.

Your examples have a _long_ way to go.  They require the user to
remember a lot of arbitrary information -- many complex tagging symbols,
and a bunch of positional names.

The cladograms require me to remember that it's top down and left to
right in time.

Considered as a user interface, the cladograms are just way easier.

Stable is a _fault_, *if* the objective is to reflect the current state
of knowledge about the relationships of organisms.  If the objective is
to have a good index, stable is not a priori a fault, but is not a
priori an advantage, either; the effort of using the index will increase
over time without reorganization.

(Don't know much about dinosaurs; do know quite a bit about data
structures and interface design.)

               To maintain the end is to uphold the means.