From Wikipedia, the free encyclopedia
Jump to: navigation, search
Former featured article Cladistics is a former featured article. Please see the links under Article milestones below for its original nomination page (for older articles, check the nomination archive) and why it was removed.
Main Page trophy This article appeared on Wikipedia's Main Page as Today's featured article on May 4, 2004.
Article milestones
Date Process Result
January 19, 2004 Refreshing brilliant prose Kept
January 21, 2007 Featured article review Demoted
Current status: Former featured article

To do list 2007[edit]

As it stands, this article confounds "cladistic" and Linnaean classification. What is being contrasted is not "cladistics versus Linnaean" but "Phylocode versus Linnaean." Fact is: any classification that contains monophyletic groups and is logically consistent with the underlying phylogeny is a "cladistic" classification. it may be Linnaean, an indented list, a classification that uses numerical prefixes, or a system that uses Phylocode. The article also confounds "Linnaean and evolutionary taxonomic classifications." A Linnaean classification that contains known paraphyletic groups is an evolutionary classification. A Linnaean classification that contains only monophyletic groups is a cladistic classification. In short, this article is in serious need of revision. Eowiley (talk) 19:57, 6 April 2009 (UTC).

  • Here are some suggestions to improve the article:
  1. I think the distinction between cladistics and phenetics should be addressed a second time once plesiomorphic and apomorphic terms are introduced. Cladistics is interested in synapomorphies. Phenetics does not distinguish between the two. An example of how the two could come up with a different tree would also help.
  2. I'm a bit unhappy with the discussion of what it means to be basal. I completely agree that usage of the term primarily refers to a taxon-poor clade that branches off early. I also think that the term gets used in reference to the ingroup, the taxon sampling, and the question being asked. For example, gibbons will commonly be said to be basal among the hominoids, yet there are 13 species of gibbons in four genera and only 7 species of great apes (also 4 genera). In this case, the research question usually being posed is really about a focus organism (us) and relationships among the gibbons is less important in that particular discussion. Bats and insectivorans are basal to the rest of the Laurasiatheria in spite of the fact that over 50% of described laurasiatherian species are bats and 20% are insectivorans. The research question is how are bats, insectivorans, carnivorans, pangolins, perissodactyls, and cetartiodactyls. From that perspective, bats and insectivorans to qualify as basal to the cetferungulates. Being "primitive" shouldn't qualify a group as basal (although it probably is used that way in some instances). Bats fly, echolocate, and look nothing like the ancestor of the Laurasiatheria.
  3. The distinction between synapomorphy and autapomorphy should be clarified.
  4. The second paragraph of the section titled "Cladistic methods" is confusing. Plesiomorphies were present in the last common ancestor of group discussed. Apomorphies arose subsequent to the last common ancestor of the group discussed. To say that an apomorphy was present in the last common ancestor of the ingroup is false. A synapomorphy was present in the last common ancestor of the clade it characterizes (and may have arisen anywhere along the branch leading to that clade). Autapomorphies are also a type of apomorphy and they weren't present in the last common ancestor of any two taxa in the analysis.
  5. Eliminate the use of "we" in the 4th paragraph of the same section.
  6. I think at least 50% of the field would consider maximum likelihood and Bayesian methods to be both Hennigian and cladistic. They are still constructed on the basis of synapomorphies, they just incorporate information about how characters evolve and attempt to incorporate the potential for additional evolution hidden in a final parsimony analysis. They are definitely not phenetic methods. I'm also amazed that there still isn't an article on maximum likelihood in phylogenetics.
  7. The total evidence approach advocated in the 6th paragraph isn't universally accepted. I think it's safe to say that >50% of the field would agree, but there are those who argue that a little bit of quality data is better than a lot of noisy data or even a little bit of quality data + some noisy data. Most (but not all) do agree that data where homology is questionable should be excluded. That should be addressed in the paragraph as well and I'm not all that comfortable with the behavior example (without expansion and clarification) for that reason. That statement that molecular, morphological, etc. not are all equal is definitely an opinion and is definitely disputed. Homoplasy is more common in morphological data? Are we sure about that?
  8. Paragraph 7. A small point, but cladistics does assume that evolution is bifurcating as opposed to hybridizing, reticulate, or having lateral transfer.
  9. In my opinion, the "Cladistic classification" section can reasonably stay, but seesm to ramble on as if it was written by several editors who had differing opinions and tried to jump back and forth in such a way as to make it sum up to NPOV. I'm not happy with the notion that about half of the text of a featured article on cladistics is spent discussing the PhyloCode and Linnean hierarchy.
  10. There is a subtle, but important philosophical difference between cladistics and parsimony. This article should address that clearly.
  11. The "see also" is an odd list. It should have links to phenetics, parsimony, maximum likelihood (phylogenetics), maybe Bayesian (phylogenetics), as well as some of what's already there.

--Aranae 02:27, 15 December 2006 (UTC)

Eyu100(t|fr|Version 1.0 Editorial Team) 15:36, 22 August 2007 (UTC)

Definition of Cladistics[edit]

The current definition of classifying based on shared ancestry is both deeply wrong and uninformative. Ancestry cannot be known in advance of cladistic analysis, so it cannot be said to be a basis for it. Cladistics is distinguished from other fields by classifying according to monophyletic groups (hence cladistics), which are based on shared derived (or unique) features—synapomorphies. The concept of grouping by synapomorphy, and only by synapomorphy, is the novel and distinguishing contribution of cladistics. The notion of grouping by ancestry is not novel to cladistics, and is in fact a better characterization of evolutionary taxonomy. Secondly, cladistics does not attempt to "construct a tree representing the ancestry of organisms and species." In fact, cladograms are completely silent on the question of ancestors, as all taxa are grouped hierarchically as sister taxa. I think we should restore my edits to the first paragraph from several months ago. —Preceding unsigned comment added by The Braz (talkcontribs) 08:14, 2 November 2009 (UTC)

I don't understand some of your criticisms. For instance, you are correct that "ancestry cannot be known in advance of cladistic analysis", but incorrect in thinking that this means a cladistic classification is not based on shared ancestry; the classification comes after inference of phylogeny, and is based on the inferred phylogeny. Your comments on ancestry appear to be rather semantic. A cladogram is a representation of the inferred ancestry of a set of (usually) extant species. The goal is to infer how species within a certain slice of time (usually, but not necessarily, the present) are related to each other, i.e., in what order they are united by common ancestors. If we have fossils that may represent some of those ancestors, these can of course be mapped onto the tree; but while they will often be quite helpful (especially in attempts to determine the ages of lineages), including ancestral taxa is not necessary to infer relationships among the extant taxa. IOW, based on shared similarities at the present, we can infer that two species shared a recent common ancestor even if we don't happen to have a fossil specimen that we believe to represent that common ancestor.Paalexan (talk) 19:17, 15 November 2009 (UTC)
A cladogram contains no ancestors, and thus is mostly silent on ancestry. A cladogram is only a summary of the most parsimonious distribution of character states. The problem of course, is that defining cladistics as a field of systematics that classifies according to ancestry is not a good start to the article. Evolutionary taxonomy also classifies according to ancestry (or some perception thereof). Cladistics groups things according to shared apomorphy (synapomorphy) which diagnoses monophyletic groups, and that is what distinguishes it from other methods that preceded it. Cladistic classifications are not based on shared ancestry, rather inferences of shared ancestry are based on cladistic classifications. I don't think this is semantic, I think it matters because one approach is purely circular and the other is not.
Cladistics might be described as a method of trying to infer evolutionary relationships, but it consequently cannot be said to offer a classification based on evolutionary relationships. Otherwise, it is circular reasoning.The Braz (talk) 15:38, 23 November 2009 (UTC)
Why must an assessment of ancestry directly include the ancestors? For instance, suppose we were grouping a set of children into families. The parents of the children are assigned no place in such a grouping; is the grouping therefore independent of ancestry? Of course not; the parents simply aren't members of the set of individuals being grouped based on ancestry. Your suggestion that cladistic classifications are based on "inferences" of shared ancestry rather than on shared ancestry itself is accurate, but does not seem relevant. All scientific knowledge that extends beyond direct observation (we could argue even on direct observation, but there would be no apparent purpose in doing so...) involves inference; should we then preface every scientific article on wikipedia with some sort of statement that the knowledge discussed in based on inferences? Pointing out that cladistics involves inference does not provide any noteworthy or unique information; the same will be true of any scientific alternative.
I also do not understand your suggestion that there is circularity involved. A phylogeny is an inferred set of evolutionary relationships, which can then be used as the basis for nomenclature. First we try to understand relationships among terminals, then we apply names to groups of terminals based on the inferred relationships. Perhaps you could explain what is circular about this?
Further, the issue is to some extent moot as far as the article itself goes; an encyclopedic article on an academic subdiscipline should reflect mainstream understanding of that subdiscipline, rather than being a venue to attack that subdiscipline. This page already suffers from that problem rather severely... hopefully I can get around to poking at it at some point.Paalexan (talk) 08:06, 29 November 2009 (UTC)
Sorry for the long delay in my response. However, I think you've almost entirely misunderstood what I said. Let's start from the beginning: You write: "Why must an assessment of ancestry directly include the ancestors? For instance, suppose we were grouping a set of children into families. The parents of the children are assigned no place in such a grouping; is the grouping therefore independent of ancestry? Of course not; the parents simply aren't members of the set of individuals being grouped based on ancestry." You haven't specified here what the grouping was based on. So your example has no relevance to the question. What justifies the grouping? If you knew that the children were descended from certain parents, then you could indeed base the clasification on common ancestry, because that is knowledge you have in advance of the grouping. In cladistics, groupings are always justified by synapomorphy. They are never justified by hypotheses of common ancestry, because the degree or level of common ancestry (what cladistics can actually discover) is not known in advance of the analysis. Common ancestry is inferred by means of synapomorphy. Thus, synapomorphies are discovered and justify the hypothesis of common ancestry. This is what sets cladistics apart from evolutionary taxonomy. The search for synapomorphy is also what sets it apart from phenetics. Therefore, cladistics is best described by what it actually does.
You write: "All scientific knowledge that extends beyond direct observation (we could argue even on direct observation, but there would be no apparent purpose in doing so...) involves inference; should we then preface every scientific article on wikipedia with some sort of statement that the knowledge discussed in based on inferences? Pointing out that cladistics involves inference does not provide any noteworthy or unique information; the same will be true of any scientific alternative." I'm not sure what this means, or where you get this from. I said nothing at all like this. There is, however, a logical direction in which inference proceeds from evidence. Cladistics, by contrast with evolutionary taxonomy, is a great advance in formalizing this relationship between evidence and inference.
You write: "I also do not understand your suggestion that there is circularity involved. A phylogeny is an inferred set of evolutionary relationships, which can then be used as the basis for nomenclature. First we try to understand relationships among terminals, then we apply names to groups of terminals based on the inferred relationships. Perhaps you could explain what is circular about this?" This is nomenclature, not classification. I never said anything about names. What I said was: "Cladistics might be described as a method of trying to infer evolutionary relationships, but it consequently cannot be said to offer a classification based on evolutionary relationships. Otherwise, it is circular reasoning". If you use cladistics to infer evolutionary relationships, and then say that cladistics is based on evolutionary relationships, you're saying something that is effectively circular (we're then back in evolutionary taxonomy land). Cladistic results are based on synapomorphy.
I think I see the problem, though, and it's made clear by your earlier statement: 'For instance, you are correct that "ancestry cannot be known in advance of cladistic analysis", but incorrect in thinking that this means a cladistic classification is not based on shared ancestry; the classification comes after inference of phylogeny, and is based on the inferred phylogeny'. You have your epistemology backwards because you see the classification as something distinct from the construction of a cladogram. At this point, we run the risk of descending into a debate about classification, etc. The point is that if you want to understand my arguments, then at least understand that I don't see classification as a distinct process from cladogram construction. I don't see why anyone should. Phylogeny, on the other hand, is a hypothetical abstraction we make beyond the inference of a cladogram/classification.
I'm not sure what sort of "attack" you're talking about. I'm in fact trying to keep this pared down to the basics, and refer to those aspects which would be least likely to cause controversy if a person were asked to describe what cladistics is. I think that by saying cladistics is based on common ancestry is vague, misleading, and not very informative. From a methodological point of view, it is wrong. The Internet, however, appears to have developed its own version of what cladistics is about. This is due, in part, to the wide spread notion that cladistics classifications are based on common ancestry. No they are based on synapomorphy. Cladistics is only a means of finding relative degrees of common ancestry. Thus, it may be described as a system of classification that makes statements of relative degrees of common ancestry, but it is meaningless to say that the classifications are 'based on common ancestry', since inferences of common ancestry are actually based on cladistic knowledge.The Braz (talk) 16:42, 25 January 2010 (UTC)


The article at present seems to suggest that cladistics and the PhyloCode are inseparably linked; that all cladists are inherently advocates of the PhyloCode. For instance, the link for "phylogenetic systematics" in the introductory sentence discussing Hennig leads one instead to phylogenetic nomenclature, i.e., the PhyloCode; Hennig never advocated such a system, but the naive reader would likely be led to think otherwise. Later, we encounter the sentence: "These traditional approaches, still in use by some researchers (especially in works intended for a more general audience[23]) use several fixed levels of a hierarchy, such as kingdom, phylum, class, order, and family. Cladistics does not use those terms, because one of the fundamental premises of cladistics is that the evolutionary tree is so deep and so complex that it is inadvisable to set a fixed number of levels." This is simply false. Advocates of abandoning ranks are "cladists" in only the broadest possible sense, and in that broad sense they are a small (but admittedly vocal) minority. Similarly, the figure demonstrating different PhyloCode approaches for delimiting taxa has no place here. Unfortunately, there are so many errors embedded throughout this article that I am not sure I can fix them without simply rewriting the whole thing! Paalexan (talk) 19:26, 15 November 2009 (UTC)


Apomorph redirects here, but is not even mentioned in the article! (talk) 05:39, 22 November 2009 (UTC)

There is a discussion of apomorphy, synapomorphy, and symplesiomorphy, all terms that are defined within the field of cladistics. Unless there is another meaning of "apomorph" that I am not familiar with, this would seem to be the appropriate place for a redirect to point to.Paalexan (talk) 08:12, 29 November 2009 (UTC)

Popularity of phenetics?[edit]

The article says of phenetic classification that

For some decades in the mid to late twentieth century, a commonly used methodology was phenetics ("numerical taxonomy"). This can be seen as a predecessor to some methods of today's cladistics (namely distance matrix methods like neighbor-joining), but made no attempt to resolve phylogeny, only similarities.

