The parts of the phylogenetic nomenclature that need to be "unique"

Question

I am not quite sure I have sorted through the diversity of approaches to nomenclature, but I had a specific question to help clarify what is going on. Basically I am just wondering what needs to be globally unique in the name of a species.

I understand how to construct a name using the Binomial Nomenclature, and am under the assumption that this "uniquely defines a species" (for simplicity, not considering all of the variations of subspecies and sub-x ranks). But I am not sure what exactly "uniquely defines" means. In addition, I am not sure if the names of all the different ranks are globally unique, and how the rank names are defined (if they are defined differently from the binomial nomenclature-defined species names) (but this is probably a separate question). This makes it hard to figure out what actually needs to be unique, and what is unique, in these names.

So to start from the top, we have Domains. Wikipedia shows 3 (Archaea, Bacteria, Eukarya). I have seen more in other places, but not too concerned with that currently. First question is, if these names are globally unique. Second (related, possibly separate) question is, if these names are decided according to the same codes as the binomial-nomenclature defined genus/species names, or if they have a different process. I'm not sure if a domain (and later, the kingdoms, phyla, etc.) is associated with one single specimen. I read that for at least the Genus', they have individual specimens that act as the model or reference point for the Genus (not sure if I can call a Genus a clade). So I'm wondering if this technique/requirement/workflow is applied to domains/kingdoms/phyla/etc. That is, there is an individual specimen in some museum somewhere that serves as the type / model for that clade (domain/kingdom/phylum/etc.). That would be interesting and helpful for better understanding how the uniqueness requirement works in the terminology.

So following domains we have Kingdoms. Wikipedia shows 7 of them. Same thing, if these names need to be unique from the Domain ones, and if they are chosen the same way.

Then Phyla has approximately 32 for Kingdom Animalia, 14 for Kingdom Plantae, etc. Same questions here, if these names need to be unique from Domains/Kingdoms/Orders/etc., and if they are chosen the same way (and have the model specimen in some museum somewhere).

Oh, in regards to that model specimen. If it turns out that each clade indeed does have a model specimen in some museum somewhere, that means it must have a genus/species name. This is a big part of the question: wondering if that means that a Kingdom like Plantae actually has a specimem along the lines of Plantae plantae. That would make me confused as to how the clade then gets its name. If there is a thing/specimen called Plantae plantae (used as the model/type representing that clade), then it seems almost circular. Plantae plantae is the name of a species, but it is also somehow [partially] defining the Kingdom Plantae! That is where I get confused. Anyway, back to the other big half of the question, about each level.

So then we have Class, Order, and Family. I see some naming rules now starting to pop up. For example, in botanical nomenclature, they have "-aceae" as the ending of the names for the most part. So this to me says that, yes, they are using the standard binomial nomenclature codes to construct the name of the Family. But there again I am confused, because the Family is only 1 word, whereas binomial nomenclature is 2 words. So don't see how those two fit together. But anyways, Class, Order, and Family start having custom endings to the words in some cases, so it seems to use binomial nomenclature codes.

Then Genus is where it starts to get a little bit more familiar. Still I don't know how the genus names are constructed (that will probably be answered with the above stuff), but I at least know that the Genus is the first part of the binomial nomenclature name.

Then Species has a name, but I guess the "full name" (as opposed to the "relative name") is <Genus> <Species>, rather than just <Species>.

So to summarize, I am wondering

What needs to be globally unique, and in reference to what. My first guess as of now is that Domain, Kingdom, Phylum, ..., Family, Genus, Species, all with 1 Latin name, are all globally unique. So you can't have a Species named the same thing as a Family or a Kingdom. But wait, no that doesn't make sense. So everything from Domain-Genus needs to be globally unique, but Species is only unique relative to the Genus. So you can have Volpus volpus (for example). But you can't have an Order named Foo and a Kingdom named Foo, that wouldn't work (that is my understanding at this point).

Answer 1

Binomial nomenclature does not uniquely define a species unfortunately, there are several duplicates.

The Wikispecies directory shows 5 duplicates: each of this is with an organism from kingdom Animalia and another from either Plantae or Fungi.

Therefore, you could conclude that binomial nomenclature is a unique identifier within a kingdom.

Answer 2

Many of the problems you are discovering are the basis for phylogenetic classification. Without necessarily answering each of your question specifically, I'll mention some of the basics of phylogenetic classification as compared to traditional Linnean classification:

1. Species are the main "currency". We classify species and species alone.
2. All taxonomic groups higher than the species level are arbitrary. The Linnean system of Kingdom - Phylum - Class, etc. has no biological meaning. For example, Reptilia and Aves can't both be classes, because birds are reptiles (a class within another class).

These two points alone address most of the questions. There are no general rules (and lots of arguments) for naming clades above the level of species. Clades are names following the main rule of "the most recent common ancestor of [taxon] and all of its descendants". This conveniently does away with all the business about "Supraclass" and "Infraorder" and such.

The implication of this scheme is that there are no (extant) species at the bifurcations, so there is no taxon that is not assignable to a "tip" on the tree. In other words, "we do not find direct ancestors in the fossil record", which basically answers your third question.