Darwin-L Message Log 3:65 (November 1993)

Academic Discussion on the History and Theory of the Historical Sciences

This is one message from the Archives of Darwin-L (1993–1997), a professional discussion group on the history and theory of the historical sciences.

Note: Additional publications on evolution and the historical sciences by the Darwin-L list owner are available on SSRN.

---

<3:65>From DARWIN@iris.uncg.edu  Mon Nov 15 19:57:56 1993

Date: Mon, 15 Nov 1993 21:04:58 -0400 (EDT)
From: DARWIN@iris.uncg.edu
Subject: Re: phenetics vs cladistics vs evol. class.
To: darwin-l@ukanaix.cc.ukans.edu
Organization: University of NC at Greensboro

Lynn Hanninen asks some large but important questions.  I don't know
whether I'll succeed in answering them briefly, but I'll give it a try.
Since they do require a bit of explanation I'll answer in installments over
a day or two, rather than all at once.

The first question is about the distinction between phenetics, cladistics,
and evolutionary classification.

The most important thing one has to do before dealing with this question is
to clearly distinguish between (1) classification and (2) phylogenetic
inference (reconstructing the evolutionary tree).  Failure to distinguish
between these two different activities has been a source of much confusion.
Phenetics and cladistics can be approaches to _either_ classification or
phylogeny reconstruction; "evolutionary classification" is (as the name
suggests) an approach to classification, not phylogeny reconstruction.

First let's talk about phylogenetic inference.

Suppose we have three species, A, B, and C.  They may be related in any one
of the following ways.  (It is important to understand that "relationship"
in this context means historical, genealogical relationship: relative
recency of common ancestry.  It is also important to understand that these
trees are "trees of history"; that is, the root represents an ancestor that
actually existed at sometime in the past.)

            /------- C             /------- C             /------- B
      /-----|                /-----|                /-----|
     /      \------- B      /      \------- A      /      \------- A
 ----|                  ----|                  ----|
     \                      \                      \
      \------------- A       \------------- B       \------------- C

Now since history only happened one way, the question is: Which of these
trees is the correct representation of the history of the species in
question?  A phenetic approach to the problem of phylogenetic inference
would say that the pair of species that are _most similar_ are the most
closely related.  Thus if B and C are more similar to one another than
either is to A, then B and C are more closely related to one another than
either is to A, and so the tree on the left represents the true phylogeny.
Advocates of this approach have proposed various ways of calculating
"overall similarity".

The cladistic approach to the problem is very different.  It decomposes the
phenetic notion of "overall similarity" into two parts: (1) _ancestral_ or
primitive similarity, and (2) _derived_ similarity.  The cladistic approach
to phylogenetic inference would say that the two species out of these three
that share the greatest number of _derived_ similarities are the most
closely related (ancestral similarities being irrelevant).  Thus if A and B
have more derived similarties in common with one another that either does
with C, then A and B are more closely related to one another than either is
to C.  The correct tree in that case would the one on the right.

Now what is the difference between ancestral and derived similarity?  That
is a topic I'll save for another installment, but I'll just say simply here
that a derived similarity (or derived character state) is an _innovation_,
and a ancestral similarity (or ancestral character state) is a _retention_.
("Ancestral character" and "derived character" are somewhat lax synonyms of
"ancestral character state" and "derived character state".)  More on the
ancestral/derived distinction later.

Let's now switch from phylogenetic inference to classification.  I should
state my own position clearly at this point: I regard the distinction
between phylogenetic inference and classification to be vitally important
because it allows systematists to ignore the whole subject of
classification, on which an enormous amount of ink has been wasted.
Consequently, I regard the discussion from this point on as moot: its value
lies only in helping us to understand some of the confused literature of
the past.  I am fully aware that there are people who don't share this
view, but I don't propose to debate them on this point.

Suppose we still have the same three species, A, B, and C, and we have
determined that this is the correct phylogeny:

                  /------- B
            /-----|
           /      \------- A
       ----|
           \
            \------------- C

The question facing someone interested in _classifying_ these three species
might run something like this.  (Notice that there is no dispute about the
phylogeny; that has already been established.)  Which of these is the best
classification for the three species?

    genus         genus         genus
     species A     species A     species B
     species B     species C     species C
    genus         genus         genus
     species C     species B     species A

(There are actually a great variety of other possibilities, especially if
we allow the rank to vary, including a genus for each species, one genus
for all three, separate families, etc., but this is a simplified example.)
Remember that all parties agree on what the phylogeny is, namely that A and
B are coordinate branches (sister clades).  A phenetic approach to
_classification_ would say that, since B and C are more similar to one
another than either is to A, then B and C should be grouped together into
the same genus, so the right-hand classification is best.  A cladistic
approach to _classification_ would say that all groups in the
classification must be whole branches (clades) in the tree, so the
left-hand classification is the best one, since it puts A and B in the same
genus.

The approach called "evolutionary classification", which has had Ernst Mayr
as its most vocal proponent, is something of a combination between the two.
It would say that the groups in the classification should be clades unless
there have been highly unequal rates of evolution in different branches, in
which case the branch that has undergone a lot of change should be singled
out and given higher rank (a family or order perhaps, rather than a genus).
Evolutionary classifications and phenetic classifications are often similar
as a result of this.  Consider the following as an example of a true
phylogeny (which it is) for three species:

                 /------- lungfish
           /-----|
          /      \------- cow
      ----|
          \
           \------------- goldfish

An advocate of phenetic classification might say that a goldfish and a
lungfish are more similar to one another than either is to the cow, and so
should be grouped together in the class Pisces.  An advocate of
evolutionary classification might come to more or less the same conclusion,
on the grounds that the cow's lineage has been highly modified since it
diverged from the lungfish's lineage, and so the cow has "left behind" the
two fish (as Pisces) and entered its own class, Mammalia.  An advocate of
cladistic classification, however, would say that while it is true that
overall the goldfish and the lungfish are more similar, the lungfish and
the cow are in fact a clade, and only clades should be recognized in a
classification, in this case perhaps as the group Choanata.

Cladistic _classification_ thus rejects the traditional taxon "Pisces", and
this example (among others such as the rejection of the traditional taxon
"Reptilia") caused many systematists to recoil from it.  The reason the
situation arises in the first place is that _classifications_ existed long
before evolution was accepted, and pre-evolutionary classifications were
by-and-large based on something like overall similarity; that is, they were
loosely phenetic.  Once the theory of descent was accepted it became
apparent that some of the pre-evolutionary _groups_ are in fact not
_clades_ (whole branches of the tree). If you read the discussion of
systematics in the _Origin of Species_ you can see this conflict becoming
apparent to Darwin as he attempts to reconcile established ideas of
classification with his new idea of "genealogical arrangement" (phylogeny).

But keep in mind that if phylogeny -- evolutionary history -- is what one
is interested in then one can disregard much of what has been said on the
subject of classification as a distinct enterprise.

Some general introductions to the issues discussed above are listed in the
file "biblio.clades" in the Darwin-L archives.  If you send the message GET
DARWIN-L BIBLIO.CLADES to listserv@ukanaix.cc.ukans.edu a copy will be sent
to you.  More on Lynn's other questions in a later post.

Bob O'Hara, Darwin-L list owner

Robert J. O'Hara (darwin@iris.uncg.edu)
Center for Critical Inquiry and Department of Biology
100 Foust Building, University of North Carolina at Greensboro
Greensboro, North Carolina 27412 U.S.A.

---

Your Amazon purchases help support this website. Thank you!