I think this exaggerates the importance of phenetic classification as a method in wide use (it certainly had an important intellectual influence). I do not recall a single year from about 1963 (when Sokal and Sneath's Principles of Numerical Taxonomy was published) to 1980 (when phylogenetic systematics was becoming dominant) that phenetic classifications were more common than other approaches such as Mayr and Simpson's "evolutionary systematics". I am also intrigued by the idea that distance matrix methods are part of "cladistics" as they are also frequently denounced as "phenetic". Which goes to show how muddled much thinking is on this topic. Felsenst (talk) 21:44, 19 December 2009 (UTC)

Long article; split into sections as per summary style[edit]

As evidenced by the size of the article and the number of related articles and references, this is a big subject, hence the main article should be a summary with links to detailed sections. I suggest that the article be broken up into sections, with summary on the main page, as per:

Nbarth (email) (talk) 01:18, 8 March 2008 (UTC)

See above, the entire "Cladistic vs Linnean" stuff really does not belong here. Dysmorodrepanis (talk) 12:50, 11 May 2008 (UTC)
This is a long-standing {{Split}} request. I will look into implementing the splits as they appear to make sense, and to have some support. I am not, however, familiar with this topic, so my efforts will need review and possibly reverting. SilkTork *YES! 12:32, 31 December 2009 (UTC)

Tallapellet's Tree of LIfe[edit]

I cut this out. I know, Tallapellet, this is not going to endear me to you. Nevertheless this is my perception in good conscience.

Tallapellet is a programmer and a WP editor; in fact, he worked on this article. He has his own web site, in which he does not identify himself, unless Tallapellet is his real name. Unfortunately the poster is part of the personal web site. With a click you can get photos of him (my what a handsome man), house plans, blogs, and his personal views of religion. The deciding factor in my mind, however, was the fact that he invites us to buy it online. We can download it for free, he says, if we take it to the copy center and pay for them to print it, or we can purchase it from the web site he specifies. Well, Tallapellet. Wonderful poster, great work of art. If this were not WP I would urge anyone interested to buy it forthwith. My walls don't have room for it. However, this is WP. In my judgement this is a personal and commercial site and is really part of his blogs and other stuff. Tallapellet, you well know this type of material is not allowed on WP. What do do you say, hey, how abut playing by the rules! I don't know what your rank is on WP but if you use it to override me I will object on grounds of abuse of power. By the way I do think it is a worthy and saleable poster and I am glad you contribute to WP.Dave (talk) 15:36, 15 January 2010 (UTC)

The first vertebrate[edit]

"the first vertebrate is the common ancestor of all vertebrates" There seems to be a confusion here between taxa and individuals. There was no first vertebrate. And, there was no first human, no first chordate, etc. This is beginning to sound like Adam and Eve. As dearly as I love the myth, it is only a figure of speech. In considering fossils, no one ever has to deal with the problem of placing the dividing line in the middle of a continuous sequence of fossils, each very slightly different, so as to say, on this side of the line is one species, and on that, another. Fossils are always plucked out of their sequences by chance so as to appear totally discontinuous; in fact exactly what the sequence was can almost never be reconstructed. Have you not been listening to the critics of evolution? Suppose in a population of individuals gradually acquiring more and more distinctive individual variations a certain number cross some defined threshold for meeting the criteria of a new species. Statistically speaking it is most unlikely that the threshold would be crossed by one individual first. But supposing it was. He would not be related to any of the other individuals crossing it subsequently, unless you hypothesize that he and he alone produced this variation that exceeded the threshold. Well, suppose a single mutation occurred instead of the thousands and hundreds of thousands that regularly occur and that one was so advantageous the individual was able to multiply and transmit that mutation. There still wouldn't be any supposed clade of individuals as they would have other ancestors, unless you suppose asexual reproduction. All clades of individuals would have to be polyphyletic.

Cladistics cannot possibly refer to individuals. "Monophyletic" etc refer to phyletic, which are groups of a kind, not to single organisms. Suppose it did. How could you ever find any clades? Human descent is not actually a tree, it is a matrix. There is no such thing as a single ancestor to a group of individuals. Our family trees are cuturally defined, not biologically defined. There may have been a single person of my name, but at that time I must have been equally related to everyone in the population in which he lived. You know, the animals went two by two into the ark, two elephants, two giraffes. Even Adam and Eve were two people. How does that fit statements such as "the first vertebrate?" You mean, the first two vertebrates?

This language is really quite confusing to the public. They are very likely to conceive of the first vertebrate slithering around, perhaps named Charlie. Exactly how the second vertebrate got here would be a bit of a puzzle, or if one did get here at the other end of the world, how did the first and second find each other? So, I am going to alter the language to such an extent it is clear to the public "the first vertebrate" is really "the first taxon of vertebrates". While we are on the subject, let's consider the manuscripts. The method is analogous, not the same. Manuscripts are individual, not classes. Now, if I err, someone be sure and correct me.Dave (talk) 03:33, 16 January 2010 (UTC)

PS. As I edit through this thing I discover chunks that are actually well-written, well-referenced and well-formatted. You will find that I did not or hardly touched those.Dave (talk) 17:32, 17 January 2010 (UTC)
PPS. As I was perusing Albert, I noticed that he defines a semaphoront as a genetic slice of an individual organism. However, he doesn't mean individual qua individual, he means species represented by this instance of it. Of course you can only work on individual animals; the species only lives in them (in Aristotle anyway). If you had a pet named Fido you wouldn't be concerned at all with the species of dog, only with your pet - how tractable, whether healthy or not, and so on. If you analyzed a sample of Fido's genes, you wouldn't care in the slightest about Fido, you would be interested in the sample as an instance of dog. Albert's choice of words is in that sense unfortunate. Organism primarily refers to the individual but he is in no way interested in individuals. As a scientist he can make his meaning clear and use the term in such as way as to avoid confusion but as Wikipedia parrots we cannot. Everything I said about the use of organism here is true. As I clarify this thing I will be making sure there is no confusion. The readers report these articles are very confusing and one of the reasons is the failure to distinguish between multiple meanings or senses of a word. Seeing that the editors think the analysis of MSS is the same thing as the analysis of species it often appears they do not understand the distinction either. The species appears in the individual but it is NOT the individuals together or collectively, it is the universal.Dave (talk) 14:42, 18 January 2010 (UTC)

Complexity of the tree of life[edit]

This purports to present an advantage and disadvantage of the Cladistic method vis a vis the tree of life. But, the reference on the advantage does not mention the advantage and the disadvantage has been cited for a reference. Nor can I find it explicitly on the Internet. However, if we read a little further, we find that the whole topic is not only covered but referenced under the table of advantages and disadvantages! Presumably support of the reference there is implicit. Is there any point in repeating the material? I think not. But, we definitely want to mention the complexity of the tree of life under that title and as far as I can see at that location. That complexity is mainly fractal evolution and Gould's punctuated idea. The reference given is not as good as Albert on fractals so I want to switch to Albert and leave the assessment to later.Dave (talk) 08:09, 19 January 2010 (UTC)

Phony cladogram[edit]

This is not a cladogram. No lineages of distribution are being reconstructed here. The Linux article says nothing at all about it. The source is a programmer's personal site. He has lost his job, he says, and he would like some help with a donation (so would I for that matter). All he does is list releases of Linux software. That's a list not a cladogram. Whether innovated by him on the spot or in use as a vogue word among software people, this is not a true use of pairwise comparison and no graphic cladogram is the outcome. Whew. Sorry, buddy, I sympathize with your economic situation, and with everyone's economic situation, and if I can't sympathize with that than I sympathize with us all for one problem or another, but you still can't use WP to raise money for yourself. Best wishes and best of luck.Dave (talk) 10:21, 21 January 2010 (UTC)

  1. ^ Lundqvist, Andreas; Rodic, Donjan (2009-03-12). "GNU/Linux distro timeline: distribution cladogram". Retrieved 2009-03-19. 

Unused bibliography[edit]

An article such as this could have a bibliography hundreds of items long. In general the topic of interest is the article not the complete bibiography (unless the article topic begins "Bibliography of ...."). So, whatever supports that belongs, whatever does not, does not. I'm removing the unused biblio items to here. Naturally there might be disagreements, in which case put the item you feel should be in there back.Dave (talk) 11:15, 21 January 2010 (UTC)

  • Felsenstein, Joseph (2004). Inferring phylogenies. Sunderland, MA: Sinauer Associates. ISBN 0-87893-177-5. 
  • Hamdi, Hamdi; Nishio, H; Zielinski, R; Dugaiczyk, A (1999). "Origin and phylogenetic distribution of Alu DNA repeats: irreversible events in the evolution of primates". Journal of Molecular Biology. 289 (4): 861–871. PMID 10369767. doi:10.1006/jmbi.1999.2797. 
  • Hennig, Willi (1950). Grundzüge einer Theorie der Phylogenetischen Systematik. Berlin: Deutscher Zentralverlag. .
  • Hennig, Willi (1982). Phylogenetische Systematik (ed. Wolfgang Hennig). Berlin: Blackwell Wissenschaft. ISBN 3-8263-2841-8. 
  • Kitching, Ian J. (1998). Cladistics: Theory and practice of parsimony analysis (2nd ed.). Oxford University Press. ISBN 0-19-850138-2.  Unknown parameter |coauthors= ignored (|author= suggested) (help)
  • Luria, Salvador (1981). A view of life. Menlo Park, CA: Benjamin/Cummings. ISBN 0-8053-6648-2.  Unknown parameter |coauthors= ignored (|author= suggested) (help)
  • Rosen, Donn; Gareth Nelson and Colin Patterson (1979), Foreword provided for Hennig (1979)
  • Shedlock, Andrew M; Okada, N (2000). "SINE insertions: Powerful tools for molecular systematics". Bioessays. 22 (2): 148–160. ISSN 0039-7989. PMID 10655034. doi:10.1002/(SICI)1521-1878(200002)22:2 (inactive 2009-11-14). 
  • Sokal, Robert R. (1975). "Mayr on cladism -- and his critics". Systematic Zoology. 24 (2): 257–262. doi:10.2307/2412766. 
  • Swofford, David L. (1996). "Phylogenetic inference". In Hillis, David M; C. Moritz and B. K. Mable (editors). Molecular Systematics (2. ed.). Sunderland, MA: Sinauer Associates. ISBN 0-87893-282-8.  Unknown parameter |coauthors= ignored (|author= suggested) (help)
  • Wiley, Edward O. (1981). Phylogenetics: The Theory and Practice of Phylogenetic Systematics. New York: Wiley Interscience. ISBN 0-471-05975-7. 
  • Zwickl DJ, Hillis DM (2002). "Increased taxon sampling greatly reduces phylogenetic error". Systematic Biology. 51 (4): 588–598. PMID 12228001. doi:10.1080/10635150290102339. 
I fully support dropping reference to my book. It was not about classification at all (as I was at pains to say in Chapter 10, in a passage that outraged many systematists). Most of it was (as the title implies) about methods for inferring phylogenies. Using nested synapomorphies, in cases where the apomorphic state is known, is Hennig's method -- all of the other numerical methods have only indirect connection to it, and they reconstruct phylogenies, and do not force one to classify monophyletically. So my book is mostly on the more interesting issues that are not what the term "cladistics" covers. Felsenst (talk) 17:50, 25 January 2010 (UTC)

Willi Hennig Society link[edit]

This is a tough call. It does contain some biographical info on Hennig and a blurb for their magazine, but no information useful in figuring out cladistics. It does plug the magazine and it gives meeting and event information with various chatty blog site material. Now, the magazine has its own article accessed through a link at the top of the page and Hennig has his own article also, so we don;t need it for that. I judge this not to be an encyclopedic source for cladistics.Dave (talk) 13:34, 21 January 2010 (UTC)

I agree. I'd drop it the link. Btw... are you planning to submit this article for FAC? –Visionholder (talk) 16:01, 21 January 2010 (UTC)

False cladogram[edit]

The cladogram illustrating polyphyly, monophyly etc is nice looking, but it's inaccurate. Pisces is not a clade. I don't want to remove the picture until there's another one to take it's place, but I don't know how to edit images. --Danger (talk) 16:09, 22 July 2010 (UTC)

I agree. Pisces is not monophyletic, it's not a good inclusion in an article talking about cladistics. Moreover, this picture is maybe difficult to understand. Using a separate picture to illustrate monophyly, paraphyly and polyphyly could make the article clearer. Naldo 911 (talk) 13:45, 22 November 2010 (UTC)

Just incert "lobe-finned fishes" or "lungishes" instead of "pisces" would solve the problem. Alternatively one could use a more hypothetical tree, like the one in the clade article.--Petter Bøckman (talk) 18:22, 22 November 2010 (UTC)

Definitions of types of group[edit]

In the version of 12:47, 4 December 2010, if the text relating to the terms Monophyletic, Paraphyletic and Polyphyletic in the Terminology section is meant to offer definitions, then they are either unclear or incorrect.

  • "Monophyletic groups are groups containing only taxa descended from a given ancestor taxon and all those descendants." This is just poor English, I think. A re-write would be something like "Monophyletic groups are groups containing all the taxa descended from a given ancestor taxon and no other taxa." However, this does not allow a group of extant taxa to be called a monophyletic group: the definition requires the group to contain all the ancestors. This is not how the term is used by many workers.
  • "Paraphyletic groups are groups excluding one or more descendant taxa of the common ancestor, and thus contain some but not all of its descendants." This is only valid if you state that the "descendant taxa" being excluded are monophyletic groups; e.g. excluding a paraphyletic taxon from a monophyletic taxon doesn't leave a paraphyletic taxon.
  • "Polyphyletic groups are groups containing taxa from two or more different monophyletic groups." A monophyletic Linnean Family which contains two or more monophyletic Linnean Genera is, by this definition, polyphyletic.

