cladism's greatest weakness

Byron J. Adams bjadams at UCDAVIS.EDU
Wed Sep 22 10:01:26 CDT 1999

Richard Jensen responded to my explanation of why we don't often recognize
groups based on the asbsence of evidence (privations):

>> I believe there is a logical reason that we do not find groups such as the
>> one described above (sister taxa; one that has evolved an apomorphy, one
>> that has not).  To do so is to recognize a privative group, or a group, the
>> essence of which is absence.  Aristotle noticed the ambiguity of
>> recognizing privative groups; they cannot be subdivided, and as applied to
>> species or lineages, neither can be logically characterized by the absence
>> of nothing.  Put bluntly, the anagenetic (static, or unchanged) taxon
>> cannot produce evidence that it is evolving independently from its sister
>> lineage.  Thus, there is no reason for us to suspect that the two taxa
>> represent distinct species (or lineages).
>I beg to disagree.  We have two lineages, X (the original) and Y (the
>peripheral isolate).  X retains all the properties that allow us to
>recognize it as X.  Y has some new property that makes it reproductively
>isolated from X.  Given that this reproductive isolation exists, we have
>every reason to suspect that X and Y are evolving separately.  The fact
>that X is unchanged does not prevent our recognition of X as a distinct
>entity different from Y; after all, Y has a unique character (call it
>anything you like (I'll call it A') that allows
>us to recognize it as different from X. This would be a case where the
>"absence of something" allows us to recognize that one taxon (X) is
>different from another taxon (Y). My key to the two taxa would simply have
>a couplet
>       A' present..........Y
>       A' absent ..........X

I do not quibble with the fact that the two entities in your example, X and
Y, cannot be diagnosed.  X is "different" than Y, and your key provides a
proof of that.  However, I think you do not make the distinction between
"diagnosing" entities, and delimiting lineages, or species.  I hope to
convince you that these are two different enterprises.  In your example,
taxa X and Y are diagnosed by their unique combinations of characters.
Unique combinations of characters may be diagnostic, but because they are
not apomorphies, they need not nest hierarchically (i.e. phenetic vs
cladistic).  Thus, extending the diagnosis of entities to the delimitation
of lineages or species can produce taxa whose taxonomic relationships are
discordant with their evolutionary, or historical relationships.  Simply
imagine a third taxon, Z, that is also delimited on the basis of negative
evidence.  The difference between diagnosis and delimitation, then, is that
the former is ahistorical, whereas the latter has meaning within an
evolutionary, or historical context.  Delimiting lineages and species based
on privations (or unique combinations of characters as opposed to
apomorphies) can divide organisms into overlapping and incompatible
lineages or species (Hull, D. L.  1997.  in "Species: The Units of
Biodiversity" Claridge et al., eds).

Concerning the logical fallacy of recognizing privative groups, Richard
Jensen wrote:

>Well, given that we now view all H. sapiens on Earth as a single taxon,
>aren't we already (as a species) simultaneously both ancestor and
>descendant?  Living members of H. sapiens are all derived from now extinct
>members of H. sapiens, but they and we all belonged to H. sapiens.

This is a clever way of arguing against the notion that there is no logical
reason to reject a "mother taxon" giving rise to a co-extant "daughter
taxon."  I suppose the argument would be valid if we rejected the notion of
lineages and species as individuals that persist as spatial and temporal
extensions through time (i.e. the "space-time worms" of Baum, Syst. Biol.
47:641-653).  But H. sapiens is a single, individual lineage, one that
originated once, and one that we predict has a unique fate.  So while the
relationships between early H. sapiens and modern H. sapiens is ancestor
decendent in a tokogenetic fashion, the lineage maintains its singular
identity -- the lineage has not given rise to successive lineages.  If
lineages or species are time extended individuals, then it is logically
absurd for them to be an ancestor and a descendent at the same time.

Richard Zander wrote:
>Interpretation of lineages as surviving ancestors is possible. If a terminal
>taxon has no autapomorphies (no branch length) it still exists, and is
>privative only in the absence of traits interpretable as advanced and not
>also in its sister group/taxon.

The scenario you describe is depicted in Richard Jensen's hypothetical
example above.  Extant lineages may be interpreted as surviving ancestors,
but I find the practice incompatible with viewing lineages and species as
time-extended individuals (arguments above).

Byron J. Adams
Department of Nematology
One Shields Avenue
University of California
Davis, CA 95616-8668
Phone: (530) 752-1404
fax: (530) 752-5809

More information about the Taxacom mailing list