Message 2002-02-0005: Crown clade definitions (was: Re: interesting style of definition)

Thu, 14 Feb 2002 14:00:23 -0600 (CST)

[Previous by date - Re: interesting style of definition]
[Next by date - Re: interesting style of definition]
[Previous by subject - Cover illustration]
[Next by subject - Crown groups]

Date: Thu, 14 Feb 2002 14:00:23 -0600 (CST)
From: jonathan.r.wagner@mail.utexas.edu
To: Mieczyslaw Wolsan <wolsan@twarda.pan.pl>
Cc: PhyloCode@ouvaxa.cats.ohiou.edu
Subject: Crown clade definitions (was: Re: interesting style of definition)

I believe the only to way to reap the benefits of crown-clade definitions is to 
adopt an *explicitly* crown-based definition. Invoking a crown clade directly 
(the most recent common ancestor of the extant membership of a clade and all of 
its descendants) will always give you the crown clade. Not even explicit 
enumeration of all known extant species in a node-based definition can achieve 
this, because there may be extant species unknown to us that lie outside of the 
clade of currently known members of any particular clade.

In early arguments, the point was made that previously applied crown-based 
definitions were node based (e.g., Aves as being a node-based name specified by 
rattites, tinamous, and neognaths... FYI: the conentuous taxon was Mammalia), 
and thus immune to the "extinction criterion," wherein extinction might 
potentially change taxon content. However, by choosing a few specifier taxa to 
anchor your node-based name, you subject your definition to the possibility 
that further taxonomic resolution may result in the definition not representing 
the crown group. For example, if Aves == vulture + ostrich + tinamou (using 
common names to appeal to the broad membership of this list), under some recent 
hypotheses of avian relationships, song birds might be considered to be outside 
of Aves. Hence, your "crown-group" definition no longer applies. The only true 
crown group name is one defined as such, and I think this new definitional 
format is a good thing.

Mike Keesey's point:
[Does this formulation violate Article 11.3 of the draft PhyloCode] Naturally, 
since the PC is a draft, we *could* address a potential violation by rewriting 
the code to accomodate it. However, there are two considerations:

a) note that the Draft Code explicitly states [Note 9.4.1] "Other wordings and 
other kinds of phylogenetic definitions are possible." We might interpret this 
as a new form of phylogenetic definition (a crown-based definition, formerly a 
stem-modified node-based definition, I believe).

b) more importanltly, [11.1] "Specifiers are species, specimens, or 
synapomorphies cited in a phylogenetic definition of a name as reference points 
that serve to specify the clade [common ancestor - JRW] to which the name 
applies." In the definition given, I don't see "all other extant organisms" (or 
species) being a specifier, nor a group of specifiers, any more than "and all 
of its descendants" is in the common formulation of the node-based definition, 
or, more relevant to the point at hand, "...all taxa sharing a more recent 
common ancestor with..." in a stem-based definition.

c) In effect, this "crown-based definition" is closer to an apomorphy-based 
definition,. The second specifier (apart from "A" in Keesey's version) is the 
specifier "extant." For the record, "existance" is plesiomorphic with the clade 
of all life, hence this is not explicitly an apomorphy based definition. 
However, we can all agree that survival to the Recent is practically meaningful 
and testable, if perhaps not phylogenetically or evolutionarily informative in 
a deterministic sense.

Mieczyslaw Wolsan's points:

1) [Exinction changes species definition] The "extinction criterion" is not 
impossible to avoid, we simply must specify a time slice. For example, 
being "extant" may be defined as possessing at least one living member organism 
on the date of publication, or on January 1, 1900 (GMT), or the year of 
publication of the xth edition of the appropriate work of Linneaus. I favor the 
former alternatives, as associating names used in neontology with the clades 
bracketed by neontological data is more useful in the context of recent study 
(our knowledge base has advanced to a great degree since the thylacine or the 
dodo ceased to be available study animals). By establishing a time cut-off, 
discovering that a species previously believed to be extinct was extant as of 
the cut-off (or vice versa) constitutes an error in our idea of the composition 
of the clade, not a change in its definition or membership.

2) [Does this include the ancestor?] The wording related by Mike Keesey is 
perhaps cumbersome, but it does explicitly include the ancestor: "the CLADE 
stemming from the... ancestor..." where a "clade" is defined as an ancestor and 
all of its descendants.

3) [What is the ancestor?] I feel pendantry is important when it illuminates 
underlying conceptual disagreement. However, the question of to which class of 
entity the term "ancestor" refers as been addressed in this forum in the past. 
Last time it was by me, on the principle that the ancestors of species are 
other species (and only other species, not individuals, not "populations"). 
However, it was made clear to me that this point was being left ambiguous for 
those workers who do not accept species as real. Essentially, if you accept 
species, then the term refers to an ancestral *species* (although some may 
debate this). If you don't, you are on your own, I cannot help you.

But, that's just my opinion, I am probably wrong.

Wagner


  

Feedback to <mike@indexdata.com> is welcome!