To be fair, I think it's very difficult to define in words what these terms mean in their current consensus usage (partly because authors aren't, in my view, consistent, and partly because they often don't give their definitions). However, what is written here is very problematic. I'm a bit reluctant to make editing changes without some input from others. There's also the difficulty of finding source references with clear and consistent definitions. Peter coxhead (talk) 22:34, 6 December 2010 (UTC)

Personnaly, I'll definite :

  • monophyletic --> based on a synapomorphy
  • paraphyletic --> based on a plesiomorphy
  • polyphyletic --> based on a homoplasy

N@ldo (talk) 11:28, 7 December 2010 (UTC)

I agree with Peter coxhead. The descriptions are vague and badly worded. This is an important article, and these descriptions should be worded carefully. Petter Bøckman (talk) 18:39, 7 December 2010 (UTC)

I agree too. The definitions based on the notion of common ancestor are rather confusing.

  • Monophyletic : Peter coxhead got the point. We can replace this with :

"All members of a monophyletic group are more closely related between them than they are with any other taxon situated outside the group. A monophyletic group is defined by at least a synapomorphy (= a character at a derived state)." Maybe we can also add that some will call them holophyletic groups.

  • Paraphyletic : "The members of a paraphyletic group are defined by a plesiomorphy (= a character at an ancestral state). As a consequence, they form a truncated group, that excludes some monophyletic taxa from a monophyletic group."
  • Polyphyletic : "The members of a polyphyletic group are defined by a homoplasy (a convergence or a reversion). In consequence, polyphyletic groups are totally artificial"

I took these definitions from diverse lectures and sources, including Darlu & Tassy 1983 (rather old, but still valid for such definitions). N@ldo (talk) 22:45, 7 December 2010 (UTC)

That looks better. I do however think that the concept of an ancestor and all the descendant need to be mentioned, as it is 1) a very common understanding of it, and 2) the definition used by Hennig. My suggestions:
  • A monophyletic group consist of all members that are more closely related to each other than to any taxon outside the group. When extinct ancestors are regarded, a monophyletic group consists of an ancestor an all of it's descendants. Such a group is defined by at least a synapomorphy (= a character at a derived state)." The terms clade and holophyletic bout refer to this understanding of the term monophyletic.
  • A paraphyletic group is defined by plesiomorphies (= characters at an ancestral state). As a consequence, they form a truncated group, that excludes one or more monophyletic taxa from a monophyletic group. An alternative name is evolutionary grade, refering to the ancestral character state within the group.
  • A polyphyletic group is defined by a homoplasy (a convergence or a reversion of characters). In consequence, polyphyletic groups are totally artificial. Petter Bøckman (talk) 08:23, 8 December 2010 (UTC)
Sounds good to me. But I think it's better to not talking about "ancestors" as this notion creates confusion. We need to talk about ancestors but in an other section, where we explain that an ancestor biologically exists, but cannot be reconstructed into a cladistic framework.
It's better to use the term "hypothetical ancestor". And personnaly, I'll define the node as a synapomorphy more than as an ancestor : the tree does NOT show speciations !
Then, let's talk about the examples and the figure, which is awful ("Pisces" is monophyletic, as "Amphibia" !). I propose to draw an other figure, with examples of species as terminal taxa (for example, "salmon" instead of "Pisces" or "Teleostei", which is more understandable for a layman). And let's stop using "Reptilia" as an example of a paraphyletic group, because in some definitions, they are a clade which is more or less synonym with "Sauropsida". We shouldn't add to the confusion !
N@ldo (talk) 10:07, 8 December 2010 (UTC)
The figure absolutely needs changing (I've already removed "Pisces" from diagrams elsewhere -- see Monophyly, but the changes need to go further). Peter coxhead (talk) 12:32, 8 December 2010 (UTC)
Petter Bøckman and N@ldo: I have a problem with your suggestions. For me, they mix what should be two quite different issues. (1) What shape are the three kinds of group, in terms of tree diagrams? This is quite independent of any of the properties of the entities being classified and so you don't need to mention any kind of character issues, plesiomorphy, homoplasy, or whatever. (2) Why would a biologist want to use a group of this shape in a classification? Here characters are crucial.
Sometimes phylogenetic analyses produce trees on which monophyletic classifications are built without it being clear that there are any synapomorphies (other than in the gene sequences). For example, the plant family Plantaginaceae has recently been expanded based on gene-based studies. Botanists have then struggled to find any morphological synapomorphies ([1]). So I don't think that it's right to define monophyletic as requiring any kind of character properties in modern (post-Hennig) use.
The problem with producing definitions which mix (1) and (2) is that it makes it harder to clearly discuss the uses, advantages and disadvantages of particular classifications because these have been embedded in the definitions. Personally, I am in favour of definitions which depend only on shape, followed by an explanation of when and why groups defined in this way are used, and what their advantages and disadvantages are. However, it's clear to me that the literature does mix shape and property in definitions. As an example, consider Budd & Jensen (2000)([2] – not public access, sorry), who, following Jeffries (1979), define a phylum as a crown group with the extant restriction. This leads them to the conclusion that there cannot be an extinct phylum, which is, to say the least, an odd use of the term 'phylum'. The oddness is produced by mixing shape with the extant property. This is only one example among many. The difficulty within Wikipedia is how far to go in trying to produce clear explanations without at least violating WP:Synth, when many (if not most) sources use different/incompatible/confused definitions. I don't know the answer! Peter coxhead (talk) 12:32, 8 December 2010 (UTC)
Hm, seems there's no simple way through this. I do see the point of "mixed" approaches. However, as bout are used, bout should be covered. Would a table be better, where on column may show "tree topography" type of definitions and another one referring to traits? I really think we ought to put "ancestor" in there as well, in that it is central to the Hennig/PhyloCode type definitions. Perhaps a third column?
From a historical point of view, is what we are seeing here the slow splitting up of the terms "monophyletic", "clade" and "crown group" into several closely related meanings? Petter Bøckman (talk) 13:54, 8 December 2010 (UTC)
A table is a clever idea, and may well be the best solution. I think that you are right that there are actually several closely related meanings of these terms. We have to try to sort this out clearly for the 'typical reader' (whoever that is!) of Wikipedia. As always it's much easier to see what's wrong than to put it right. Certainly it's worth trying a two-column table first, to see how it works – I guess in user space rather than publicly at this stage. Peter coxhead (talk) 14:08, 8 December 2010 (UTC)
The table is a good idea : as we already said, it will make it easier to discuss the "ancestor" concept.N@ldo (talk) 15:10, 8 December 2010 (UTC)
Ok, I've attempted a table-based set of definitions at User:Peter_coxhead/Sandbox. Please edit/comment -- I suggest at that page not here. If you think the table works and we can agree on its contents, it can be copied to the article. Peter coxhead (talk) 09:28, 9 December 2010 (UTC)
I'm for replacing the current text! I think you have done a good job! Petter Bøckman (talk) 08:14, 10 December 2010 (UTC)
The revision is good ! However, I believe that our definition of paraphyly and the example both need to be revised. The plesiomorphy quoted here is the backbone which is a synapomorphy of the Vertebrates, but a plesiomorphy when you are inside the Vertebrates. I think it's important to emphasize that apomorphy and plesiomorphy are relative notions, that appear to us when we polarize a tree, using an extra-group (or, more rarely, ontogeny). So, the status of plesiomorphy depends on the ingroup considered.

The example, I think, is confusing. True, the backbone is a plesiomorphy if you are inside the Vertebrates. But no matrix can group birds and mammals as a paraphyletic group if you use backbone as a character at a plesiomorphic state, because there is no apomorphic state for the same character in the group considered. As an example of symplesiomorphy which leads to paraphyletic groups, we can use for example the pluribasal member. Inside the Vertebrates (or at least, the Craniata), it is a plesiomorphic character. The apomorphic state for Sarcopterygii will be "absence of pluribasal member", or "monobasal member". As a consequence, a group defined by the character state "pluribasal member" (Chondrichtyes + Actinopterygii if you don't include fossils) will be paraphyletic, as it excludes Sarcopterygii. Glad that this discussion had made the article better ! N@ldo (talk) 21:58, 11 December 2010 (UTC)

Note that there is a paragraph stressing the relativity of synapomorphy/plesiomorphy. One problem which comes up over and over again in writing this material is its circularity: it's very hard to write a linear explanation, because so many terms depend on the others for a full understanding. So the reader really must work through the whole section. However, I agree that the example isn't the best. However, it must relate to the diagram (which Chondrichthyses + Actinopterygii wouldn't, given the present diagram). As was commented earlier in the discussion, we really need a slightly better example diagram, but hopefully one which doesn't require the whole of the section to be re-written. Is there an example based on the present diagram which is better? Peter coxhead (talk) 23:50, 11 December 2010 (UTC)
Ok, I've tried again, removing the "backbone" example altogether. I think 'cold-bloodedness' works better. Note that we can't say, at this point, that "symplesiomorphy leads to paraphyletic groups" (which I nearly did) because the '-phyletic' terms haven't yet been introduced. Comments on this change, please. Peter coxhead (talk) 00:14, 12 December 2010 (UTC)
~Yeah, "Cold-bloodedness" is better, at least as a provisory example.N@ldo (talk) 10:28, 12 December 2010 (UTC)

James & Pourtless citation[edit]

In a discussion with user Peter coxhead we came to the conclusion that the bird-dinosaur example is not entirely appropriate. If no one objects, I'll remove it. Does anyone know of another case, easily understandable to the general public, in which homology might have been assumed because of a presupposed phylogeny?--MWAK (talk) 14:00, 26 May 2011 (UTC)

The bird-bats link may do? It was Linnaeus who solved that pickle. In Erasmus Darwin's time, the hot blooded creatures (birds-mammals) were generally thought to be related. Another possible candidate may be the "panda thumb" in pandas and red pandas, which both have despite not being very closely related. Another possibility may be falcons and eagles. Petter Bøckman (talk) 14:10, 26 May 2011 (UTC)
Just to make the point that the example needs to serve a very specific purpose, namely to point out that sometimes cladistic analysis can involve a circular argument (but of course doesn't always). A character is excluded because it is thought to be a plesiomorphy or included because it is thought to be an apomorphy. The resulting analysis (manual or computerized) then confirms the assumption. The best example I know (as a "plants person") is the early evolution of land plants. At one time it was thought that the algal ancestors of land plants had more-or-less isomorphic alternation of generations. On this basis, the gametophyte-dominant lifestyle of the bryophytes was an apomorphy, leading to the use of this character in deciding that bryophytes were monophyletic. However, molecular evidence later showed that that this was wrong, so now the plesiomorphic state is considered to be gametophyte dominance, leading to this character playing no part in deciding whether bryophytes are monophyletic. I don't know enough about the examples Petter suggests to know whether they fit this scenario or not. Peter coxhead (talk) 15:38, 26 May 2011 (UTC)
All my suggestions are off then, as they are all more or less pre cladistics. What about the debate over Ida the fossil, where the two sides really debate the character state for the early primates? Petter Bøckman (talk) 16:34, 26 May 2011 (UTC)

Turtles are too controversial[edit]

In view of mounting evidence to the contrary (such as this source), to imply that there is a clade containing birds and lizards but not turtles violates WP:NPOV. The diagram in the Clades section therefore needs revision. I suggest eliminating the Testudines side branch, eliminating the Reptilia and Diapsida legends, and specifying Sauropsida as the node for the Lepidosauria branch. If five side branches is not considered enough, one for the Hyperoartia (lampreys) could be added. Peter M. Brown (talk) 21:04, 24 September 2011 (UTC)

One possibility is to substitute the image shown here, which avoids naming the "traditional reptile groups". The text may needs some changing. Peter coxhead (talk) 09:31, 25 September 2011 (UTC)
Actually, the second paragraph of the section is devoted to emphasizing the tentative nature of cladograms, with an example of a possible result that would force a revision. We could leave the image alone and modify the paragraph so as to present a genuine possibility rather than a speculative one.
In the Monophyly, Paraphyly, and Polyphyly articles, where the image is also used, the caption could be modified to make the point.
My suggestion and Peter coxhead's evade the issue; why not confront it? Peter M. Brown (talk) 17:20, 25 September 2011 (UTC)
Cladogram (family tree)
The root of the reptilian (Amniote) tree is kind of fussy. Presenting the tentativeness of cladograms with the reptiles/turles/sauropsids as an example is a good idea. Peter Coxhead's suggestion to find another "standard" figure for the three -phyly articles would on the other hand also make the articles better in my view. The terminology in the present figure is vague ("reptiles" have multiple meanings, so has "Sauropdids" btw), we really should find another section of the tree where the mapping is a bit less back and forth. Perhaps we should go for a purely hypothetical figure in stead? The figure to the right might serve as an example of the latter. Petter Bøckman (talk) 06:37, 27 September 2011 (UTC)
Hm. Well, apes—on one common use of the term, excluding humans—are paraphyletic. Simians are monophyletic. Monkeys are polyphyletic. Will that do? Is there a better example? Peter M. Brown (talk) 19:25, 27 September 2011 (UTC)
Are monkeys as such polyphyletic? By what definition of monkey? Petter Bøckman (talk) 14:02, 28 September 2011 (UTC)
Old World monkeys and New World monkeys are each clades respectively, but the clade including both (Simiiformes) must also include great apes, which are generally not considered monkeys. So, yes, monkeys in the colloquial sense are polyphyletic. --Danger (talk) 14:20, 28 September 2011 (UTC)
That would be paraphyletic rather than polyphyletic. The monkeys have a single point of origin, so that make them monophyletic in the evolutionary taxonomic sense. A unit made up of night active primates (tarsiers and lorises) wpould be biphyletic. Petter Bøckman (talk) 17:15, 28 September 2011 (UTC)
Ugh, you're right. I don't know why I talk before tea. Danger (talk) 17:19, 28 September 2011 (UTC)
The monkeys are the Platyrrhini together with the Cercopithecidae. The last common ancestor, among the basal Simiiformes, was neither. The monkeys are therefore polyphyletic.
The fact that Petter Bøckman has to ask, though, means that this is not a good example. We're trying to explain what a clade is, and we can't clutter up the explanation with a definition of "monkey."
How about: Boreoeutheria monophyletic, Insectivora polyphyletic, Soricomorpha paraphyletic? Peter M. Brown (talk) 01:15, 29 September 2011 (UTC)
Are you really sure, Peter? This depends on how you define Platyrrhini and Cercopithecidae. If you use a crown group definition for the two you are right, but crown groups are not universally accepted. Using a trait or stem based definition will give you other results. I'd say the mundane understanding of monkey is a monophyletic grade, including their common ancestor and excluding their ape descendants. I would be careful imposing subtle scientific definitions on it, given that "monkey" is very much a vernacular term. The primate tree is useful for our purpose though, as it is well understood and branches fairly straight forward. Insectovores are a bit more fickle, not to mention there are alternative definitions about. Petter Bøckman (talk) 07:06, 29 September 2011 (UTC)
Prowling on Google (in English), I find some bloggers who say that humans are descended from monkeys and, not counting creationists, quite a few who say that they are not. Only the former (who generally seem less well-informed) conform to your assessment of "the mundane understanding." All the same, it would be best if we were neutral on the point.
If we do use the primate tree, though, what group—other than the monkeys—do we use to illustrate polyphyly? Do we just pick a homoplasy within the primates and point out that those with the character constitute a polypheletic group? We could use the absence of a tail, for example.
Alternatively: porcupines are a polyphyletic group, and rodents are monophyletic; is there a paraphyletic group closely associated? Peter M. Brown (talk) 22:03, 29 September 2011 (UTC)
Well, humans are descended from monkeys, if you go back far enough (just like we are ultimately descended from reptiles, amphibians, fish etc). We'll need to be careful when wording things. Prosimians are definitely paraphyletic any way we twist it, perhaps they'll do? For a polyphyletic unit is was thinking about a hypothetical unit for nocturnal primates. Then it dawned on me that primates may be primitively nocturnal, rendering "Nocturnalia" a grade. If so, a unit comprised of the diurnal primates would be nicely polyphyletic. I'll take it up with a primatologist friend up and the university and see what he can tell me. I'm making a primate version of the "standard cladogram" for such use. Petter Bøckman (talk) 19:53, 2 October 2011 (UTC)
Primate cladogram
Edit, here's the raw tree. Petter Bøckman (talk) 09:47, 3 October 2011 (UTC)

