Nameless Taxonomy and Senseless Babble

Richard Pyle deepreef at BISHOPMUSEUM.ORG
Sat Feb 1 13:19:08 CST 2003

I tried to stay out of this discussion.  Honestly -- I really did.  I wanted
to let it go.  But I just couldn't.  So here is my senseless babble about
nameless taxonomy:

First, let's not forget the ultimate goal:  to understand the natural world
around us.  The word "understand" in this context applies to both "facts"
(individual instances of observation and documentation -- otherwise known as
knowledge), and "patterns" (models and algorithms derived from the
assimilated collection of facts, that allow us to more accurately predict
what we have yet to, or in some cases cannot, observe directly -- otherwise
known as hypotheses).

Taxonomy, Systematics, Nomenclature, etc. are tools we use to work towards
this ultimate goal -- accumulating facts and deriving models and algorithms
about living organisms.  The "facts" are the individual specimens, their
morphological characteristics, the sequences of DNA within their cells, and
relevant observations about place, time, and interaction with the
environment and other living organisms. The "patterns" include both the
abstract (clusters of individual organisms about which we discover or assign
affinity to [taxa]; the predictions [postdictions?] about the historical
origin of the shared ancestry of those organisms), and the more tangible
(predicting distributions; changes to diversity over time; etc.).

In the present context, we need to keep in mind the distinction between the
*mechanism* of information organization (nomenclature), and the information
itself (organismal facts and taxonomic patterns). Nobody (I don't think)
would argue that the further accumulation of facts is a bad thing.
Documenting both morphological characteristics and DNA sequences can only
assist us in our goal of understanding the natural world.  The arguments
come in the form of how to effectively manage and organize those facts in
ways that lead us to the derivation of patterns with the most consistently
reliable and accurate predictions. Linnaeus established a *mechanism* to
organize and manage information about biodiversity via a consistent
hierarchical nomenclatural system.  This mechanism has evolved only slightly
over the years, and thus has the advantage of providing an infrastructure
for the organization of vast quantities of facts accumulated over the past
two and a half centuries. The problem is that this mechanism was established
without the foresight of evolutionary theory or the existence of DNA.

I think that future history will show the latter part of the 20th century
(and early 21st century) to be an inflection point in biodiversity
information management.  We now find ourselves struggling to merge the
knowledge of the past (written in Latin nomenclature), with the knowledge of
the present (written in DNA sequences and cladograms). In an effort to
bridge this gap, Linnaean nomenclature has been stretched, pulled, prodded,
and otherwise contorted to represent the burgeoning new pool of facts and
patterns derived there from.  The development of PhyloCode, and more
recently Hebert's bar-coding, are signs that we're approaching the breaking
point of the Linnaean system.  When you reflect on the past, and consider
the present, I think the answer to this dilemma can be in part obtained by
looking toward the future.

Reflecting on the history of information management technology in general
(i.e., computers), and biodiversity fact-generation in particular (i.e., DNA
sequencing), most rational people would agree that the ability to extract
VAST quantities of DNA sequence data at very little cost in terms of both
money and time is going to continue to accelerate during the next several
decades at least (unless, of course, we encounter a catastrophic collapse of
modern society). Based on my admittedly limited grasp of this frontier of
systematic biology, I suspect that access to such enormous quantities of
data will allow us to develop models and algorithms for recognizing patterns
with very profound implications.  Going out on a limb a bit here, I'll
predict that two such patterns will be: 1) observation of the presence or
absence of a detectable inflection on the continuum of very distant to very
recent genealogical affinities among individual organisms (thereby lending
insight on whether "species" are objective entities to be "discovered", or
merely convenient and subjective entities that are "defined"); and 2) a
renewed appreciation for Richard Dawkins when we can demonstrate the
distinction between the evolutionary history of individual genes, and the
evolutionary history of the organismal machines that have perpetuated those

Both of these patterns, I think, will have great relevance to taxonomy,
systematics, and nomenclature in the years to come -- for obvious reasons.

Having just re-read this, I'm no longer certain about what my original point
was when I began writing it.  I guess it boils down to my oft-repeated
sermon that the primary crisis that Linnaean nomenclature faces involves a
fractioning of interpretations of how best to use it for organizing
information. Should we continue to attempt to bridge the old knowledge to
the new by modifying and contorting the Linnaean system to meet the myriad
of needs?  Or, should we establish new systems better suited for managing
new information (e.g., PhyloCode for managing phylogenetic information, or
DNA Bar-coding for managing genetic data)? I don't have the answers, but I
think it would be helpful if we shared some common sense of the questions --
and what it is we are really arguing about.

One final point:  regarding the issue of matching DNA sequences to species
names -- there's really only one objective way to do this, which is to get
the sequence from the primary type specimen of the name itself.  Everything
else -- from hard-core identification of the sequenced specimen by the
world's greatest taxonomic expert on the group, to a SWAG by a student just
before tissue-munching -- is really just part of a continuum of differing
degrees of subjectivity.  The point is, when it comes time to benchmark DNA
bar-codes against Linnaean-based name definitions, you've really got to
sequence the primary type (or change the Code to allow re-assignment of a
type to a sequencable specimen).  Future technology might allow this to be
more feasible than it currently is (i.e., overcoming the problem of
formalin-fixed specimens; allowing sequences from just a few cells, thereby
leaving the morphology of even the smallest primary types intact, etc.). But
in any case, for the issues that we are discussing here, I believe that
thinking ahead towards the future will help us overcome the growing conflict
between the past and the present.

O.K....that's enough senseless babble from me today...


Richard L. Pyle
Ichthyology, Bishop Museum
1525 Bernice St., Honolulu, HI 96817
Ph: (808)848-4115, Fax: (808)847-8252
email: deepreef at
"The opinions expressed are those of the sender, and not necessarily those
of Bishop Museum."

More information about the Taxacom mailing list