[Taxacom] A Cladist is a systematist who seeks a natural classification

Sat Apr 7 16:03:40 CDT 2018

Ken said: "there is no problem with cladistic analysis as an hypothesis generator"

Actually, I think that there is a problem. A method for generating hypotheses does just that, i.e. generates hypotheses, and nothing more. Now, it doesn't actually matter where a hypothesis comes from (i.e. it doesn't matter how it is generated). The (only) value of any hypothesis lies in subsequent testing. Therefore, what advantage does cladistics/phylogenetic analysis have over any other method for generating hypothesis, such as "taxonomic intuition"?

Stephen

--------------------------------------------
On Sun, 8/4/18, Kenneth Kinman <kinman at hotmail.com> wrote:

 Subject: Re: [Taxacom] A Cladist is a systematist who seeks a	natural	classification
 To: "taxacom" <taxacom at mailman.nhm.ku.edu>, "Stephen Thorpe" <stephen_thorpe at yahoo.co.nz>
 Received: Sunday, 8 April, 2018, 1:23 AM

 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.

  _______________________________________________