Re: does hubey have a point?

H. M. Hubey (hubey@pegasus.montclair.edu)
10 Oct 1995 23:41:26 -0400

David Froehlich <eohippus@curly.cc.utexas.edu> writes:

>On 9 Oct 1995, H. M. Hubey wrote:

>Once again I can't believe I am answering thia guy. (in no way is my
>anserwing this to be considered a validation of Mr. Hubey's ability to do
>science (or at least this science))

And I will continue to answer yours because I'm a teacher and
I've had lots of dummies in my classes over the years :-)..

>> measure "distances" (morphologically) of skeletons from
>> other skeletons.

>fundamental assumption here that the transition was continuous and monotonic

No there isn't. It says nothing about the transition. It could have
been but it doesn't have to be. The description is only of
the metric and not evolution. The instrument should not be
confused with the thing being measured.

>> reasoning that is used is based on "similarity". And
>> the word "similar" really means that the "distance" between
>> the two is shorter than the distance between two that are
>> not similar.

>Once again, Mr Hubey demonstrates his ignorance of the methodology that
>we use. We are not looking for "similarity" or for "morphological
>distances".

Don't make a bigger spectacle than you deserve. Analogical
reasoning is used all the time. And we do look for distances.

As a systematist, we use shared derived features,
>evolutionary novelties to unite groups of organisms into clades.

Cladistics is an attempt to produce some reasoning principles
to prop up what seemed like arbitrary rules. You can think
of them as axioms or postulates. What you are doing
is still finding distances. Except now you have discrete
distance measures and not continuous ones. That's perfectly
reasonable.

Why
>does throwing numbers at a problem help at all? This does not mean that
>I disagree that quantification is not something to be strived for, rather
>I disagree that a number is necessarily better than a word.

We can't do without words. But mathematics is also a language.
And it has shown itself to be valuable in science. In this
case it will produce a precision (at least in the language.)
That is one of the problems with natural language. Its
strengths are also its weaknesses. Its imprecision also gives it
flexibility and allows us to eventually get the point across
somehow but in the last week or two much argument could have
easily been avoidedk if people were asked to simply produce
some numbers.

Mr. Hubey
>wants us to quantify bipedality-quadrupedality, while I think this is a
>gross oversimplification of the problem. Just within ungulates there are
>numerous ways of being quadrupedal, and Mr. Hubey wants us to lump all of
>that diversity into a single number.

WE can add them too. We can make more dimensions. This is one
of the reasons why fields such as this are more difficult than
things like physics. OTOH, you want to solve a more complicated
problem with a less powerful method. Where is the logic in that?

CAn you fix a car with a hammer?

Ultimately my concern is that
>numbers without understanding is no better than drivel (which I submit
>has not been in short supply on these posts).

You are right but drivel without numbers is probably worse.
Nobody likes drivel.

-- 

Regards, Mark
http://www.smns.montclair.edu/~hubey