────────────────────────────────────────────────────────────────────────────────────────────────────Part of the problem is that terms like "paraphyly" and "polyphyly" aren't used consistently in the literature; they may include or exclude hypothetical ancestors. The commonest uses of these terms in papers that I read are to classify groups of terminal nodes in cladograms produced by either morphological or molecular cladistic analyses. In such cases there is no question of ancestors being included. (This is well-discussed by Podani (2010) – reference here.) So if you use the term "paraphyly" solely in terms of the terminal nodes of this cladogram then traditional monkeys are a paraphyletic group. Whether or not they are paraphyletic if ancestors are included isn't clear to me; you would have to define "monkeys" as all the whole simian clade (i.e. down to its earliest ancestor) minus the whole hominoid clade. I don't think that the traditional term "monkey" would be used in the kind of source which discussed the fossil ancestors. Peter coxhead (talk) 17:06, 4 October 2011 (UTC)

I see your point. Again, there's subtle language phenomena at work here (words have ever so slightly different meanings in various languages). However, Prosimian should be a clearcut grade, no matter how you twist it, and Simian an equally clarecut clade. Petter Bøckman (talk) 18:01, 4 October 2011 (UTC)
Yes, I think it's better to use scientific names, not common ones like "monkey" which don't have clear meanings. Prosimii and Hominoidea are perhaps the best examples in the primates. The definition of the Suborder Prosimii, e.g. here, is a good example of a grade/paraphyly, so isn't accepted in current systems. Whereas everyone agrees that Hominoidea form a clade. Peter coxhead (talk) 18:28, 4 October 2011 (UTC)
They even have straight up vernacular names (lower and higher primates). I'll see what I can do with the cladogram tomorrow. Petter Bøckman (talk) 19:13, 4 October 2011 (UTC)
Yes, good point. I guess my point should be that "secondary names", i.e. ones that scientists and popularizers of science make up for scientifically defined taxa, generally have clear meanings. Whereas "real" vernacular words like "monkey" or "ape" don't have tight biological definitions (although people try to give them ones). So it may be better to avoid the latter. Peter coxhead (talk) 20:09, 4 October 2011 (UTC)

Monophyly, paraphyly and polyphyly
Sorry for being late, Gents! I have now made a primate version of the mono-, para- and polyphyly figure. Petter Bøckman (talk) 11:22, 9 March 2012 (UTC)
With Petter's OK, I have implemented his diagram with minor modifications. (Simia was abandoned by the ICZN in 1929). This is my first venture at uploading, so I welcome improvements. Peter M. Brown (talk) 19:06, 5 July 2012 (UTC)
Now someone needs to modify the section next to this diagram since it refers heavily to the previous diagram. (I know way too little about the topic to try myself ...) Ardik (talk) 12:16, 15 July 2012 (UTC)
I've made some changes to the text next to the diagram. The section Cladistics#Terminology for characters, however, is seriously in need of attention.Peter M. Brown (talk) 17:19, 15 July 2012 (UTC)
Real life events have unfortunately kept me from contributing much lately, but I just want to say you guys have done a very good job of clearing up this article! Petter Bøckman (talk) 06:51, 2 August 2012 (UTC)

Merged Clade[edit]

I have determined that the Clade article and the Clades section of this article contain basically the same information, therefore making it pointless for Clade to have its own article. There is really no way to expand information on the term so as to necessitate a separate main article, since any discussions on clades refers directly to cladistics. This is just a heads up to anyone wondering where the Clade article went. Cadiomals (talk) 03:34, 2 March 2012 (UTC)

The new section on clade in the cladistics aricle lacks a lot of the information found in the old clade article. Though they necessarily will contain much of the same information, the two are not the same. I will revert if no-one comes up with a better reason for a merge. Petter Bøckman (talk) 07:40, 2 March 2012 (UTC)
I support a revert. The concept of a clade is derived from cladistics but logically independent from it. Information on what a clade is belongs at Clade. Information on how clades are used in the field/activity of Cladistics belongs here. If there's unnecessary overlap, the move of content should be the other way. Peter coxhead (talk) 14:50, 2 March 2012 (UTC)
Agreed. Is anyone up for moving the content the other way? Cladistics is a method and a controversial point of view while the concept of a clade is generally acknowledged to be useful, or at least well-defined, by all concerned. A reader who runs into the term in other reading and goes to Wikipedia for enlightenment is not looking either for a method or for a POV. Peter M. Brown (talk) 15:10, 2 March 2012 (UTC)
I've reverted the removal of the content of Clade. I do agree with User:Cadiomals that there is an issue with the overlap of the two articles, but as per User:Petter Bøckman and User:Peter M. Brown, removing Clade was not the solution. I hope that one or other of us will have time to look at these articles soon. Peter coxhead (talk) 15:17, 2 March 2012 (UTC)
My view: The content of the clade article is relevant to several parts of the cladistics article. Eventually, the content will defuse a bit, and anyone looking for a concise text on clade will have to trawl through the (necessarily) quite long article on cladistics. In my view we need both, even though there will be a bit of doubling of information. Petter Bøckman (talk) 19:56, 2 March 2012 (UTC)
Whatever; I don't ultimately care. Cadiomals (talk) 21:30, 2 March 2012 (UTC)

Linnaean definition of tetrapod[edit]

According to Cladistics#Example, Linnaean taxonomy defines the Tetrapoda "morphologically as vertebrates with four limbs (as well as animals with four-limbed ancestors, such as snakes)." The definition is not morphological if it includes the part in parentheses, which brings in the matter of ancestry. Without the parentheses, however, it is inadequate because it excludes snakes, moas, etc. Peter M. Brown (talk) 17:00, 6 August 2012 (UTC)

Well, it's how the term is used in Linnaean taxonomy. We just need a suitable way to express it. Petter Bøckman (talk) 20:23, 10 August 2012 (UTC)
I would like to replace the first paragraph of Cladistics#Example with something like the following.
The difference between approaches can be illustrated by the varying treatments of Tetrapoda.
Linnaean taxonomists do not have a firm definition of Tetrapoda or, indeed, of any higher-level taxon. Systematists like Benton see, as an important strength of the approach, the ability of the taxonomy to change based on the progress of science. The ICZN takes this flexibility as an underlying principle: "Nomenclature does not determine the inclusiveness or exclusiveness of any taxon".
Historically, the classes Mammalia, Aves, Reptilia, and Amphibia were united in the superclass Tetrapoda based the evident fact that most animals in these classes have four limbs. All animals in these classes were therefore included despite the fact that some, including whales, moas, snakes, and caecilians, have lacked this feature. In addition, four-limbed animals like the Diadectidae known only from fossils were included even when they do not fall into the four extant classes. Going forward, the discovery of a new animal may require the scientific community to decide whether to call it a tetrapod or not; when the transitional fossil Tiktaalik was first described in 2006, for example, the discoverers called it a fish, not a tetrapod, but no firm rules required this designation.
Is my history accurate? Can it be sourced? If not, can something along the same line be used? Peter M. Brown (talk) 00:44, 11 August 2012 (UTC)
No, it seems that all early sarcopterygians were classified as fish, amphibians, or reptiles; no tetrapod was considered neither amniote nor amphibian. I'm working on it. Help would be appreciated. Peter M. Brown (talk) 14:16, 11 August 2012 (UTC)
OK, done. Please review and correct, or at least comment on the shortcomings. Peter M. Brown (talk) 23:01, 11 August 2012 (UTC)
Looks good, only there is actually a "firm definition" of Tetrapoda, i.e. having feet rather than fins. One might say that the back and forth over internal gills and fin rays (both found in Ichthyostega) would make the definition "unfirm", but the fins/feet definition has there all along. The same with mammals, where the definition eventually settled on the jaw joint/ear osicles.Petter Bøckman (talk) 20:19, 12 August 2012 (UTC)
Really? On the matter of defining Tetrapoda, Clack & Coates (1995) wrote, "Possession of digits remains the most compelling character in practice at the present, but we point out its potential inadequacies for the future." I might add that as a reference supporting my contention that delimiting Tetrapoda in a Linnaean framework is an ongoing process; that there is not a firm definition. Notice that they do not mention feet. Clack adds elsewhere that "limbs with digits" is unsatisfactory since it can be broken down into many separate characters. Peter M. Brown (talk) 21:12, 12 August 2012 (UTC)
I suppose it depends on the abruptness of the change. If the single-mutation scenario is right, then all, or most of, Clacks characters are linked. If it was a more gradual process, then more fossils will lead to new definitions. In that regard I believe the term "unfirm" is apt. Petter Bøckman (talk) 12:19, 13 August 2012 (UTC)
Do not confuse definition and diagnosis. Diagnoses – lists of characters that allow systematists to tell whether a specimen belongs to the taxon in question or not – belong to taxonomy, not to nomenclature; the ICZN (like its sister codes) stipulates that, when a taxon name is erected, a diagnosis must be provided, but this diagnosis can be changed at any time; the diagnosis need not ever be correct, and a change of the diagnosis is not a nomenclatural act.
Taxon names in the species, genus and family groups of ranks do have definitions under the ICZN; but those consist of a type and a rank. Homo sapiens is defined as the name of the taxon of species rank that Linnaeus belongs to, Homo is defined as the name of the taxon at genus rank that H. sapiens belongs to, and so on, till Hominoidea is defined as the name of the taxon at superfamily rank that Hominidae belongs to. And then it ends. Tetrapoda is high above the family group of ranks; on the rare occasions that it's given a rank at all, that is usually superclass or an unconventional rank.
The fun thing is that most rules of the ICZN only hold for the species, genus and family groups of ranks. Even the Principle of Priority does not hold for names of higher rank. They don't have types either. Therefore, Tetrapoda does not have a definition at all under "Linnaean" nomenclature.
There is a discussion about how best to define the name Tetrapoda = which clade to attach it to; but this discussion is entirely within phylogenetic nomenclature. As you say, some favor applying the name to a crown-group, which means they want a node-based definition, while others favor an apomorphy-based definition that lets Ichthyostega stay a tetrapod.
David Marjanović (talk) 13:27, 23 September 2012 (UTC)
A diagnosis is nomenclatural, not just taxonomic, as it is required as part of naming at the species level. If taxa were defined by necessary and sufficient conditions, it would indeed be an objection that the diagnosis may change and may not actually be applicable to the type. A diagnosis does not supply necessary conditions for membership, however; rather, it is a statement intended to provide characters that differentiate the taxon from others with which it is likely to be confused. It provides guidance to subsequent researchers and may be replaced by another formulation that gives better guidance. Peter M. Brown (talk) 16:57, 6 October 2012 (UTC)

ostriches and rheas[edit]

Hi Peter Brown - you asked if you were missing something in removing the sentence "The fact that ostriches and rheas both have four limbs does not provide any support for putting them into a separate group of 'flightless birds'." Perhaps. Isn't it an important concept that while ostriches and rheas share this character, it provides no evidence of their being closely related, because the trait is plesiomorphic for the larger clade containing those species - tetrapoda?

