[Taxacom] [tdwg] Semantic Web: What is a species?

Richard Pyle deepreef at bishopmuseum.org
Mon Jan 26 14:04:53 CST 2009

Richard Zander wrote:

> Well, operationally, a species needs to be described and 
> published according to the rules, then survive the 
> marketplace of ideas. 

If by "rules" you mean codes of nomenclature, then they only apply to the
name (label), not the species.  It's the "marketplace of ideas" that serves
as judge & jury for the definition of the species concept.

> This is close to "what a community of 
> taxonomists say it is" and rather different from "what a 
> taxonomist says it is" (which I have also heard).

An unfortunate reality is that in many cases the relevant "community" of
taxonomists consists of only one individual taxonomist. And actually, it's
very often the case that only one taxonomist "said" it (in the form of
describing new species, ansd asserting synonymies); the rest of the
community simply chooses to follow or not.

Richard Jensen said:

> If a species exists in the world, but no human is there to 
> observe it, is it real?

That's the crux of the question.  But I think a better way to think of it
is:  If "extra-terrestrial beings of a super-intelligence" arrived on our
planet and classified all living things, would they come up with the same
units as we have? Would they come up with a hierarchical classification
scheme? If so, would they recognize one unit of classification in the
hierarchy as being "special", moreseo than the others?  Would those units
correlate well with our "species"? Would they draw stick-figure cladograms
as representations of evolutionary affinity?  I suspect that in some cases
there would be good congruency, and in other cases, not so much.

A parallel test we can do without the extra-terrestrials is to compare
traditional nomenclatures created by native peoples, independently of
western science, and see what the correlation is.  In my experience, the
correlation is reasonably good in some cases, and in other cases ... not so

> I am not vain enough to believe that all other species exist 
> solely because we humans are here to make observations about 
> our surroundings. 
> Trees falling in forests make noise whether anyone is there 
> or not - the physical disturbances created by the falling 
> tree, which our brain perceives as noise, still exist. 
> Unless, of course, you believe that nothing exists until or 
> unless it is perceived by a sentient being. If that's the 
> case, then the light coming from a star one million light 
> years away didn't exist until it was sensed. But if it didn't 
> exist until it was sensed, then how did it make the journey? 
> I find it terribly ego-centric to think that nothing else 
> exists except in the context of human perception.

I don't think that's the issue.  Nobody is claiming that organisms don't
exist. The organisms would be here whether or not humans were here to
observe them and (attempt to) classify them.  I think we can all agree with
that (at least for the purposes of this conversation).  

The question is whether there is a special circumscription of organisms
(across time and space) whose boundaries are more clearly defined than other
broader or more restrictive circumscriptions of organisms.  Another way to
visualize it is as some sort of "inflection point" on the evolutionary path,
where the rate of divergence, or propensity to reproduce, or whatever
fundamental metric of "species-ness" you want to use takes a sharp bend,
relative to an otherwise smooth line before and after the bend.  This would
be the "speciation event", and would be objective evidence that species
actually exist as units in nature with some basis in reality other than our
human subjective interpetations. There would always be some fuzziness at the
bend, but the point is that demarcations prior to the bend are in the realm
of subspecies and other such arbitrary clusterings defined by humans, and
demarcations after the bend are in the realm of genera, families, and other
such arbitrary clusterings defined by humans, but the "species" demarcation
has some basis in biological reality that is independent of human subjective

> Exactly what is a "natural" unit that exists in nature 
> independently of human interpretations? Does an atom of gold 
> qualify? How about a planet? 
> Unit implies something that can be individuated, most 
> typically as a consequence of connections, however defined, 
> between its constituent parts. You have suggested that 
> populations may be natural units, but the very concept of 
> population is linked to one or more of the many available 
> species definitions. If the point is that species are not 
> natural, then we have to know what "natural" means. Care to 
> give it a shot?

"Natural" in this sense means "not invented by a human".  The gold atom and
the planet (probably) exist outside of human imagination.  As mentioned
above, individual organisms (probably) exist outside of human imagination.
Populations, subspecies, species, genera, families ... these are all sets of
individual organisms defined by humans.  The question here is whether one of
these kinds of sets (species) has more distinct boundaries based on
properties of the organisms themselves, and is hence more "real"
(independent of human imagination) than the other sets.

My assertion is that "species" do not hold a special place on the scale of
hierarchically nested sets.  I would argue that, as a general rule
(exceptions abound), the more restrictive the set, the more objective its

The set of organisms that includes two parents and all their offspring has
fairly clear and objective boundaries.  Expand that back several generations
(great-great-great grandparents and all of their descendants), and the sets
start to become a bit fuzzy and uneven (especially if individual organisms
produce offspring through multiple mating events with different partners).
Expand back further to what most of us would think of as a population, and
the boundaries become even more fuzzy and arbitrary (i.e., rates of
gene-flow among different populations may not be even or consistent over
time, etc.)  The further back you go (i.e., the more inclusive the sets
become), in my perspective, the less clear the boundaries of the sets

This is contrasted with the "species are real units of nature" perspective,
which maintains that there is a particular zone -- a sweet-spot -- an
inflection point -- a circumscription boundary -- that is "natural"
(independent of human imagination), and therefore "real".  From this
perspective, the task of taxonomists is to discover these circumscription
boundaries, and define species boundaries to be congruent with them.  The
fundamental premise is that the boundaries are defined by some metric of
"propensity to reproduce" (or something similar).

If we define a speciation event as the shift from high "propensity to
reproduce" to low "propensity to reproduce", then cases where this shift is
abrupt (proportional to the time between such shifts along the evolutionary
path) are much easier to visualize as demarcations between species.  And in
such cases, it is very useful and practical to think of species as "real"
units of nature, that exist in nature of human definitions, and are marked
by these abrupt shifts. In cases where the shift is not so abrupt, and
perhaps even continues long enough for another shift to begin (i.e., the
tranisiton periods overlap), then species boundaries become much less clear,
and it is much less useful to think of species as "real" entities in nature.

My contention is, and has been, that almost none of these shifts are
instantaneous (i.e., there are likely to be very few cases where an
individual organism would be regarded as being a different species from its
parents), and as such, there will always be some fuzziness in the

I'd love to ramble on (and on, and on, and on).  But I suspect that only a
very small percentage of readers have made it this far; and I've got other
things to do today.

Sorry for the bandwidth.


More information about the Taxacom mailing list