[Taxacom] Are species real? Doesn't matter.

Richard Pyle deepreef at bishopmuseum.org
Thu May 31 15:42:14 CDT 2007

> As I read through these posts, I couldn't help but wonder if 
> those of you responding view species as classes, rather than 
> individuals. 

Yes, I do view species as classes, rather than individuals.

> My view is that species are real, but the best we can do with 
> our names is to designate hypotheses about what these real 
> entities are [e.g., Homo sapiens is, hypothetically, a real 
> entity, a product of a natural process (evolution); and, it 
> appears to be a pretty good hypothesis given that there are 
> few instances of questionable human beings - I know, certain 
> deans and administrators are probably beneath being human, 
> but they're a special case]. Because these are hypotheses, we 
> must accumulate evidence to test our hypotheses and reject 
> them when when the evidence indicates that we were mistaken. 
> And, there are competing hypotheses as to what constitutes a 
> given species. Until such time as we can develop a consensus 
> about the status of each named species, it is important that 
> we be aware of each hypothesis so we can disentangle the 
> historical treatments of each named species.

I still don't buy the "species as hypothesis" argument.  What are we
testing?  That no individual organism exists on the planet today that is
ambiguous as to whether it belongs to the species H. sapiens?  That no such
organism ever existed in the past?  That no such organism will ever exist in
the future?  Would it matter if such an organism never developed past the
embryonic stage?  Would it matter if it was created in a laboratory?

Or, is the hypothesis about frequency or probability of gene flow between
two given organisms or populations; one within the species H. sapiens, and
one outside?  Must gene flow occur through mating of two individual
organisms, or does it count if it is lateral, through an intermediate (e.g.,
a bacterium)?  And how would gene flow be defined in this context -- would a
single exchange count, or must there be a pattern of multiple events of gene
exchange (through generations and/or across multiple individuals)?

Or, is there some other metric against which we can test whether a
particular individual falls inside, or outside, of a "real" species?

As I've said before, we can reasonably establish hypotheses about pedigrees
of individuals, and even (for the most part) phylogenetic affinities, and we
can test these hypotheses with current and future techniques -- based on the
assumption that there was some "real" series of historical events over large
time scales that led to some observed pattern of heterogeny or clustering of
similar organisms observed in the present day.  But unless you can
articulate some biological metric by which we could, with perfect knowledge,
unambiguously distinguish whether (for example) individuals in one
population "are" (as opposed to "should be treated as") a separate species
from the individuals in another population, then taxa (of all ranks) are
classes that we define, not real entities that we discover through
ever-increasing accumulation of knowledge.

If we start with the premise that our reason for establishing taxonomic
units (including species) and putting labels on them is in order to
facilitate information exchange among biologists, then you could construct a
hypothesis along the lines of: "treating individual organisms among these
two populations as separate species, rather than as separate subspecies,
allows more effective and efficient communication amongst biologists".  But
this wouldn't be a scientific hypothesis, it would be a sociological one,
and the methods to test the hypothesis would involve examination of
information exchange patterns among humans, rather than any attributes of
the individual organisms within the two populations.

No matter how I look at it, I keep coming to the conclusion that taxa of all
ranks (including species, as well as supraspecific and subspecific ranks)
are simply a human-defined mechanism to allow us to more effectively
communicate with each other about what is otherwise a near-perfect continuum
of genetic information flow, starting with the very first self-replicating
genome and continuing, uninterrupted for billions of years, to the spectrum
of diversity we see around us today. The only real granularity to that
otherwise smooth continuum are the individual DNA-replication events, within
every cell that has ever reproduced.  A reasonable case could be made that
clusters of these cell-reproduction events can be encapsulated into what we
would all think of as a single multicelluar organism (but even this has
caveats -- e.g., only a small fraction of the cells in your body contain
human DNA; how long does the average DNA molecule exist intact within your
body, compared to your entire lifespan; etc.).  So already at the level of
an individual organism, things can be a bit dicey -- and we still have
several constructs of "populations" to go through before we get to

It's theoretically possible to construct a set of objective criteria by
which we might be able to test whether two individual organisms belong to
the same species, or different species, according to those criteria.  The
problem is, as far as I can tell, every attempt to do this in any sort of
universal way has failed for one reason or another.  None of the methods for
defining a species that I've read about adequately accommodate all organisms
(e.g., both sexual and asexual reproducers, both unicellular and
multicelluar, etc.) across time (fossils and extant organisms).  Personally,
I think that the defacto definition of a species for the past 250 years or
so (i.e., a species is what a competent taxonomist or community of
taxonomists says it is) is far and away the most practical and useful
(though certainly not perfect), in terms of serving as a mechanism to allow
biologists to communicate with each other.

But if one thing is overwhelmingly obvious, it is that I've spent WAY too
much time writing emails this morning (not the first time), so as much as
I'd like to carry on this line of philosophical reasoning, I really ought to
cut it off here.  Many of these points (e.g., species as classes, rather
than individuals) have been on this list before, and should be somewhere in
the archives.

My apologies for the cluttering of inboxes this morning.


More information about the Taxacom mailing list