[Taxacom] A Cladist is a systematist who seeks a natural classification
John Grehan
calabar.john at gmail.com
Sat Apr 7 12:45:16 CDT 2018
I don't think I recall Richard posting on Taxacom of a basic summary of his
non-cladistic method for constructing 'evolutionary trees'. I think that
would be of some interest. Also it would be interesting to have some
citations of various systematists who have published evolutionary trees
using Richard's methodology.
John Grehan
On Sat, Apr 7, 2018 at 1:30 PM, Richard Zander <Richard.Zander at mobot.org>
wrote:
> I cannot find the macro key that I once I used to list 20 reasons why
> cladistics, phylogenetics, and molecular systematics were a scam, a grift,
> and a shuck, respectively. Read my books (available for cheap on Amazon)
> which explain in detail the above, with diagrams, aphorisms, copious
> bulleted lists, worked out examples, and interesting descriptive verbiage.
>
> I think a cladist is a person who seeks to substitute a dichotomous tree
> for an evolutionary tree because that person l o v e s his or her computer,
> which only burps up dichotomous trees. Multifurcations are as cursed as is
> paraphyly. Evolutionary theory be damned.
>
> -----Original Message-----
> From: Taxacom [mailto:taxacom-bounces at mailman.nhm.ku.edu] On Behalf Of
> Kenneth Kinman
> Sent: Saturday, April 7, 2018 7:24 AM
> To: taxacom <taxacom at mailman.nhm.ku.edu>; Stephen Thorpe <
> stephen_thorpe at yahoo.co.nz>
> Subject: Re: [Taxacom] A Cladist is a systematist who seeks a natural
> classification
>
> Hi all,
>
> There is no problem with cladistic analysis as an hypothesis
> generator. The problem is that phylogenetic systematists only formally
> recognize taxa which are clades. By branding paraphyletic taxa as
> unnatural and refusing to recognize any of them, they often fail to put in
> the added work of incorporating divergence information into their
> classifications when it would make them more stable and usable (as
> advocated by Mayr, Ashlock, Cavalier-Smith, and other evolutionary
> systematists).
>
>
> This is especially true of higher taxa (families to kingdoms). It
> is therefore no surprise that it is at the level of Kingdoms, Phyla, and
> Classes that the debate between evolutionary systematists and phylogenetic
> systematists is most heated. Phylogenetic systematists have too often
> generated instability at those levels, and thus severely affecting
> usability.
>
>
> That is why Ernst Mayr called them cladifications (not
> classifications). At the level of species and genera, cladifications often
> turn out to be good classifications, but the same is too often not true at
> higher taxonomic levels. The worst case is the Three Domain cladification
> which was (and continues to be) horribly simplistic. It is people like
> Cavalier-Smith who is putting in the hard work of attempting to construct
> more natural, stable, and usable classifications.
>
> --------------Ken
>
>
> ________________________________
> From: Taxacom <taxacom-bounces at mailman.nhm.ku.edu> on behalf of Stephen
> Thorpe <stephen_thorpe at yahoo.co.nz>
> Sent: Saturday, April 7, 2018 1:01 AM
> To: taxacom; John Grehan
> Subject: Re: [Taxacom] A Cladist is a systematist who seeks a natural
> classification
>
> The issue that I still haven't been able to get a clear answer to is
> whether cladistics is just a way of generating hypotheses for future
> testing (which, as we all know, is ongoing and never conclusive), or
> whether it somehow generates something which can be more or less thought of
> as a "fact", i.e. something which is at least more likely to be "true" than
> not. My own suspicion is the former, i.e. just a hypothesis generator,
> based on various assumptions (such a s parsimony) and given values of
> certain variables (weightings, etc.) which may themselves be quite
> subjective.
>
> Stephen
>
> --------------------------------------------
> On Sat, 7/4/18, John Grehan <calabar.john at gmail.com> wrote:
>
> Subject: [Taxacom] A Cladist is a systematist who seeks a natural
> classification
> To: "taxacom" <taxacom at mailman.nhm.ku.edu>
> Received: Saturday, 7 April, 2018, 6:10 PM
>
> Since there are at various times some strong opinions on cladistics and
> on natural classification I have pasted below the text of a recent
> article that might be of interest to some (some typos may have crept in
> during the copy/paste).