Reflecting on your response, I now think that the problem with the sentence is not so much that it is irrelevant as that it is false. Restricting the animals of interest to those with four limbs excludes most flying animals (insects), thereby providing some evidence that they are flightless; it also eliminates the most common animals of all, the nematodes, so every non-nematode taxon—including birds—becomes more likely in light of this evidence. Your formulation also seems incorrect: the fact that ostriches and rheas all possess four limbs places both groups among a tiny percentage of eukaryote genera, thus providing excellent evidence that they are closely related. Peter M. Brown (talk) 16:28, 10 August 2012 (UTC)
The context is within the tetrapoda. Yes, four limbs in rheas and ostriches is evidence that they are related, but only at the level of tetrapoda, and at no finer of a level. Having four limb is synapomorphic for the tetrapods, but plesiomorphic if you consider just rheas and ostriches. Four limbs provides no evidence for a closer relationship, even though there are tetrapods with other than four limbs (snakes, skinks, caelicians etc). de Bivort 17:18, 10 August 2012 (UTC)
Does this observation point the way to reformulating the deleted sentence in a way that is true, relevant, and not redundant? The text already says that "for a group within the tetrapods, such as birds, having four limbs is a plesiomorphy." I do not see a point in replacing the deleted sentence with one noting that this trait is a plesiomorphy for the ratites. Peter M. Brown (talk) 17:42, 10 August 2012 (UTC)
Well, clearly the point of the deleted sentence was to provide an example, to make the concept easier to understand for readers. And also clearly it didn't succeed, since it generated confusion. I think an example, either the original one re-worked, or a new one, is a good idea. Hopefully others will weigh in. de Bivort 18:11, 10 August 2012 (UTC)

Excessive commas[edit]

I am removing some commas recently added to the article. All are permitted by the rules, but many are optional and interfere with the flow of the text. For example, the edit added a comma after each occurrence of "e.g."; while this is always permitted, it is generally optional. The Blue Book of Grammar and Punctuation, for example, uses

You may be required to bring many items, e.g. sleeping bags, pans, and warm clothing.

as an example of acceptable usage; a comma is admissible after "e.g.", but it is not necessary. Peter M. Brown (talk) 12:37, 25 August 2012 (UTC)

Edit 23 September 2012[edit]

Argh. This whole series of articles was, and largely still is, a mess. So many articles on so narrow topics, all of them repeating large chunks of the same stuff (independently written, not copied & pasted!) that only fits one of them... Cladogram explains phylogenetic analysis at considerable length, the very topic of the Phylogenetics article, when a cladogram is simply the result of a phylogenetic analysis...

I have tried to begin unraveling this mess. Yesterday, I extricated Phylogenetic nomenclature from it; that was fairly easy, because I had created that article independently of the mess. I also restricted Clade to what it says on the tin; that article was relatively short.

Today, I:

  • Wrote a new introduction, trying to distinguish phylogenetics from the principle of only naming clades.
  • Reorganized the article according to that distinction. This distinction allows me to keep the Cladistics and Phylogenetics articles separate (I have elsewhere threatened to merge them).
  • Made the paragraph about the problem of homology a "controversy" subsection within the "phylogenetics" section. The rest of that paragraph is now in the "controversy" subsection of the "taxonomy" section, see below.
  • Moved the history section up, because it concerns both phylogenetics and taxonomy, and rewrote the part on phenetics.
  • Exported the explanations of "nested", "sister group" and "basal" to Clade, because they belong neither quite to phylogenetics nor to taxonomy.
  • Changed "Terminology for characters" to "Terminology for character states".
  • Changed "Terminology for groups" to "Terminology for taxa". Apparently, it was meant to compare pattern cladistics to (process) cladistics, but it never once mentioned those terms, and pattern cladistics is dead, dead, dead. (It doesn't even have a WP article, unlike, say, the quinarian system!) I've still kept the subsection, because it is a nice conclusion to the rest of the taxonomy section, but I've removed the pattern-cladistic language with its "groups of nodes". If you want to bring it back, write an article on pattern cladistics. – Also, the stipulation that an apomorphy must be inherited by all members of a clade was flat-out wrong.
  • Moved "Phylogenetic definitions of a clade" to Phylogenetic nomenclature, where it belongs. Once more, in chorus: Clades are not defined, they exist outside our skulls and are discovered by science; names are defined, they are conventions. Science and nomenclature aren't the same!
  • Integrated part of the "Cladograms" section, which was in part a repetition of stuff higher up on the page, in the new "Phylogenetics" section. – "In cladograms, all species lie at the leaves" is flat-out false; datasets for cladistic analysis can have larger taxa than species as their OTUs, as well as smaller ones down to individuals (or genes) within one or a few species.
  • Deleted the other part of "Cladograms", because it was just repetition of what it says in Phylogenetics and Phylogenetic tree.
  • Removed the stuff about phylogenetic nomenclature. PN is not synonymous with the principle of only naming clades – it's not even a subset of it. The rest of the "cladistics in taxonomy" section is fine; I've made it a "controversy" subsection within the "taxonomy" section. This subsection also includes text from higher up; I've removed the reference to Mayr (1974) writing of "the essential unity of the reptiles", though, because almost 40 years later that's just silly – the fossil record has improved to such a point that the long, long transition from "reptiles" to birds is now known in itty-gritty detail, never mind the one from "reptiles" to mammals which was already pretty well understood in 1974. If you want to have a gradist statement, find a much more recent reference (Mayr himself only died in 2010, right?).
  • Edited that subsection. As somebody said higher up on this page, Hennig's work didn't trigger any discussion about phylogenetic nomenclature; the first paper that used PN came out in 1986!
  • Removed the "complexity of the tree of life" section because it's off topic. Calling the tree "the cladistic tree of life" doesn't make it cladistic.
  • Changed "Application to other disciplines" to "Application to disciplines other than biology". Removed the simply wrong statements about pairwise comparison; phylogenetic analysis compares the entire dataset at once.
  • Clarified the small distinction between cladistics and the comparative method of historical linguistics.
  • Moved the ethology example high up, because ethology is part of biology – ethological data can be, and have been, used in the reconstruction of the relationships of animals.
  • Removed "Bauplan", "Evolution of Mollusca" and "PhyloCode" from the "See also" section.
  • Cut down the enormous bibliography to the references and a few other important works... which should probably be cited somewhere in the article or moved to other articles.

The article as it now is does not have enough references. I'll try to add more, but I won't have much time in the coming weeks. Feel free to scoop me!

Now on to Phylogenetics... but I won't do much there, it's late.

David Marjanović (talk) 18:19, 23 September 2012 (UTC)

Hi. I rewrote the subsection "Some criticisms of phylogenetic nomenclature" last week, spending some time researching criticisms. No question, it didn't belong in the Cladistics article. I don't see it in the new Phylogenetic nomenclature article either, though. Where are you planning to put it? Or will you be rerewriting it? Peter M. Brown (talk) 19:43, 23 September 2012 (UTC)

Opening definitions[edit]

The opening definitions are very confusing and appear to be confusing/confounding cladistics with phylogenetics, among other things.

Definition #1- Cladistics is not "the science of phylogenetics." This sentence has two problems. First, cladistics is a starting point method for figuring out a phylogeny. It is not the actual making a phylogeny, several of which could come from one cladistic analysis (see Cladograms and Phylogenetic Trees, Wiley EO, Systematic Zoology, 1979). Second, and in a similar vein, it is not "the" method as in the only one for figuring out a phylogeny (again, it doesn't even figure them out in the first place it is only a starting point). Third, and MOST importantly, this definition really doesn't tell the reader anything about cladistics. It just tells the reader that it is "a method", but never actually describes the method (in fact no where in the article is this so called "method of cladistics" ever defined or described, not even telling the reading what a shared derived characteristic is). Therefore I wanted to replace the opening definition with: "A method for establishing the relationships of various individuals, items, concepts, etc through the use of shared characteristics." This definition is backed up with the following citations:

"Cladistics: Theory and Practice of Parsimony Analysis" page 1: "Cladistics is a method of classification that groups taxa hierarchically into sets and subsets." page 2: "The aim of cladistics is to establish sister-group relationships, and the concept of two taxa being more closely related than either is to a third." "Sister groups are hypothesized through the analysis of characters"

"Transformed Cladistics, Taxonomy and Evolution" page 68: "On the basis of a comparison of organisms, perceived similarities are utilized to choose those characters that will be used for more detailed comparisons leading to the construction of a cladogram. Central to this endeavor is the recognition of shared derived characters."

UCMP University Of California Museum of Paleontology Website on cladistics: "Cladistics is a particular method of hypothesizing relationships among organisms." "The basic idea behind cladistics is that members of a group share a common evolutionary history, and are "closely related," more so to members of the same group than to other organisms. These groups are recognized by sharing unique features which were not present in distant ancestors."

Definition #2- This is confusing due to its wording. Its original form leaves it unclear if the words "phylogenetic systematics" are included because they are somehow a synonym for cladistics (if so why is that down here in definitions) or whether it is a whole different field and hence no longer a definition of the main heading word of the page. I tried to emphasize that the second definition is a narrower form of the first definition and that it goes by a specific name (phylogenetic systematics). Hence my definition: "phylogenetic systematics, is a narrower and more formal version of the above which uses the principle of only naming clades as its source of information for taxonomy. This is sometimes pejoratively called "cladism."" — Preceding unsigned comment added by Nathank2 (talkcontribs) 21:21, 29 October 2012 (UTC)

Certainly, your critiques are well founded. Your quotations, however, don't actually provide a definition and cannot be used in support of one. None of the three actually defines cladistics; the first and the third say that it is a method with certain characteristics but don't say just which such method should be called "cladistics" while the second doesn't use the term.
I see no support for the statement that the term "cladism" is perjorative. Donoghue and Cantino have no problem with it, as they have produced an article entitled "Paraphyly, ancestors, and the goals of taxonomy: A botanical defense of cladism".
You say that cladistics does not establish a phylogeny but does establish relationships between individuals. What is there to phylogeny besides relationships? Again, you object to the characterization of cladistics as a method and your cure is seemingly to characterize it as a method. . . .what am I missing?
There are indeed two senses of "cladistics". The first is a widely accepted research technique. The second is cladism, an approach to nomenclature. Neither is a "version" of the other.
Revision is definitely needed. It is well that you have started this discussion.
Peter Brown (talk) 22:52, 29 October 2012 (UTC)
I didn't add in the cladism part. It was originally introduced by Mayr and Darlington in the 1960's and 70's due to their perception of unfounded philosophical belief and borderline zealotry by followers of Hennig's newly created field of cladistics. But it's likely a long since forgotten concern.
Phylogeny is not just relationships between individuals (aka who is the most similar). Phylogenies shows the ancestor-descendant relationships that result from the various events of evolution. Phylogenies take into consideration hypotheses of evolutionary events like budding and anagenesis which a cladogram (the result of cladistics) cannot account for or show. Cladograms also don't show who actually evolved from who on a species/population level. For example the famous horse evolution phylogeny [picture] is not a cladogram, but it is a phylogeny. Cladograms don't have real entities in the internal nodes only at the leaves, but phylogenies can have real entities at the internal nodes. Hence different evolutionary theories about ancestor-descendent relationships will result in different phylogenetic trees from a single cladogram. (Again, see "Cladograms and Phylogenetic Trees, Wiley EO, Systematic Zoology, 1979" page 1, 1st paragraph.)
I see your point about simply restating that it is a method. It is a method, but the original definition is incomplete. The new definition explains that the method is to compare shared characteristics for the purpose of figuring out relationships. I wanted to put shared derived characteristics and then add a page link, but haven't gotten around to writing up a whole thing on shared-derived characteristics yet :)
Why does definition #1 not work? "The aim of cladistics is to establish sister-group relationships, and the concept of two taxa being more closely related than either is to a third." "Sister groups are hypothesized through the analysis of characters" I haven't been here long so I don't know but does wikipedia require only direct quotes as definitions? Can't we reword that mess into something a layman could understand? That's basically what I tried to do.
How about a direct quote from the OED? Cladistics (OED 2nd Edition)- "Systematic classification of groups of organisms on the basis of shared characteristics thought to derive from a common ancestor." It's just too narrow because it only includes organisms.
"Phylogenetic systematics" was Hennig's original term for his methodology which is now referred to as cladistics. His system emphasized "clades" to the exclusion of other grouping. "Phylogenetic nomenclature" is the new nomenclature based on cladistics, it has it's own page already."

Nathank2 (talk) 00:06, 30 October 2012 (UTC)

Looking back over the old revisions of this article, things seem to have been clearer before. The very old version from [09:22, 6 June 2003] is almost exactly what I am suggesting. It's clear and understandable for a lay person. Version [20:23, 19 January 2007] is a little more convoluted and difficult to understand at the beginning, but the entirety of the opening section is very good and does a good job of giving a basic explanation of the whole concept. Nathank2 (talk) 16:14, 30 October 2012 (UTC)

The 2003 version uses the phrase "derived similarity", which I don't think the lay person will find at all clear and understandable. The 2007 version, however, is admirably clear. It does use the term "organisms", which you object to; perhaps you would not mind "items (usually organisms)"?
What it doesn't do is distinguish the two senses of "cladistics". I would urge that the article be limited to cladistics1. As I understand cladistics2, it is simply cladistics1 as practiced by those favoring phylogenetic nomenclature. Rather than give cladistics2 any significant attention, let's just refer the reader to the Phylogenetic nomenclature article.
By the way, you misunderstood my use of the term "relationships". I didn't mean just "who is most similar" but included ancestral relationships. I suggest that the use of "relationships" in your proposed Definition #1 would be misunderstood in exactly the same way.
Peter Brown (talk) 19:22, 30 October 2012 (UTC)
Yeah, the "items" idea sounds good. I'm still not fond of the 2007 definition. It just seems jargony ("morphological similarity", "branching", "evolutionary tree")? However, if you like that definition, could we perhaps add something at the end which mentions that the process is carried out through the construction of a cladogram? The definition never mentions characteristics or that they are shared and derived or that they are even used in cladistics, but a link to cladograms could (if also cleaned-up) or a further explanation below as in the 2007 version.
"Cladistics is a philosophy of classification that arranges items only by their order of branching in an evolutionary tree and not by their morphological similarity [1]. This is carried out through the construction of a Cladogram."
  1. ^ Luria, S. E., S. J. Gould, and S. Singer. 1981. A View of Life
I totally agree that we can just dump def 2. It has really been replaced by the term cladistics in the general vernacular (It was just the original/official word for what we now call cladistics). A redirect from "Phylogenetic systematics" to here would be appropriate. Phylogenetic nomenclature is not however Phylogenetic systematics. Phylogenetic nomenclature came out in the mid 80's, Phylogenetic systematics was the title of Hennig's book on cladistics in 1966 (it was his official name for what is now called cladistics).
Not sure what you meant by the "relationships" comment, but it looks like that will be resolved with the current definition.

