In the children's picture book The Kuia and the Spider by New Zealand author Patricia Grace, an elderly woman (kuia in Maori) is challenged by a spider living in her kitchen, who keeps saying to her, "Hey, old woman! My weaving/cooking/etc. is better than yours!" Each successive challenge leads to frantic competition, as the woman and the spider try to outdo each other ("cooking" for the spider, of course, refers to the process of catching, wrapping and breaking down flies). However, each of these competitions ends in an effective draw - they only way they could end, because neither competitor is willing to admit the other's superiority in these ultimately subjective comparisons.
As my regular readers will probably be aware, I'm not a big fan of rank-based classifications (see here for one of my earlier rants on the subject. The ultimate problem with rank-based taxonomy is that the urge to inject some sort of "reality" into the concept of ranks is irresistible. Defenders of the rank system claim that this is not a problem - ranks only indicate relative positions, and it does not matter that a "genus" of insects is not directly comparable to a "genus" of mammals, because that's not the point of ranks. Unfortunately, they then forget this point themselves - instantaneously, even:
Everyone accepts that Linnaean ranks are subjective, and yet there is no benefit in abandoning ranks because they have proved to be of such value to users of classifications, and genera and families, for example, act as valuable surrogates for species in large−scale evolutionary and ecological studies. (Benton, 2007)
If ranks are subjective (as supposedly everyone says they are), then higher ranks cannot possibly act as surrogates for species for the simple reason that one author's rank-ometer may be (and, in practice, usually is) calibrated differently from another author's.
The point I really wanted to make today, though, is another aspect of the fallacy of thinking of ranks as "real". The current taxonomic ranking system, as we all know, is anchored on the principle that Kind People Can Often Find Good Sex. The seven ranks referred to by that mnemonic are the primary and more or less mandatory ranks of the system, while all other more optional ranks are conceptualised in their relationship to the primary seven. Unfortunately, this leads directly to the idea that taxa at those "primary" ranks are somehow more significant than taxa at the "subsidiary" ranks. Take this comment made recently on the Taxacom mailing list:
Ideological extinction is the fate of autophyletic (descendant groups recognized at same taxonomic level as ancestor) taxa, and of paraphyletic taxa split into non-recognition. Although taxa of high visibility (Aves, polar bears) seem immune to ideological extinction, many groups simply disappear from classification because they were embedded in an ancestral group of the same rank. In my own field, bryology, three families (Ephemeraceae, Cinclidotaceae, and Splachnobryaceae) have been sunk in a recent influential phylogenetic classification into one larger one (Pottiaceae) without discussion because they were autophyletic in previously published molecular trees; their names do not appear anywhere in the classification, which offers no synonymy.
The author of the comment, Richard Zander, is complaining that the subsumation of the three smaller families into the larger family obscures the distinctiveness of the three smaller groups. There is no particular reason why this should be so - if the taxa originally labelled by those three names are still good taxa, then they should still be perfectly recognisable whatever rank they are put at. The only reason why it should make a difference that they have lost their status as separate families is if there is some particular significance to being a "family" as opposed to a subfamily, tribe or whatever. If ranks are truly only relative, and there's no real significance to them, then why are you complaining?
Benton, M. J. 2007. The PhyloCode: beating a dead horse? Acta Palaeontologica Polonica 52 (3): 651-655. (Thank you to David Marjanović for pointing this out to me.)