Consensus

[Ken Kinman]

Paul,
      I can certainly understand your frustration about higher taxon names having varying circumscriptions.  That is one reason I stuck so long with the traditional division of Phylum Magnoliophyta into just two classes, Magnoliopsida (sensu lato, i.e. all dicots) and Liliopsida (monocots).

     However, I finally decided that a Three Class consensus was better, so in my February 2004 classification posted here, I finally split off Class Rosopsida (for the eudicots).  One could certainly try to split Rosopsida into various subclasses, Rosidae, Asteridae, and a few others, but the interrelationships are still so controversial that a consensus would presently be very difficult to achieve.  I still just code it as one large Class without any formal subclasses at this time.  I doubt that I will ever divide the smaller Classes Liliopsida and Magnoliopsida into formal subclasses.

     Meanwhile, I am still quite happy with the Three Class consensus classification, and I am glad I finally switched to it.  A Google search for "Rosopsida" yields an enormous number of hits, and the vast majority seem to refer to it as being equivalent to eudicots.  Some may restrict its meaning to something like rosids only (= subclass Rosidae), but I see no good reason to recognize more than three Classes for angiosperms (and as Paul noted, it only confuses matters to have different circumscriptions popping up all over the place).  As for the Class Magnoliopsida, it has always been paraphyletic and always will be, so I doubt the APG will ever agree to a formal consensus classification above the ordinal level.  Whether it be classes, subclasses, or even superorders, something near the base of angiosperms has to be paraphyletic (unless you want to keep splitting them up and creating a new cycle of instability and confusion).  I don't recall having ever seen how Cavalier-Smith would subdivide angiosperms, but I wouldn't be surprised if he divided them into three classes (since he employs occasional paraphyly as well).
  -----Cheers,
          Ken Kinman
P.S.  I certainly agree with APG that traditional classifications of angiosperms have far too many Orders.  That is why on the ordinal level, I chose to adopt the APG classification and then resplit the large ones where I think they got a little carried away.  :-)