>
> Biol Philos (2018) 33:10
> https://doi.org/10.1007/s10539-018-9621-7
>
> David M. Williams & Malte C. Ebach
>
> A Cladist is a systematist who seeks a
> natural classifcation: some comments
> on Quinn (2017)
>
> Abstract. In response to Quinn (Biol
> Philos, 2017.
> https://doi.org/10.1007/s1053
> 9-017-9577-z) we identify cladistics to be about natural
> classifications and their discovery and thereby propose to add an eighth
> cladistic defnition to Quinn’s list, namely the systematist who seeks to
> discover natural classifications, regardless of their affiliation,
> theoretical or methodological justifications.
>
> Derived from various permutations of
> phylogeny, biology, philosophy,
> methodology, sociology, loyalty etc.,
> Aleta Quinn recently proposed “seven
> specific definitions that capture
> distinct contemporary uses” of cladistics (Quinn 2017, p. 1). Our own
> efforts, based on the same criteria, yielded a further seven, which we do
> not intend to bore our readers with here. We are sure more could be found
> and more people could be found who subscribe/correspond to them. Suffice
> to say, one might find definitions for anything—and in any case, Quinn
> was clear about her motives:“I do not intend to classify individuals,
> ideas, or research programs. Rather, I clarify distinct things that
> speakers mean by the term ‘cladist’” (Quinn 2017, p. 1). Depending on
> one’s outlook—philosopher, historian, biologist, even sociologist (Hull
> 1988)—the definitions might help progress their subject. As biologists,
> we found much to think about but rather than dissecting the minutiae, we
> seek to clarify by attempting to simplify.
> We need first to dispense with one
> misconception. Quinn draws upon a
> commonly preconceived notion, namely
> that systematics requires evolution as
> a prior condition:1
>
> “What that theoretical foundation may
> have been [in reference to de
> Candolle’s
> view on characters] is not relevant to
> my points about contemporary
> systematics,
> whose conceptual framework presupposes
> the concept of evolution” (Quinn
> 2017, footnote 11).
>
> Consider the concept of a cladogram,
> which everyone might agree is a
> branching diagram commonly included as
> part of the results of a cladistic
> analysis. One might derive from this
> diagram which taxon is more closely
> related to itself than to any other.
> One might explain this relationship by
> common descent. The cladogram, however, need not be constructed with
> any evolutionary assumptions in mind; rather, the evolutionary
> assumptions serve to explain why one taxon is more closely related to
> itself than any other.
>
> The search for a natural classifcation
> was established prior to the
> adoption of
> any theory of evolution. In fact
> Augustin P. de Candolle’s had a great deal to say on the matter,
> especially the differences between natural and artificial classifications
> (Candolle 1913). But de Candolle was working some time ago, so what, if
> anything, might be his relevance today? Methods of systematics change as
> time passes. But all methods fnd cladograms, in the sense that the
> results yield sets of relationships, either as a branching diagram or as
> a written classification. Regardless of method, which of these
> relationships might be considered to reflect something that actually
> exists, rather than a product (an
> artefact) of the method? How can any
> method achieve that without knowing the answer beforehand? Obviously it
> can’t. One might play around with simulation studies to judge the
> performance of any suite of methods, or one might delve into philosophy
> to create justification, but in the court of last resort all that remains
> are sets of cladograms that either agree or disagree to a greater or
> lesser extent in terms of common relationships found. That is, they agree
> in the cladistic parameter, the relationships specified—that the signal
> to noise ratio is working in our favour, as is evident from
> classifications of the past. Here we might argue that natural
> classification is the result derived from several cladograms, regardless
> as to how they were arrived at; artificial classifications are derived
> from a specific method, be that Wagner parsimony, UPGMA, maximum
> likelihood and so on, or from a specific source of data (DNA,
> ultrastructure, etc.), and so on. Why are these artificial? Because a
> method, any method, assumes the results that are required (the shortest
> tree; or the most similar taxa grouped together; or the most similar taxa
> grouped together via a weighted model of character change, etc.); for a
> data source, they assume those data are privileged over other data (DNA
> must be the source of ‘true’ relationships, etc.).