Nathank2 (talk) 21:11, 30 October 2012 (UTC)

You're right, characteristics are essential to the concept and must be mentioned in the lead. I reiterate that talk of derived traits can be most confusing. A trait can only be derived with reference to a clade (feathers are a derived trait within Archosauria but not within Aves) but the term is often used in a way that fails to make clear the relativity of the attribute. Peter Brown (talk) 21:47, 30 October 2012 (UTC)
completely agree. the "derived" concept can be really confusing and probably too much for an opening definition. Nathank2 (talk) 22:02, 30 October 2012 (UTC)
After reviewing still more versions of this article it kinda feels like we are just recreating the wheel and rehashing out the same stuff from the last 10 years :( The massive edit on "18:19, 23 September 2012" with no sources and no community discussion made many changes that were very good, but also introduced a lot of changes that make the article very confusing and inaccurate (as we have discussed here the last few days). Our current discussion seems to be leading us back to the exact same article opening paragraphs that were already there prior to that massive edit. Nathank2 (talk) 15:37, 31 October 2012 (UTC)


Hello, I need a help from people with knowledge in systematics and taxonomy to discuss the Template:SysTax. This template was removed from the pages where it appears for discussion and improvment. Thanks Zorahia (talk) 15:22, 22 March 2013 (UTC)

Reference Errors[edit]

RE: Maas, Philipp (2010), "Text Genealogy, Textual Criticism and Editorial Technique", in Jürgen, Hanneder; Maas, Philipp, Wiener Zeitschrift für die Kunde Südasiens (Vienna: Austrian Academy of Sciences), 52-53: 63–120, doi:10.1553/wzks2009-2010s63

The article title is wrong. At pages 63-120 of this volume, Maas's article title is "Computer Aided Stemmatics — The Case of Fifty-Two Text Versions of Carakasaṃhitā Vimānasthāna 8.67-157."

The title given to this issue of the journal, which Maas CO-edited, is indeed Text Genealogy, Textual Criticism and Editorial Technique. But in the references, that title for the journal is given as the title for the article. It is unclear what is being cited. — Preceding unsigned comment added by (talk) 05:15, 19 May 2013 (UTC)

I have access to the article by Maas, and that seems to be what is being cited. It develops "A hypothetical stemma codicum, i.e. a branching diagram that reflects the transmission history of a given text" using "two complementary approaches: the computer-based cladistic analysis of variant readings (i.e. a quantitative approach) and the philological discussion of selected variants (i.e. a qualitative approach)." MacClade software turned out to be useful. That fits with reference 6 (I wish people wouldn't split bibliographies into two parts like this, as seems to be becoming more common in wikipedia). Sminthopsis84 (talk) 15:41, 19 May 2013 (UTC)

Neutrality issue[edit]

Peter Brown deleted my edits to the cladistic pages here, suggesting that they gave too much credit to Steve Farris. I see that he is not a systematist, and I wonder just how he justifies his editorial choices, given that the section called for greater detail, which I provided. Is he suggesting that the historical information I provided is not true? - Andy Brower — Preceding unsigned comment added by Abrower (talkcontribs) 13:00, 20 May 2013 (UTC)

Yes, I think that's the general idea. Not that Steve Farris didn't write what you said he did, but that his personal approach was somewhat divisive, and the legacy that you claimed for him is largely due to other people. Sminthopsis84 (talk) 17:11, 20 May 2013 (UTC)
I support the deletion, too. The issue is maintaining WP:NPOV and WP:UNDUE. You can say a bit about what Farris wrote, but his importance must be based on what reliable third party sources say. Peter coxhead (talk) 19:12, 20 May 2013 (UTC)

Etymology: minor edit-war[edit]

See here. A similar example, which thankfully hasn't escalated to an edit-war due to the timely intervention-edit of another user, can be found here. From my point of view, this is something trivial... Thanatos|talk 19:59, 24 May 2013 (UTC)

Stray citation and WP:3RR[edit]

The first sentence of the page currently has two citations for the parenthetical statement that the word "cladistics" derives from Greek κλάδος, klados, i.e. "branch". I see the first one as spurious ("cladistics". Online Etymology Dictionary. ) because it says nothing about this, but the editor who inserted it has repeatedly reverted me. Could someone else please look at this matter? Thanks. Sminthopsis84 (talk) 20:01, 24 May 2013 (UTC)

  • A. See section above.
  • B. WP:3RR reads "more than 3 reverts" (emphasis mine); so please count before invoking it; I haven't in fact even reached 3 reverts yet; I'm presently at 2...
  • C. If you object so much (can't understand why) to this reference-entry-link, why haven't you instead e.g. proposed citing, linking instead to the hyperlinked one therein, i.e. cladism, or even linked it yourself without first proposing anything?? ;-)
Thanx. Thanatos|talk 20:08, 24 May 2013 (UTC)
Maybe we were seeing different things but when I followed the original first reference (to "cladistics" in the Online Etymological Dictionary) I got nothing (I think because there is only an entry for "cladistic"). The link I've used now shows – for me anyway! – both "clade" and the derived terms "cladistic" and "cladism". In my opinion we don't need a reference to a dictionary of classical Greek to permit the transcription between κλάδος and klados; transcriptions between writing systems are regularly given in Wikipedia without sources. So I think this is the only reference needed. Peter coxhead (talk) 19:21, 25 May 2013 (UTC)
Thank you Peter, that also looks good when I access the link. Sminthopsis84 (talk) 13:29, 28 May 2013 (UTC)


What is the difference (if any) between cladistics and classic biological taxonomy? How do they relate to one another? What was there before cladistics was invented? Or is cladistics just a new name for an old discipline? The article on "Taxonomy (biology)" mentions "cladistic" only in passing, without any explanation of its role. — Preceding unsigned comment added by (talk) 07:49, 29 July 2013 (UTC)

I would say that cladistics is just one method which can be used as part of taxonomy to derive hypotheses about the evolutionary relationships of taxa and hence provide evidence on which classifications are based. However, the word "cladistics" seems to be used very loosely in modern biology (e.g. cladograms produced by computer programs employing parsimony or maximum likelihood criteria are sometimes considered part of "cladistics" whereas their methodology is very different). What was there immediately before (but after Darwin) were judgements made by specialists in that group, based on characters and attempting to reconstruct evolutionary relationships, but in a less "algorithmic" and hence arguably more "subjective" fashion. Now, at least for living organisms, there is a related but different method based largely on patterns in their genes. Peter coxhead (talk) 09:26, 2 August 2013 (UTC)

Character-based clade definition[edit]

In Cladistics#Terminology for taxa is the statement:

A clade is characterized by one or more apomorphies: derived character states inferred to have been present in the first member of the taxon, inherited by its descendants (unless secondarily lost), and not inherited by any other taxa.

Is saying that the character states were "inherited by its descendants (unless secondarily lost)" just a fancy way of saying that the states were inherited by its descendants unless they weren't? If so, the phrase is vacuous and could be omitted. Peter Brown (talk) 02:12, 2 August 2013 (UTC)

I think your interpretation of the phrase is correct, but I don't think it's vacuous. Firstly, if you omit the entire phrase, you get A clade is characterized by one or more apomorphies: derived character states inferred to have been present in the first member of the taxon and not inherited by any other taxa. This isn't right, because it doesn't say that the apomorphies were inherited by that taxon, only that they weren't inherited by any other. If you omit the parenthesised part, you get A clade is characterized by one or more apomorphies: derived character states inferred to have been present in the first member of the taxon, inherited by its descendants, and not inherited by any other taxa. Since bare plurals in English often logically mean "all", this can be misread as ... inherited by all its descendants ..., so the qualification is needed to avoid this interpretation. A fuller statement might be something like A clade is characterized by one or more apomorphies: derived character states inferred to have been present in the first member of the taxon, present through inheritance in most of its descendants (but not necessarily all since they may have lost it), and not inherited by any other taxa. Peter coxhead (talk) 09:14, 2 August 2013 (UTC)
A clade is characterized by one or more apomorphies: derived character states inferred to have been present in the first member of the taxon, present through inheritance in most of its descendants (but not necessarily all since they may have lost it), and not inherited by any other taxa.
This "fuller statement" of yours introduces most as a brand new element. ". . .most of its descendants"— are we counting individuals, species, families or something else? Of course, imprecision in a definition is acceptable so long as we are willing for it to be inherited by the defined term, "clade" in this case. Is imprecision in the concept of a clade tolerable? Notice that whether a character state satisfies the condition will vary through time; ovipary was relevant as an apomorphy (sensu Coxhead) for Synapsida as of the end of the middle Jurassic, but not today.
"Vacuous" was admittedly the wrong word. To say that a character state is either inherited or not inherited is, rather, a tautology, a proposition that cannot consistently be denied. When a tautology appears as part of a conjunction, it can simply be dropped. "I am American and all Kurds are Kurds and you are English" is equivalent to "I am American and you are English". If the result of dropping a tautology in a conjunctive context is unacceptable, then so is the original proposition. We agree that the result of dropping "inherited by its descendants (unless secondarily lost)" from the definition in the article is inadequate; it follows, I submit, that the definition is unsatisfactory as it stands. Peter Brown (talk) 18:16, 2 August 2013 (UTC)
Ah, right, I see you were making a slightly different point than I thought. First, a character state might not be (simply) inheritable; e.g. as I know from experience, many alpines have consistent growth habits in the wild which don't persist in lowland gardens. The genetic propensity to grow in a certain way when exposed to high levels of UV, snow cover, etc. may be inherited, but not the observed characters. Second, there is the word "secondarily" in the original. To say that a character is either inherited or secondarily lost isn't a tautology: "secondarily lost" doesn't just mean "not inherited".
However, I do agree that the definition you quote is unsatisfactory, because it appears to be circular. I think that what "secondarily lost" really means is "this is a clade, so the fact that some members don't have the character means that they lost it secondarily". Thus apomorphies of a clade not present in a member of the clade must be "secondarily" lost as opposed to the "primarily" acquired apomorphies. Can "primary" and "secondary" in this context be defined independent of "clade"?
Also, are the character-based 'definitions' really definitions in the context of this article? The 'real' definitions of a clade are the "all descendants" ones. In Hennig's method, characters are a way of coming up with hypotheses as to what are clades. Character-based definitions seem to be a feature of the PhyloCode, not cladistics more generally. Peter coxhead (talk) 08:05, 3 August 2013 (UTC)
The propensity to grow in a certain way when exposed to high levels of UV, snow cover, etc. is not an observed character but it is one that you have observed???
Yes, bad example on my part. Although it does make the point that "character" requires very careful definition: a "propensity" is a character, but might not obviously be thought of as one. Peter coxhead (talk) 17:07, 3 August 2013 (UTC)
The only definition I quote is the one you offered; I would say that both it and the one in the article are circular for the same reason. However, suppose we adopt the the "all descendants" definition of "clade" so that it doesn't depend on "secondarily lost". Circularity disappears, but does this permit a principled way of saying whether oviparity was secondarily lost in the mammaliaform lineage leading to Theria or whether it simply disappeared? I doubt it.
The article's character-based definition of paraphyly does not use the term "secondarily" but, perhaps for that reason, it's in even worse shape.
A paraphyletic assemblage is characterized by one or more plesiomorphies: character states inferred to have been inherited from ancestors but not present in all of their descendants.
I take Monotremata to be monophyletic, though there are few careful studies of the matter. The combination of hair and oviparity is a good candidate for a characterizing feature. It was inherited from the last common ancestor of the crown mammals. But some descendants of that ancestor lack oviparity; is the characteristic then a plesiomorphy, so that the monotremes are paraphyletic as well as being monophyletic? No, the definition of plesiomorphy is unsatisfactory.
I do not conclude that character-based definitions are all unsatisfactory, for I think that the character-based definition of polyphyly, as I have now updated it, works. It is essentially the definition given in the articles Monophyly and Polyphyly.
Monophyly and paraphyly, in my view, are tree-based concepts and should have tree-based definitions. Polyphyly is a character-based concept and should have a character-based definition. The attempt to approach all three in a uniform manner is a mistake. Further, "was secondarily lost" is just a misleading way of saying "was not inherited".
Peter Brown (talk) 16:28, 3 August 2013 (UTC)
Two different issues here:
  1. was secondarily lost" is just a misleading way of saying "was not inherited – well, this is at the heart of our disagreement. Clearly "was lost" is just a way of saying "was not inherited". However, the addition of "secondarily" here is an attempt to say more, although I suspect it fails through circularity. (But this is precisely one of the accusations that critics made against cladistics: the decision that a character is an apomorphy is informed by a prior view that the group it defines is a clade.) You and I may agree that adding "secondarily" fails to say anything extra, but I think that reliable sources do not, so we should be cautious about removing it.
  2. I profoundly disagree that monophyly and paraphyly are tree-based concepts and polyphyly is character-based. First, the literature doesn't support this view. Hennig's definitions were all essentially character-based. (As elegantly summarized by Oosterbroek, monophyly = synapomorphous similarity; paraphyly = symplesiomorphous similarity; polyphyly = convergent similarity.) Later, others produced tree-based ones. I can source both kinds of definition for all three concepts, if you want. Second, it's very easy to give a tree-based definition of polyphyly: any group of taxa that is not either a monophyly (tree defined) or a paraphyly (tree defined) is a polyphyly (tree defined). Only Oosterbroek (1987) has managed to give a positive rather than a negative tree definition of paraphyly which can be shown to precisely match the negative one ("[a] group of taxa is polyphyletic if their most recent common ancestor has given rise to excluded taxa of which at least one of the sister groups is only partly included in the group").
Peter coxhead (talk) 17:43, 3 August 2013 (UTC)

Subjectivity and polyphyly[edit]

An earlier version of the article said:

"A polyphyletic assemblage is characterized by one or more homoplasies: character states which are inferred to have converged or reverted so as to appear to be the same but which the cladogram implies have not been inherited from a common ancestor."

Peter Brown changed this to remove the 'subjective' wording "inferred", "appear to be" and "implies", with the edit summary "Deleted phrases implying that polyphyly is subjective. Polyphyly does not depend on human inference."

