[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Subject Index][Author Index]

Re: Princeton Field Guide

 > I have 1) a general statement on the matter, and 2) a direct reply
 to one of Greg's arguments.


 I mostly agree with Jaime here, although I differ slightly in
 interpretation. I think that "species" can be defined so as to make
 it a term describing real entities (phylogenetic species concept,
 i.e. distinct evolutionary destinies), but I would also say that
 there are, practical sense, so many hurdles to actually identifying
 and measuring species as distinct entities - and to applying any
 single species concept across all of biology, without special
 pleading for umpteen different subfields - that the concept of
 "species" in any hard-and-fast sense is useless. It is good only as
 an approximation to ease communication somewhat.

I think the real problem is Linnaean nomenclature, which requires us to shoehorn every organism into a species (and a genus) if we are to put it into any kind of system at all.

If we ignore this requirement, and the PhyloCode will let us, everything suddenly looks a lot different. For instance, "phylogenetic species" turns out to be just a synonym of LITU (Least Inclusive Taxonomic Unit = smallest recognizable clade; a term not used by the PhyloCode, but fully compatible with it). Why use the separate term "species" for it? There is no reason.

Other species concepts, too, describe real entities. If you're interested in gene flow between populations, the good old Biological Species Concepts (both of them*) describe useful entities (ring species included), laborious as they are to discover; they're just often not congruent with ecological species, morphological species, or phylogenetic species. All these concepts have nothing in common except the _word_ "species". _There is no_ Platonic ideal of "species" that these concepts approach, as so many people still seem to think.

* _Can_ they produce fertile offspring in captivity vs. _do_ they do so in the wild.

 I believe "species" can be addressed logically, as stated, but not
 observed and measured adequately. That still gives us some basis for
 saying it is a real concept. "Genus," however, has never been
 measured, nor to my knowledge has it ever been given any definition
 that can validate its existence through logic alone.

There is a definition of "genus". It's an extrapolation of the Biological Species Concepts: if they (can) have fertile offspring, they're in the same species; if they (can) have sterile offspring, they're different species in the same genus; if they can't have offspring at all, they're in different genera.

Unsurprisingly, almost nobody has ever tried to use this definition. Like the BSC, it's impossible to use for anything but extant sexually reproducing organisms, and very difficult in practice to apply even to those; like the BSC, it would trigger a lot of lumping, in some cases to the point that several traditional subfamilies would have to be merged into a single genus; and in addition to ring species, we'd get sphere genera.

 The concept is somewhat useful as a sort of "handle" by which we can
 mentally "pick up" various similar species-ish taxa simultaneously -
 a sort of shorthand, really - but that's about it. It lets us
 conceptually grab bunches of species together instead of listing the
 whole lot or constantly laying out the definition of their clade.

The _name_ of the clade in question does this, not the _rank_ of genus (or any other rank).

 Genera are therefore subjective. If we agree on their content by
 convention, then so be it, because we know what we mean by it and it
 hypothetically facilitates communication.

Problem is, there is no consensus, and the ICZN doesn't even try to help:


There are certain underlying principles upon which the Code is based. These are as follows:

(1) The Code refrains from infringing upon taxonomic judgment, which must not be made subject to regulation or restraint.

(2) Nomenclature does not determine the inclusiveness or exclusiveness of any taxon, nor the rank to be accorded to any assemblage of animals, but, rather, provides the name that is to be used for a taxon whatever taxonomic limits and rank are given to it.<<


Anyone can publish a classification that assigns however many species to any genus as they damn well please. And then anyone else can come along and publish a classification that contradicts it. And to all eternity both are correct, and neither is wrong, provided they keep the type species in their genera and obey priority.

(...Never mind that the mandatory suffixes for family-group ranks and the mandatory formats for genus-group and species-group names contradict Principle 2 rather blatantly.)

So, if I say "*Bufo*", do I mean the mid-20th-century *Bufo* which includes all hundreds of bufonids that aren't too weird, or the much, much smaller Frost et al. (2006) concept of *Bufo* which is limited to a couple of Eurasian species, so that the cane toad is *Rhinella marina*? The name alone doesn't tell you, and neither does the (unchanged) genus rank of that name.