> Cladistics, in its most general sense,
> does not associate with any one
> method, or any one data source. It
> applies to sets of relationships—it is
> the set of relationships. This is
> effectively what de Candolle argued for, and has been the basis of
> systematics for decades, if not centuries:
>
> “For the last 50 years and
> more—even now continuing into the realm of nomenclature—in the name of
> the modern and the new, Visionaries aim, as it were, to confine the past
> to a dustbin of history, and to bolt and lock the lid upon it. As if
> without it, we be in some way better, even born again more whole-some;
> as if Carl Linnaeus really were among the last of the Ancients, and not,
> rightly, the first of the moderns, and so related to us—of a group
> inclusive of us” (Annual Review of the Linnean Society, 2001).
>
> These words, not readily accessible,
> were spoken by Gareth Nelson after
> receiving the Linnean Gold Medal and
> re-cast above as part of the 2001
> Annual Review of the Linnean Society,
> London. Linnaeus as the first of the
> moderns? Among other matters, Linnaeus
> spoke of the differences between
> artificial and natural classification,
> a subject taken up and developed by
> de Candolle (1913). One might cast that debate in very simple terms:
> artificial classifications are found by imposition, natural
> classification is discovered. Imposition implies some method or
> motivation to erect a particular classification, such as a field guide or
> handbook for identifying specimens—today it is more likely those would be
> websites, or online interactive guides. There is nothing wrong with
> artificial classifications. We both use them all the time, almost every
> day ( https://www.trilobites.info/; http://naturalhistory.
> museumwales.ac.uk/diatoms/). But whatever merits they have, and there
> are many, they are created by acts of imposition. We ask our readers,
> then, if they would consider analysis of some data with one or another
> statistical program, or with one or another parsimony program, or with
> one or another phenetic program, whether this is an act of imposition or
> an act of discovery? We see it as an act of imposition. How could it be
> otherwise? Cladistics, then, is about discovery, about finding repeating
> patterns,finding the same relationships, finding relationships that are
> not method dependent, finding relationships that are reflections of the
> world as it is:
>
> “What, then, of cladistics in
> relation to the history of systematics? If cladistics is merely a
> restatement of the principles of natural classifcation, why has
> cladistics been the subject of argument? I suspect that the argument is
> largely misplaced, and that the misplacement stems, as de Candolle
> suggests, from confounding the goals of artifcial and natural systems”
> (Nelson 1979, p.
> 20).
>
>
> For us, cladistics is about natural
> classifcations and their discovery, an
> activity
> that occurs with or without
> “knowledge of process”. Look in museums, herbaria, universities and
> other institutions that still hire systematists and you will see:
>
> Cladist (viii): A cladist is a
> systematist who seeks to discover natural classifications.
> _______________________________________________
>
> _______________________________________________
> Taxacom Mailing List
> Taxacom at mailman.nhm.ku.edu,
> http://mailman.nhm.ku.edu/cgi-bin/mailman/listinfo/taxacom
> The Taxacom Archive back to 1992 may be searched at:
> http://taxacom.markmail.org
>
> Send Taxacom mailing list submissions to taxacom at mailman.nhm.ku.edu To
> subscribe or unsubscribe via the Web, visit:
> http://mailman.nhm.ku.edu/cgi-bin/mailman/listinfo/taxacom
> You can reach the person managing the list at:
> taxacom-owner at mailman.nhm.ku.edu
>
> Nurturing Nuance while Assaulting Ambiguity for 31 Some Years, 1987-2018.
> _______________________________________________
> Taxacom Mailing List
> Taxacom at mailman.nhm.ku.edu,
> http://mailman.nhm.ku.edu/cgi-bin/mailman/listinfo/taxacom
> The Taxacom Archive back to 1992 may be searched at:
> http://taxacom.markmail.org
>
> Send Taxacom mailing list submissions to taxacom at mailman.nhm.ku.edu
> To subscribe or unsubscribe via the Web, visit:
> http://mailman.nhm.ku.edu/cgi-bin/mailman/listinfo/taxacom
> You can reach the person managing the list at:
> taxacom-owner at mailman.nhm.ku.edu
>
> Nurturing Nuance while Assaulting Ambiguity for 31 Some Years, 1987-2018.
>
More information about the Taxacom
mailing list