Now I agree that polyphyly is objective: a set of taxa either is or is not polyphyletic (based on a given definition). However, my worry is that stated baldly this doesn't convey the science involved. We don't know for sure whether a character is a homoplasy; it's a scientific hypothesis that it is and hence it's a scientific hypothesis that a set of taxa is a polyphyly. Polyphyly does not depend on human inference, but whether or not we believe that a set of taxa constitutes a polyphyly does depend on human inference. I'm not quite sure how this is best conveyed, though. Peter coxhead (talk) 15:17, 3 August 2013 (UTC)

It's common on Wikipedia to hide the effort involved in establishing properties and relations. The Author article, for instance, ignores the immense effort involved in establishing the authorship of medieval works. On the other hand, the Atomic number article does discuss the means by which an atomic number is established. If someone wants to augment Author or Cladistics or Polyphyly with discussions of history or methodology, fine, but I don't see the absence of such augmentation as a shortcoming. The prior wording in Cladistics did imply that polyphyly was subjective, and it was therefore appropriate to remove that implication. Peter Brown (talk) 16:22, 3 August 2013 (UTC)
I don't think that the prior wording did necessarily imply that polyphyly was subjective, although it could be read that way. Another way of reading it is that the first part ("[a] polyphyletic assemblage is characterized by one or more homoplasies") is an objective statement, but that the second part says that the concept of a homoplasy is a subjective one. This does seem to me to be a view taken in some of the literature.
The problem for me is that what we should be doing in Wikipedia, namely paraphrasing reliable sources and not making up our own definitions, is difficult in relation to cladistic concepts, because reliable sources are regularly confused and contradictory. Peter coxhead (talk) 18:41, 3 August 2013 (UTC)
If homoplasy is subjective and polyphyly is defined in terms of homoplasy, then I think it follows that homoplasy is subjective too. And to define homoplasy in terms of what people have inferred makes homoplasy, and therefore polyphyly, relative to human endeavor.
I think you mean "... I think it follows that polyphyly is subjective too." if so, this is a nice semantic point. Is it subjective that there are eight planets? The definition of 'planet' seems to be subjective, but once the definition is accepted then I'm reluctant to say that the number (now eight, once nine) is subjective. Peter coxhead (talk) 21:37, 3 August 2013 (UTC)
I do have to acknowledge that some homoplasies are based on terms like "yellow" or "gracile" that are vague enough that whether two specimens both possess it is a subjective matter. The polyphyly of Diandria though, seems devoid of subjectivity; either a plant has two stamens or it doesn't.
As far as what we should be doing, the policy is that, if reliable sources conflict, we should be presenting both sides. It is surely OK, though to ignore sources that are self-contradictory or totally confused. The trick to to decide what sources can be used. Despite Hennig's historical importance, I'm doubtful as to whether he should be cited except in a strictly historical context; after all, he did die in 1976 before molecular sequencing became practical, and such sequencing dominates current literature (along with parsimony, maximum likelihood, and Bayesian techniques). You have not mentioned Oosterbroek except on talk pages, which I think wise; not only are his formulations too complex for the average reader, Google Scholar lists only 12 citations for his 1987 paper, less than one every two years. Even though he published in a major journal, I think his work must be considered fringe for that reason. He certainly does not represent a consensus.
Yes, some sources describe what's now done using genes and computer programs as what Mayr called cladistics. But its logic is entirely different. It doesn't begin by looking for apomorphies; it doesn't use the character concepts of Hennig at all. All it has in common are some prior assumptions (binary branching, all actual taxa terminal, etc.) Only tree based definitions make sense in the molecular phylogenetic context. Peter coxhead (talk) 21:37, 3 August 2013 (UTC)
You say in the previous section, " I can source both kinds of definition for all three concepts, if you want." Well, I do want sources, but Cladistics#Terminology for taxa does provide sources, which may be adequate. I'll get back to you on that.
By the way, thanks for setting extended quotations off in separate paragraphs instead of using {{tq}}. Personally, I don't find the serif/sans-serif distinction easily to spot and green doesn't show up well on my screen.
Peter Brown (talk) 20:27, 3 August 2013 (UTC)

Definition of symplesiomorphy[edit]

Ignoring parentheses, the first two sentences of the definition at Cladistics#Terminology for character states read:

A plesiomorphy or ancestral state is a character state that a taxon has retained from its ancestors. When two or more taxa that are not nested within each other share a plesiomorphy, it is a symplesiomorphy.

Both crocodilians and insects have certainly inherited jointed legs from their ancestors so, from the first sentence, it seems that jointed legs are a plesiomorphy of the crocodilians as well as of the insects. Neither Crocodilia nor Insecta is nested within the other. Does that mean, from the second sentence, that the possession of jointed legs is a symplesiomorphy of these two clades? If so, this would be a good example to use. Peter Brown (talk) 23:01, 12 October 2013 (UTC)

As far as I understand it, it has to be the same plesiomorphy. Crocodiles and insects didn't inherit jointed legs from their LCA. Peter coxhead (talk) 00:34, 13 October 2013 (UTC)
Tricky! The most obvious Wiktionary definition of "share" is "have in common", but jointed legs are certainly common to insects and crocodiles. These sentences need to be reworded if homology is relevant.
No, I think you are mis-reading the paragraph. It says that they "share a plesiomorphy" not that they "share a character", and a plesiomorphy is necessarily homologous. (However, "retained" has to be read as "retained by inheritance"; to be really precise perhaps this should be said.) Peter coxhead (talk) 01:07, 14 October 2013 (UTC)
There are further problems with this paragraph. The last sentence reads:
Since cold-bloodedness is a plesiomorphy, inherited from the common ancestor of traditional reptiles and birds, and thus a symplesiomorphy of turtles, snakes and crocodiles (among others), it does not mean that turtles, snakes and crocodiles form a clade that excludes the birds.
After a long subordinate clause, one finally reaches the subject of the sentence, which turns out to be "it". To me and perhaps other readers, the reference of "it" is quite unclear. Why should anyone think that it (whatever it is) means that turtles, snakes and crocodiles form a clade anyhow, never mind one that excludes the birds? Further, pelycosaurs are cold-blooded "traditional reptiles"; are they included in the group of which cold-bloodedness is a symplesiomorphy?
Why are birds mentioned at all?
Someone might say, "Hey, Peter Brown, you're a competent writer and the #2 editor of this article; if you don't like the paragraph, why not fix it up?" The answer is that I am not clear on what a symplesiomorphy is and therefore cannot provide a definition with any confidence. Peter Brown (talk) 18:05, 13 October 2013 (UTC)
I don't think that symplesiomorphy explains it very well either. Here's my take. Suppose we have three groups of organisms, A, B and C, and it seems that all members of A and B have a particular character and all members of C don't have the character. We are trying to decide whether A and B fall into a clade which excludes C. Evidence that they do would be that the character was an innovation (apomorphy) in an ancestor of A and B that was not an ancestor of C and that the character has been inherited from this ancestor by A and B (i.e. it is a shared apomorphy = synapomorphy of A and B). What would not be evidence that A and B fall into a clade which excludes C would be that A and B retain some character from an ancestor of A, B and C, but that C or C's ancestors lost it. Such a character is a plesiomorphy – a character inherited by many groups from an ancestor sufficiently distant that it has been lost in many of that ancestor's decendants. The character would then be a shared plesiomorphy = symplesiomorphy of A and B, and it would offer no evidence that A and B fall into a clade which excludes C.
Let A = cartilaginous fish (Chondrichthyes), B = ray-finned fish (Actinopterygii) and C = tetrapods (Tetrapoda). The character "possesses internal gills" was not an innovation (apomorphy) in an ancestor of cartilaginous fish and ray-finned fish that was not an ancestor of tetrapods. It was an innovation in an ancestor of all three, and has been lost along the line leading to tetrapods. Hence it is a plesiomorphy (a "basal" or "primitive" character) and a shared plesiomorphy = symplesiomorphy of cartilaginous fish and ray-finned fish. It offers no evidence that cartilaginous fish and ray-finned fish form a clade which excludes tetrapods.
To me an under-emphasized issue in many explanations of symplesiomorphy is that three groups need to be considered, since to call a character a symplesiomorphy of two groups we have to rule out its being a synapomorphy, and two groups alone won't do this. "Possesses internal gills" is a symplesiomorphy of cartilaginous fish and ray-finned fish relative to tetrapods, but is a synapomorphy of cartilaginous fish and ray-finned fish relative to exchinoderms.
This need for a third group is why birds have to be mentioned in saying that "cold-bloodedness" is a symplesiomorphy of traditional reptiles. (Actually I think this is a poor example, since non-homeothermy isn't really a character, but rather the absence of one. You can see that if you ask "of what groups is 'cold-bloodedness' a synapomorphy?" No ancestor of clade-defined reptiles evolved the non-ability to control its body temperature.) Consider "possesses epidermal scales on the head and body" (not a terribly good character either). It's a symplesiomorphy of traditional reptiles relative to birds, but a synapomorphy of the clade Reptilia (=Sauropsida) relative to synapsids (or at least I think so). Peter coxhead (talk) 01:07, 14 October 2013 (UTC)
I'm not finding your four paragraphs easy to understand.
Paragraph 1: What would not be evidence that A and B fall into a clade which excludes C would be that A and B retain some character from an ancestor of A, B and C, but that C or C's ancestors lost it.

The fact that Gila monsters and monitor lizards have legs retained from an ancestor of Gila monsters, monitor lizards, and snakes but snakes' ancestors lost them is evidence that Gila monsters and monitor lizards fall into a clade (specifically Anguimorpha) that excludes snakes, isn't it?

No, it's not (in the strict Hennig approach). Replace "Gila monsters" by "crocodiles". Crocodiles and monitor lizards have legs retained from an ancestor of crocodiles, monitor lizards and snakes. Having legs is a plesiomorphy of quadrupeds; it's a symplesiomorphy for crocodiles and monitor lizards exactly as it is for Gila monsters and monitor lizards. Therefore it offers no evidence either that crocodiles and monitor lizards form a clade excluding snakes or that Gila monsters and monitor lizards form a clade excluding snakes. Peter coxhead (talk) 12:46, 16 October 2013 (UTC)
Paragraph 2: The character "possesses internal gills". . .offers no evidence that cartilaginous fish and ray-finned fish form a clade which excludes tetrapods.

As with the Anguimorpha, I'd say that it does offer strong evidence. As it happens, there is even stronger evidence that there is no such clade. I suggest that consideration of evidence be excluded from these formulations; what matters are the traits and the phylogeny.

Again, it does not, because an ancestor of cartilaginous fish, ray-finned fish and tetrapods had internal gills. Sharing the retention of an ancestral feature which is also retained by other groups or has been lost in other groups is a symplesiomorphy by definition. In the Hennig approach symplesiomorphies never offer evidence for a clade (rather for a para- or polyphyletic group). Peter coxhead (talk) 12:46, 16 October 2013 (UTC)
Paragraph 3: "Possesses internal gills". . .is a synapomorphy of cartilaginous fish and ray-finned fish relative to exchinoderms.

For symplesiomorphies, your preferred schema is "σ is a symplesiomorphy of A and B relative to C". For synapomorphies, though, do we want the "relative to" part? Can't we just say that "Possesses internal gills" is a synapomorphy of cartilaginous fish and ray-finned fish?

No, not if we know that lobe-finned fish (for example) inherited the same feature from the same ancestor. That by definition makes it a symplesiomorphy. There is (by assumption) a common ancestor for all deuterostomes and hence for cartilaginous fish, ray-finned fish and echinoderms. That ancestor did not have internal gills (so far as I know). So internal gills were an apomorphy of some ancestor of cartilaginous fish, ray-finned fish (and other groups, but not echinoderms). All the descendants of the first ancestor with internal gills form a clade; the evidence is that they all either have internal gills or have clearly lost them. If you want not to specify a "relative to" part for a synapomorphy then you have to include the total clade which is descended from the ancestor for which the feature was an apomorphy. Peter coxhead (talk) 12:46, 16 October 2013 (UTC)
Paragraph 4: Same problem as in Paragraph 3. Also, are you saying that synapsids don't have epidermal scales on the head and body? What about armadillos?
I'm far from an expert on vertebrates! My understanding was that synapsids and mammals, including armadillos, have a different kind of scale, but this may be wrong. Ignore the example if my facts are wrong. Peter coxhead (talk) 12:46, 16 October 2013 (UTC)
Peter Brown (talk) 18:29, 15 October 2013 (UTC)

Another attempt[edit]

I think I need a diagram to help me explain what I mean (which is hopefully what Hennig meant, but I find this stuff confused in the literature).

Consider three features distributed among four groups like the following. "Yes" means "has or has lost".

Group Feature 1 Feature 2 Feature 3
A yes no no
B no no no
C yes yes yes
D yes yes no

The universe of discourse consists only of A, B, C and D and Features 1, 2 and 3.

Then in the Hennigian approach, we look for apomorphies.

  • Feature 1 is a synapomorphy of A, C, D, i.e. A+C+D are a clade
  • Feature 2 is a synapomorphy of C, D, i.e. C+D are a clade
  • Feature 3 is an autapomorphy of C

Then the only possible cladogram (bar rotations) is:


B: −−−


A: 1−−


D: 12−

C: 123

So far, so good. What about (sym)plesiomorphies? Since Feature 1 is a synapomorphy of A+C+D, it's a symplesiomorphy of any subset, i.e. A+C, A+D, C+D. C+D do form a clade but their shared possession of Feature 1 offers no evidence for this in the Hennigian approach because it's a symplesiomorphy with respect to A (in my terminology). [In other phylogenetic methods, e.g. parsimony, it would add support for this cladogram.] If Feature 3 is a radical innovation, and Feature 2 only a minor change, in traditional Linnean classification we could argue for putting A and D in one (paraphyletic) group and C in another. But there's no Hennigian evidence for this: all that A and D share is Feature 1, which is a symplesiomorphy (with respect to C). What about the absence of Feature 3? That's a (kind of) symplesiomorphy, in this case inherited by A, B and D from their common ancestor, but C has "lost" the absence. The absence of Feature 3 can define a paraphyletic group, but not a Hennigian clade, which must have a synapomorphy.

In the strict Hennigian approach, the only groups of real interest are monophylies; paraphylies and polyphylies are just different kinds of non-monophylies. In the same way, the only features of real interest are synapomorphies – those features which define a clade because they are shared by all members of the clade and are derived from the ancestor of the clade (although they may have been lost later). Symplesiomorphies are just features which aren't synapomorphies and hence characterize non-monophylies.

A synapomorphy necessarily involves at least one group outside the clade – an outgroup in modern phylogenetic studies – from which the "morphy" is "apo". That's what I mean by "with respect to". (For example, in my cladogram above, A is outside the clade C+D because it doesn't have the synapomorphy Feature 2. Feature 2 is a synapomorphy of C+D with respect to A – or indeed with respect to any other group.)

