Predictivity vs Useful

Michael A. Ivie mivie at MONTANA.EDU
Sun Feb 13 14:37:34 CST 2005


Swore I wouldn't do this, but the flesh is weak....

[For non-entomologists who care to follow this, the Nematocera are a grade of fly families at the base of the Order Diptera, from which the purportedly monophyletic Brachycera emerged.]

Ken,

Of course I use Nematocera, because it is useful at the current state of knowledge, but it is not predictive, which was the topic (again, you have trouble following the topic at hand). To show how this differs, I give an example: I also use Republican, which is useful, but not predictive.  Luckily the children of Republicans, through education, can be redeemed, and unfortunately, some misguided children of intelligent parents grow up to be Republicans.  Thus, you cannot predict anything about the phylogenetic characteristics of a Republican, even though you know they are scary (the useful part).

You repeatedly confuse "Useful" with "Predictive."   I define a "Useful" classification as one that provides a pigeon hole to which a group can be returned repeatedly and with certainty.  Vermes was useful in its day, and while Nematocera is more useful than Vermes, it still does not rise to the level of the most useful taxon, i.e. one that is also "Predictive."  If you know something is a Nematoceran, you only know that because it IS a fly, but is NOT Brachycera.  Thus, the only characters the Nematocera have in common are those that are already present in the ground plan of the Diptera.  The only predictive taxa in this process are Diptera and Brachycera. You cannot predict anything further about a Nematoceran of uncertain position BECAUSE it is a Nematocera.  Simple.

FYI, in fact, the vast majority of families that I have published on are useful, but not predictive.  We simply don't know that much about insect phylogeny yet.  Thus, you are also greatly wrong on what I will and will not recognize.  As I tell my students, "keep your eye on the goal, but remember that incremental progress is the best we can hope for tomorrow."  However, why would we ever give up a predictive taxon for a merely useful one?

Now, go back and try to understand Curtis' points.  They will help you if you can be helped.

Mike

Ken Kinman wrote:

> Hi Don (and others),
>       Michael has obviously mixed apples and oranges (Vermes and Nematocera) in his examples.  He calls them both non-monophyletic, and we all know that Vermes is horribly polyphyletic and worthless.  On the other hand, Nematocera is paraphyletic, BUT it is an extremely useful taxon (even some strict cladists still use it).
>
>       So is Nematocera predictive?  You betcha it is!!!  If you find a fly (possessing Dipteran synapomorphies) which lacks the synapomorphies of Brachycera (= Orthorrhappa and Cyclorrhapa), you can be almost certain it is a Nematoceran.  Or conversely, any new nematoceran you find will almost always have only the nematoceran plesiomorphies.  Neither holophyly or paraphyly is 100% predictive due to homoplasies (especially those pesky reversals).   The literature is full of references to Nematocera (a paraphyletic taxon), so refusing to put it into a classification (just because it is paraphyletic) would indicate an extremely obstinate and close-minded strict cladist is the culprit who wants to exclude this additional information just because their religion (Hennigian "rules") forbids it.
>
>       I can fully understand Michael's displeasure with polyphyletic groups (everyone can), and even his displeasure with unannotated paraphyletic groups.  But when they are properly labelled as such, paraphyletic groups actually add more information to a classification.  And if strict cladists would look beyond their own particular branches of the Tree of Life, they would realize that we can't have a real Tree of Life without paraphyletic taxa.
>
>      Here is the initial part of my classification of Diptera (the full classification was posted on Taxacom on 17 March 2004; see the Archives).  The first version has suborders inserted and the second version doesn't (same topology, just less information).  Both are cladistic in an informational sense, but the first is divided into three very useful suborders (and the first two of those---shown here---are both paraphyletic).  In a group like Coleoptera (which is cladistically very poorly known by comparison), paraphyly can be even more necessary and useful.  The old inexplicit paraphyly should definitely be discarded, but that does not mean this modern crusade to attack every single paraphyletic taxon is much better (and in some ways, it is even worse).
>
> ORDER DIPTERA
>   1  Nematocera% (primitive flies)
>           1  Tanyderidae
>           B  Ptychopteridae
>           2  Thaumaleidae
>           b  Simuliidae
>           c  Ceratopogonidae
>           d  Chironomidae
>           B  Dixidae
>           C  Corethrellidae
>           D  Chaoboridae
>           E  Culicidae
>           3  Nymphomyiidae
>           B  Deuterophlebiidae
>           C  Blephariceridae
>           4  Tipulidae (sensu lato)
>           5  Psychodidae
>           ?  Axymyiidae
>           6  Perissommatidae
>              Scatopsidae
>              Canthyloscelidae
>           7  Pachyneuridae
>           B  Bibionidae
>           C  Mycetophilidae
>           D  Cecidomyiidae
>           ?  Rangomaramidae
>           E  Sciaridae
>           8  Anisopodidae
>           9  {{Brachycera}}
>               (= Orthorrhapa + Cyclorrhapa)
>
>  _1_ Orthorrhapa (primitive brachyceran flies)
>           1  Pantophthalmidae
>           B  Xylomyidae
>           C  Stratiomyidae
>           2  Vermileonidae
>           3  Xylophagidae
>           B  Ragionidae
>           ......... and so on...
>
> *************************************
>
>   ORDER DIPTERA (paraphyly eliminated; yes, I know how to do that too) (but therefore less balance; and less information; AND also more potential instability, especially in other groups that are less well-known than this)
>
>           1  Tanyderidae
>           B  Ptychopteridae
>           2  Thaumaleidae
>           b  Simuliidae
>           c  Ceratopogonidae
>           d  Chironomidae
>           B  Dixidae
>           C  Corethrellidae
>           D  Chaoboridae
>           E  Culicidae
>           3  Nymphomyiidae
>           B  Deuterophlebiidae
>           C  Blephariceridae
>           4  Tipulidae (sensu lato)
>           5  Psychodidae
>           ?  Axymyiidae
>           6  Perissommatidae
>              Scatopsidae
>              Canthyloscelidae
>           7  Pachyneuridae
>           B  Bibionidae
>           C  Mycetophilidae (sensu lato)
>           D  Cecidomyiidae
>           ?  Rangomaramidae
>           E  Sciaridae
>           8  Anisopodidae
>           9  Pantophthalmidae
>           B  Xylomyidae
>           C  Stratiomyidae
>          10  Vermileonidae
>          11  Xylophagidae
>           B  Ragionidae
>          ........and so on...

--
__________________________________________________

Michael A. Ivie, Ph.D.
Department of Entomology
Montana State University
Bozeman, MT 59717
USA

(406) 994-4610 (voice)
(406) 994-6029 (FAX)
mivie at montana.edu




More information about the Taxacom mailing list