From:    Wayne Throop                           
To:      All                                    Msg #77, Feb-14-93 05:00PM
Subject: Re: "Bad" design

From: throopw@sheol.UUCP (Wayne Throop)
Message-ID: <729746626@sheol.uucp>
Newsgroups: talk.origins

:: From: hamilton@hydra.gmr.com (William E. Hamilton)
:: Message-ID: <97583@rphroy.ph.gmr.com>
:: We humans have enough trouble designing "simple" things
:: like satellites.  So something that looks like bad design may serve a  purpose
:: we don't know about, and serve it admirably.

: From: tycchow@riesz.mit.edu (Timothy Y. Chow)
: Message-ID: <1993Feb10.014250.16932@galois.mit.edu>
: I proposed basically the same argument not long ago.  

This is also Bob Bales' basic argument from design, when confronted by
(the usual example) the Panda's thumb, whale legs, etc, etc.

To me, the problem isn't so much that the design is bad, per se.
But there *are* two problems that such examples pose for the argument
from design.

First, if you are saying that the goal of the design is unknown to you
as an explanation of designs that seem bad, then you are also saying
that designs you find "good" may really involve goals unknown to you
also, and what seems to you "good" design is not.  That is, saying that
the functional purpose is unknown is to undercut the whole argument from
design at its foundation.  It is no longer persuasive evidence of a
designer at all, since you have to throw out the baby (the "good"
designs) with the bathwater (the "bad" designs), in order to be
self-consistent.

But more significantly to me, it isn't the designs in isolation that
point away from a designer.  It is designs in the context of how similar
functional problems are solved by other organisms.  For example, why do
whales have leg bones, and sharks not? Why do octopus eyes have no blind
spot and human eyes do? Why do so many mammals use the fifth digit for a
thumb, but the panda uses a specialized bone for the purpose? Why do
chicken genes code for teeth that are never realized? These things pose
severe problems for a single designer hypothesis, and rescuing from
these problems by disclaiming knowledge of the functional role of the
feature in question doesn't really help at all, as described above. 

And pursuing further the question of the context of these "bad" designs,
consider the patterns they fall into.  Chicken teeth are reptile teeth
when expressed, never mammal teeth.  Vestigial bone structures in whales
are always mammal structures, never bird or reptile or fish or anything
else.  In particular, the heirarchical nature of feature distributions
among species is essentially *never* violated, even when it would
strongly seem to make functional sense to do so. 

Of course, some have observed that current "object oriented" designs of
computers decompose features into a heirarchical decomposition.  They
are overlooking that this is so (somewhat simplifying things) for units
of *implementation*, not of *function*.  And this distribution of
feature/implementation in living things is *still* just exactly the
opposite of what one would expect of a designed thing. 
--
Wayne Throop  ...!mcnc!dg-rtp!sheol!throopw