 If we start talking about inclusion or exclusion of taxa according to
 phylogeny as elucidated by traits, we're applying objective reasoning
 to a subjective construct. That does not work. Only subjective
 consensus applies to defining subjective phenomena. So saying that we
 should or shouldn't include taxa in a genus based "only" on traits
 like frill horns or whatnot is absurd, because we can subjectively
 attach as much or as little significance to any given trait as we
 collectively agree to do.

We don't even need to agree. It's enough that the ICZN lets us.

 Greg's subjective preference that we give less value to such
 features is not wrong by any means, it's just not the standard agreed
 upon by a larger set of people-interested-in-the-issue.

There isn't a standard in the first place.

 Greg's standard has therefore been excluded by subjective consensus;

That consensus consists only of which species to include in which genera of Mesozoic dinosaurs. There's not much, if any, of a general concept behind it.


 Some point-by-point responses to a particular line of reasoning:

 "It is not possible to define species either, but species are real."

 As far as science is concerned. all real things have a definition.


 "It is not possible to define life, but it is real."

 "Life" is a loaded word. I would argue that reasonable definitions
 for it do exist, as long as we're careful about "unloading" it first
 to avoid confusion.

It's a continuum that our languages fail to express. Therefore, where in the continuum we decide to draw the line is completely subjective. Are viruses living beings? Depends on how we choose to define "life". :-| If we want viruses to count as "alive", we'll likely define "life" along criteria of information; if we don't want them to count, we'll likely use things like independent metabolism as defining. It's up to us.

 "There is no way to precisely define battleship versus battlecruiser
 (HMS HOOD for example) versus heavy cruiser (Alaska class for

 That's because, philosophically, there's no such thing as a
 "battleship," "battlecruiser, or "cruiser." Let me clarify. The
 designs for those vessels exist; the vessels themselves exist as
 physical objects. They are real and they are mathematically
 definable. But the categorizations for them are ultimately arbitrary
 human labels. Yes, one could say "semi-arbitrary," because tonnage
 and certain other factors are measurable as portions of a continuum,
 which is what such terms are attempting to encapsulate. Yet, like
 language in general, the labels themselves have no fixed meaning, and
 further, there are so many continua being lumped together (tonnage,
 dimensions, # of guns, etc) that a huge number of variations along
 one or all continua are possible. We could define every possible
 combination along every last axis if we had the time and inclination
 to do so, but it's such a complex question that no one has the
 wherewithal to attempt it, and that would still leave us with a
 radically different form of classification.

 In point of fact, there is no *empirical* dividing line along any
 continuum at which one category turns into another, except in our
 own minds due our own subjective preferences (because in this case
 we *want* categories, so we delineate them at our convenience. We
 could choose to lump them all into a single category if we desired to
 do so, or we could break them down until we only talk about specific
 designs, specific mission profiles, or specific vessels). The
 physical world only shows that one ship has X guns and another ship
 has Y, *not* that this difference has significance in any absolute
 sense, and therefore *not* that this difference demands any
 *specific* nomenclature for categorization. The only significance is
 in our minds, and hence the terms we apply to individual designs are
 still arbitrary designations - except insofar as we wish to
 communicate useful information (so while we could call a
 "battlecruiser" a "flying spaghetti warship" if we so wished, we
 prefer to use the term that will convey meaning more effectively to
 those who might otherwise be confused).

You are, thus, saying that, "philosophically, there's no such thing as" life.

I'm fine with that.

 "Many terms are approximations, it does not mean that they cannot be
 used to describe and distinguish basic, comparable types."

 Right. Approximations. General notions. Not things we can be
 definite about. So not, for example, things we can definitely say
 include or exclude any one taxon.

Phylogenetic nomenclature escapes this by using objectively discoverable events to define clade names. Rank-based nomenclature doesn't even try.

That's because rank-based nomenclature is deliberately theory-free. Both Charles Darwin and Richard Owen, together with many others, are among the authors of the 1842/3 Strickland code of zoological nomenclature, so it's a worst-of-all-worlds compromise. It treats organisms as battleships, battle cruisers and heavy cruisers. Phylogenetic nomenclature, on the other hand, assumes evolution and uses tree-thinking; this makes everything easier.