Philosophical views of phyl. reconst.

Joe Felsenstein joe at evolution.genetics.washington.edu
Mon May 22 18:55:59 EST 1995

In article <wfl1-140595101149 at>,
Warren Frank Lamboy <wfl1 at cornell.edu> wrote:
>Can somebody please define or point to a reference that defines the three
>philosophical views underlying phylogenetic inference that were recently
>characterized by Joe Felsenstein as:
>         hypothetical-deductive
>         logical parsimony
>         statistical?

Ron DeBry has commented on the last. The first (really "hypothetico-deductive")
is a view which sees inferring phylogenies as an example of Popper's scheme
of scientific inference.  Individual characters are seen as refuting each
other if they conflict (if they cannot both be unique and unreversed on the
same tree).  Alternatively, and equivalently, the characters can be seen as
refuting trees.  For expressions of this view, which was popular among
phylogenetic systematists in the early 1980's, see, for example, Eldredge and
Cracraft's ("Phylogenetic Patterns and the Evolutionary Process", 1980)
   "If one views the science of systematics as being subject to the same
    rules of inference as other branches of hypothetico-deductive science,
    then these assumptions and expectations take on the nature of an axiomatic
    methodological principle: because we cannot empirically have knowledge of
    the true historical pattern, science must formulate a criterion by which
    to judge the relative merits of our close approximations (hypotheses).
    That criterion, in effect, is parsimony, and it specifies the the most
    preferred hypothesis to be the one exhibiting the most congruence in the
    synapomorphy pattern." (p. 67)
   "In the scientific literature, one will also find the word 'falisfied' to
    refer to hypotheses such as those represented in cladograms b and c.
    'Falsified' implies that the hypotheses are proven false, but this is not
    the meaning we (or other phylogenetic systematists) wish to convey.  It
    may be that, eventually, the preferred hypothesis will itself be 'rejected'
    by some synapomorphies.  Another word that is occasionally used is
    'refute', but this has essentially the same inplications as 'falsify'.
    (page 69, footnote)
   "the criterion of parsimony specifies our acceptance of the least
    rejected hypothesis."  (p. 70)
Ed Wiley ("Phylogenetics", 1981) declares on page 20 that "I shall adopt the
    hypotheticodeductive approach throughout this book." and that "the
    principle of simplicity (parsimony
J. S. Farris (Advances in Cladistic, vol. 2, 1983, p. 8 says that "Wiley
    [in a 1975 paper] discusses parsimony in a Popperian context,
    characterizing most parsimonious genealogies as those that are least
    falsified on available evidence.  As I shall discuss below,, any such
    falsifier engenders a requirement for an ad hoc hypothesis of homoplasy
    to defend the genealogy.  Wiley's concept is then equivalent to mine."
  and he also explicitly disavows the statistical inference approach (p. 17)
    "The statistical approach to phylogenetic inference was wrong from the
    start, for it rests on the idea that to study phylogeny at all, one must
    first know in great detail how evolution has proceeded."
These sources will be helpful for much other explanation of this postion.

As for the second approach, it is documented very little in the literature,
but at he end of 1989 Arnold Kluge gave a talk at a Society for Systematic
Zoology symposium, and I asked him in the question period specifically about
this.  He said that the hypothetico-deductive approach to phylogenies was
"dead and buried, years ago", agreed with the characterization of his view as
that the reason for using parsimony was not because of its correspondence to
hypothetico-deductive criteria but because of its correspondence to
Ockham's criterion of parsimony.  This is known as the "logical-parsimony"
approach.  When asked where it could be found in the literature he cited
Elliott Sober's book "Reconstructing the Past".  However that does not
put forward this view, but instead a statistical one.  Likewise some people
have claimed that the recent paper by Kluge and Wolf in Cladistics lays out
the foundations of such a view, but it does not that I can see.  Kluge gave
another talk in 1994 at the Evolution meetings in Georgia that mentioned the
approach, calling it instead "logical probability" which has an ambiguous
ring to it (to my ears, anyway).  It would seem important, in a field where so
much emphasis is laid on being explicit about one's assumptions, to have
somewhere that this view is laid out.  If anyone can point out where it is,
they should let us know.  I believe that this view is not unique to Kluge but
is also shared by Farris and others.

Sorry for the delay -- I was waiting to see if others would step forward
and join the discussion.

Joe Felsenstein         joe at genetics.washington.edu     (IP No.
 Dept. of Genetics, Univ. of Washington, Box 357360, Seattle, WA 98195-7360

More information about the Mol-evol mailing list

Send comments to us at biosci-help [At] net.bio.net