Can we describe Feature 1 as a synapomorphy of C+D? The simplest answer is "no". An "apo-morphy" must be "apo" something. In this case it's "apo" the absence of Feature 1. So the only group of which Feature 1 is a synapomorphy is A+C+D. Another answer, and here I think I may be departing from the Hennigian approach (although I find explanations of it unclear on this point) is that Feature 1 is a synapomorphy of C+D with respect to B. Feature 1 is a shared (i.e. syn) difference (i.e. apomorphy) of C and D from B. But it's not a shared difference of C and D from A, so it's not an "absolute" synapomorphy.

A symplesiomorphy necessarily involves at least one group within a clade whose sharing of the clade-defining feature (or possibly having lost it) is being ignored – that's what I mean by "with respect to". (For example, in my cladogram above, if we group A and D on the basis of sharing Feature 1, we're ignoring C which also has Feature 1. Feature 1 is symplesiomorphy of A and D with respect to C.) Peter coxhead (talk) 14:22, 16 October 2013 (UTC)

Key summary

Once you have defined a clade on the basis of a shared inherited feature which differentiates it from all non-members of the clade (i.e. a synapomorphy), that feature is of no value whatsoever in determining whether groups within the clade form subclades. From that point on it's a symplesiomorphy. Peter coxhead (talk) 14:31, 16 October 2013 (UTC)

Thank you for all your work on the above. I'm certainly clearer on some points. I'm baffled, though, by the concept of "Hennigian evidence". Hennig was a recognized expert on the Diptera; surely, he had to adopt his colleagues' sense of evidence to be taken seriously. X is evidence for Y just in case Y is more probable given X than otherwise. The fact that cartilaginous and ray-finned fish both possess internal gills and tetrapods don't is evidence in this sense that the two fish groups are members of a clade that excludes tetrapods. The latter proposition is false, to be sure, but the former is still evidence for it. Facts, like the fact that an ancestor of cartilaginous fish, ray-finned fish, and tetrapods had internal gills, can be introduced as additional evidence for or against a proposition, so that the evidence relations get more complicated, but prior to their introduction, we have only:
X: cartilaginous and ray-finned fish both possess internal gills while tetrapods don't, and
Y: cartilaginous and ray-finned fish belong to a clade excluding tetrapods.
Here, X is evidence for Y. Hennig could not have been a respected scientist if he insisted on a nonstandard sense of evidence. Nor do I remember encountering this sense in my scattershot reading of Hennig. Peter Brown (talk) 00:01, 17 October 2013 (UTC)
Frustrating to try to communicate in this way! (In case it's not clear, by "Hennigian" I mean only "following the method espoused by Hennig" not that Hennig himself said this.)
  • cartilaginous and ray-finned fish both possess internal gills while tetrapods don't It's not so simple because features can be transformed and so apparently lost. There don't have to be fully functioning internal gills; the question is whether there are any features of tetrapods which are homologous to internal gills and which can't be explained other than by having ancestors which did have internal gills. Similar features may not be homologous and homologous features may not be similar.
  • The fact that cartilaginous and ray-finned fish both possess internal gills and tetrapods don't is evidence in this sense that the two fish groups are members of a clade that excludes tetrapods. If tetrapods really did have no trace whatsoever of internal gills, then sure. However, embryological development as well as adult morphology shows that actually tetrapods do have features which are homologous to the internal gills of the two fish groups. (Gill slits appear in mammalian embryological development, for example.) There's no doubt that tetrapods are descended from ancestors with internal gills.
However, you are, I think, right to raise these questions. There have always been those who worried that there is a circularity in the Hennigian method. Synapomorphies determine clades, but are we really sure that we can decide what is a synapomorphy independent of what is a clade? Given that we know that cartilaginous fish, ray-finned fish and tetrapods are part of the same clade we're likely to look for evidence that tetrapods' ancestors had internal gills and that tetrapods retain vestiges of internal gills, because that's what should be the case. Similarly we discount symplesiomorphies as evidence for a set of groups forming a clade, but is it only because we know that the total clade includes other groups that we are able to decide that a feature is a symplesiomorphy?
All this is of doubtful modern relevance, since the computer-based methods now used, whether based on molecular data or on character data, don't use strict synapomorphies to determine clades. Peter coxhead (talk) 01:28, 17 October 2013 (UTC)

What is the topic of this article?[edit]

I find this article confused/confusing because it seems to me that it's not clear what its topic should be: is it "phylogenetic systematics" as used by Hennig (and called "cladistics" by its opponents), or is it the whole field of modern methods grouped under this term?

"Phylogenetic systematics" as Hennig used the term is not identical to "cladistics" for two reasons: firstly, "cladistics" was the preferred term of those hostile to "phylogenetic systematics"; secondly, "cladistics" is now used to cover a much wider range of methods than those of "phylogenetic systematics", some of which are directly contradictory to Hennig's methods.

Hennig advocated two main ideas which are related but logically different:

  1. Systematics should be based purely on evolutionary relatedness and not on morphological similarity. The only groupings (taxa) of value are monophylies (clades). The key aim should be to construct a hierarchy which corresponds to the evolutionary history of the species concerned.
  2. The only morphological features (characters) that are of value in determining phylogenetic relationships are (syn)apomorphies. Overall morphological similarity is irrelevant, since this may be due to symplesiomorphies, which are valueless in determining relationships.

(1) is now accepted as the aim of phylogenetic studies, although some taxonomists are then prepared to name a few paraphyletic groups in special cases, but with full awareness of their nature.

(2) in the way Hennig described it is simply not a feature of modern computer-based methods. Parsimony, which still seems to be the dominant method for analyses based on morphological characters, considers all characters used in the analysis equally, although the choice of characters may be influenced by whether they are considered derived or primitive. Statistical models (Bayesian, etc.) used with molecular data treat all sequences equally, although again the choice of sequences to be used may be based on the selection of homologs. But careful selection of synapomorphies which define clades simply doesn't feature. Although these methods aren't exactly the same as phenetics (numerical taxonomy), they are in many ways more similar than it to Hennig's manual methods.

However, I'm unsure how to fix the article. The Methodology section is particularly confused between the first part – an explanation of Hennig's terminology and methods, and the second part – an over-brief survey of the very different methods used now and their development. Peter coxhead (talk) 15:50, 19 October 2013 (UTC)


Having thought about the article and the related ones in Wikipedia, I propose that:

  • This article be restricted to "phylogenetic systematics", i.e. "cladistics" in sense of Hennig and his opponents.
  • Material which relates to the broader sense of "cladistics" be moved to "Phylogenetics".

Peter coxhead (talk) 18:12, 19 October 2013 (UTC)

I completely agree. I am not going to edit the Cladistics page, as I am a partisan in various disputes with people who think of themselves as practicing cladistics. But I do want to express myself here. There seem to be various positions:
  1. Cladistics is the position that all groups in the classification system be monophyletic.
  2. Cladistics is that, plus reconstructing the tree by nested synapomorphies.
  3. Cladistics is those, plus using parsimony methods when you have reversals or parallelisms in your data.
  4. Cladistics is all those, but only if you're a paid-up member of the Willi Hennig Society.
  5. Cladistics is numbers 1-3, plus using likelihood or Bayesian methods to infer trees.
  6. Cladistics is numbers 1-3 and number 5, plus inferring trees by distance matrix methods.
This page seems to take position number 6, or perhaps number 5. To me this seems deeply incoherent. The straightforward and coherent definition seems, to me, to be #1. It is a position on classification, not on how the phylogeny should be inferred. As such it is a sensible position, and certainly the dominant one these days. This page is basically talking about two things at once -- how we infer phylogenies and how we define groups in the classification system. It is a mess. But people should realize that my view on this is my own, and not shared by most systematists, who are wildly divided among the various posiitons listed above. The confusion in the main article about what "cladistics" is simply reflects the state of the field and the wildly mixed messages that it gives to the outside world. 'Nuf said. Felsenst (talk) 11:36, 31 October 2014 (UTC)
Our task is not to take any position, but to say what the positions are (in one article, or several). That allows us to say that the term is used in different ways, and indeed that there is confusion about it. What an article must not do is to get sucked into the confusion. The way to avoid this is to say clearly and verifiably who said what. Chiswick Chap (talk) 12:25, 31 October 2014 (UTC)
@Chiswick Chap: I don't think it's quite so simple. As Felsenst's analysis above demonstrates, I think, a large part of the problem is that the term "cladistics", when interpreted broadly, doesn't correspond to a topic, and a good Wikipedia article – indeed any encyclopedia article rather than a dictionary entry – needs to be about a topic. The term "cladistics" has been used in so many different senses that trying to cover them all in one article would still produce a confused and confusing article even if each sense could be clearly explained using reliable sources (which I doubt).
Going back to my suggestion, what exactly distinguishes this article from Phylogenetics? If there are two articles, there must be some demonstrable distinction between them. What exactly makes the Methodology section and the two Terminology sections part of "Cladistics" rather than "Phylogenetics"? Peter coxhead (talk) 14:49, 31 October 2014 (UTC)
So we just say clearly up front that this article covers Cladistics in some narrow sense (say, #1), and see Phylogenetics for the rest? That sounds sensible and defensible (and I'd be happy to !vote for that). It means removing all uses of 'cladistics' in the broad sense from the article. Chiswick Chap (talk) 16:07, 31 October 2014 (UTC)
The problem is that among systematists there is general confusion as to which of the above definitions of cladistics is correct. To me #1 is the only really defensible one, but my views on this are controversial, to say the least. No matter what position the page takes a lot of people in the field are going to say it is wrong. Enjoy, folks. Felsenst (talk) 05:50, 2 November 2014 (UTC)

Stem-based taxon[edit]

The term Stem-based taxon redirects to this article, which however does not even mention it! Could the term please be placed in boldface in the lead, with a brief explanation, and then discussed somewhere in the article? Chiswick Chap (talk) 15:51, 23 October 2013 (UTC)

It shouldn't redirect here. This concept is covered at Phylogenetic nomenclature#Phylogenetic definitions of clade names. I'll alter the redirect. Peter coxhead (talk) 22:27, 23 October 2013 (UTC)

Parallel evolution: divergence versus convergence?[edit]

It gets even worse than the article and talk page seem to have suggested. You could in theory have 2 different (yet maybe closely-related) species/varieties experience a common environmental factor and thus have a preference for the same mutation/s. Eyes for example seem to have evolved several times since animal life evolved. Granted, they aren't physically identical forms of vision. Actually, I was wrong. It's more like dozens. O_o

How do you contend with timelines of family/group/genus/species/subspecies trees that seem paradoxical because of this issue? The article implies that only one of the cladistic trees is accurate, but it seems that simplistic trees might ALL be wrong for a given group of species. 2601:1:9280:155:6155:1881:4B26:F3D4 (talk) 12:25, 6 March 2014 (UTC)

This is dealt with by the parallel optimization of many characters on the same tree. The thought is that convergence is unlikely to happen in enough characters to dominate the tree topology, provided enough are included in the analysis. For the purposes of examples in the article however, it makes most sense to talk about one character at a time. de Bivort 14:26, 6 March 2014 (UTC)
It's also important to select and define "characters" carefully. There are many different kinds of eye, as noted above, so "possesses eyes" is not a useful character in a cladistic analysis. "Simplistic" trees (i.e. trees constructed using simplistic characters) are highly likely to be wrong. This is one of the serious objections to the original Hennigian method which relied on individual characters to identify branches in the cladogram, rather than statistical aggregation over many characters as in modern computer-based methods. Peter coxhead (talk) 21:02, 12 March 2014 (UTC)

For those interested in phylogenetics, and clade presentations[edit]

Could you have a look at this effort, here, to use clade diagrams to summarize pharma business mergers and acquisitions (M&A). My take at present is that the images created are devoid of standard quantitative meaning—nothing is captured by vertical and horizontal line lengths, as far as I can tell—and so they are a misapplication of this maths/graphic presentation method. Moreover, I argue that they are misleading (presenting a time axis, but not making spacing of events proportionate to the historical time differences), much harder to maintain (consider adding entries to a std Table versus this graphic), more likely to diminish article quality (in their ambiguity of content, again, over a std Table with clear headings), and therefore practically amenable to decay as a result. I would add to this, in this esteemed cladistics context, that they would make those who trained us, and other purists in methodology and meaning… turn in their graves/beds. After having a look at the User page and at a couple of pages linked on that sandbox page, leave your opinion [at the link give below], regarding the overall effort? Thanks for your opinion. Cheers. Le Prof Leprof 7272 (talk) 01:37, 23 March 2016 (UTC)

Jytdog deleted the above posting (edited slightly here for clarity), and moved the question to a new forum, the Business area, now here.
I reply in greater detail with my views there.
I stand by the fact that this Cladistics article is a gathering place for experts in cladistics, and so an appropriate venue to call out for experts, regarding how a tool of cladistics is being cross-applied (I argue misapplied) to an area of business graphics.
I ask your input, and apologize for your being left out of the conversation, by the overstepping deletion. (I acknowledge Jytdog's right to have an opposing opinion on the matter, but not to limit discussion, by deleting my Talk page entry. Enter a post saying inappropriate, but do not delete. That oversteps.)
All coming from the Cladistics area, to address the cross-application at the link above: I would appreciate if you state for the record there, the real practical knowledge you have on on this matter (e.g., if you have ever actually worked on a project involving cladogram computations/presentations), for sake of transparency, please. Le Prof Leprof 7272 (talk) 15:45, 23 April 2016 (UTC)

External links modified[edit]

Hello fellow Wikipedians,

I have just modified one external link on Cladistics. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

You may set the |checked=, on this template, to true or failed to let other editors know you reviewed the change. If you find any errors, please use the tools below to fix them or call an editor by setting |needhelp= to your help request.

  • If you have discovered URLs which were erroneously considered dead by the bot, you can report them with this tool.
  • If you found an error with any archives or the URLs themselves, you can fix them with this tool.

If you are unable to use these tools, you may set |needhelp=<your help request> on this template to request help from an experienced user. Please include details about your problem, to help other editors.

Cheers.—InternetArchiveBot (Report bug) 02:23, 20 May 2017 (UTC)