I don’t see how PN is a substitute for classification. It’s still a system that is used for categorizing and naming objects. It’s just that the criteria for naming has changed, and the definitions have become fuzzier.

In the end, these are my two main problems with cladistic based classification. Under PN rules, if I go fossil hunting and come across a fossilized scute, I can’t say what it belongs to. At least not in the conceptual sense. Under the “old rules” of classification I could look at this hypothetical scute and see that it was oblong with lots of deep scarring. A character that would fall in as diagnostic for a member of Crocodylia (actually crocodylomorpha, but I’m trying not to get too bogged down in details here). So I could then say that this scute belongs to some kind of crocodylian because of these features.

Under the rules of PN, where names are based off of whose related to who, and the character diagnoses get removed, I’m forced to look at this oblong scute with its deep scarring and say just that. It is an oblong scute with deep scarring. Short of putting these characters in a data matrix (and a fitting one at that. Not some invert data matrix, or one featuring just sharks) I am unable to say exactly what this scute belongs to.

Of course in the real world, I would still classify this scute based off of my initial observation. All paleontologists and biologists do this regardless of what classification scheme they use. Even the cladistic programs get their results from analysis of the characters.

So in the end, if the name still appears to be defined by the characters. Why bother removing them from the definition?

I get that this way is less stable. As more specimens show up, the characters that diagnose a certain group are likely to change. Still this seems a small price to pay for the ability to say that taxon x is defined as those creatures that share the following morphological traits. Traits that one can actually see.

The fuzziness of the names comes from the all inclusive nature of PN. For example, if one used Osteichthys (or Sarcoptergyii, or whatever) as an example. The original definition limited one to “bony fish.” Right off the bat there is a concept in place that helps one more quickly find what one is looking for. If one views osteichthys as being those aquatic critters with bony skeletons, then one is not going to go looking for an osteichthyan in a terrestrial setting. Obviously there are cases where concepts can get in the way. Reptilia and dinosaurs being a prime example. Still these should only serve as cautionary tales of what not to do with a conceptual name (i.e. concepts should help guide ideas, but they shouldn’t control them). In the case of Reptilia, the concept is in dire need of an overhaul. Removing it altogether due to baggage seems pointless. Especially given its inertia in the popular mindset, as well as a lack of any real replacement names (sorry but I don’t see sauropsida taking off anytime soon). Returning to my example, if one instead, pulls for a more phylogenetic based approach, then suddenly we have mammals, reptiles and birds getting shoved into the Osteichthys group. Now the concept associated with name is destroyed. The term becomes fuzzy and essentially meaningless. I can now find osteichthyans on land and in the air. Heck there are even a few that make sojourns into space. So what does the term mean? Well, nothing really. Not anymore.

If you take away the concepts associated with the names, then don’t you remove the purpose of giving names at all? How can I talk to someone else about bony fish, if there is no name for these fish anymore? Okay, I could just use “bony fish” when talking to another person about bony fish, but then all I’ve done is take a common name and used it to replace a taxonomic name. Alternatively I could stick some unwieldy qualifier in front of it like non-tetrapodan osteichthyan (or non-tetrapodan sarcopterygian, etc) in order to bring the name closer back to its original meaning. So, then, what is the point of the taxonomic name at all?

I believe it is this reason why we don’t see paraphyletic groups disappearing from biology. Despite your misgivings there are those of us out there that do find these groups interesting and worth talking about. Paraphyly only gets muddy at the ends, where creatures don’t easily fit in one group, or the other. These, though, will always be the exceptions rather than the rules. If necessary one could stick a qualifier on these guys like how the old term “mammal-like reptile” was once used (a term that would be just fine if it weren’t for the fact that mammals didn’t evolve from reptiles). It’s still an unwieldy and rather annoying qualifier, but at least it refers to fewer taxa.

Another alternative is that one could come up with a specific name for that “transitional” group. Conceptually this could prove chaotic as one find more and more taxa that blur the lines, thus requiring more and more new names. Realistically, though, this would be unlikely to happen for most groups as this would require a fossil record with a resolution well and above the one we actually have.

Okay, I think I’m going to stop now. This has been enough of a brain dump for one night. Besides I’m pretty certain this comment is now longer than the original post. :)

The real problem here is that they are trying to simultaneously use arbitrary ranks and phylogenetic nomenclature (what you are semi-erroneously calling “cladistics”–”cladistics” generally refers to the scientific process of analysis, not to a system of nomenclature). They feel a need to designate something a “class”, but this is a useless designation that tells us nothing. It doesn’t do any harm, though … at least, until you consider fossil taxa.

Which class does _Ankylosaurus_ belong to? _Stagonolepis_? _Pareiasaurus_? _Ichthyosauurus_? The only options are:
1) Expand the classes to be total groups: make _Ankylosaurus_ an avian, _Stagonolepis_ a eusuchian, etc.
2) Make hundreds of impossibly small classes: Class Deinonychosauria, Class Lagosuchidae, Class Ophiacodontidae, etc.
3) Shoehorn them into paraphyletic classes, thus defeating the whole purpose.

Note that, even under #1, the ancestors of crown groups cannot be classified beyond a certain rank without creating paraphyletic taxa.

So, it seems to me that we must lose ranks or clades. The decision is fortunately an easy one: ranks are almost completely uninformative and subjective, while clades reflect propinquity of descent, a useful bit of information if there ever was one.

Also, it’s a mistake to say that advocates of phylogenetic nomenclature are confusing phylogeny with classification. Quite the contrary, we see phylogeny as a *substitute* for classification! Instead of spending time dithering over whether Eusuchia should be a class or a subclass or an order or a parvorder, we just say that it’s the last common ancestor of _Alligator_, _Crocodylus_, _Gavialis_, and _Hylaeochampsa_ plus all descendants thereof. Then the phylogeny does the work for us — no classification necessary. Under this system, when people argue about what belongs to which taxon, they are actually arguing about something scientific, not just being fussy.

Kudos to CNAH for showing us how silly rank-based nomenclature can be!