[CPN] Revision of proposed changes in Art. 21

[David Marjanovic]

> Of course we could have an  arbitrary rule that the species name must
 > agree with the last clade name ("address") in a series, or the first,
 > or N - 1, but there really isn't any point to doing this given that
 > the combinations are no longer being treated as binomina (wherein the
 > species name is an adjectival modifier of the genus name and
 > therefore must agree with it in gender, number, and case)--they are
 > just species names with clade "addresses".

As I said, this is a good argument. However, as far as the rank-based 
codes are concerned, adjectival species names/epithets don't have a 
default form that could be used without having to make a decision. If we 
want such a default form to exist, _we_ have to put that into the 
PhyloCode. Well, if so, then how? As I explained at the beginning of 
this discussion, we could go with the version in the original 
publication, or we could go with currently prevailing usage. The 
original can be very hard to find, and is often different from the 
prevailing usage of the last 100 or 200 years, so we'd probably go with 
current usage. But if so, we run into trouble where different forms are 
in current widespread use, and sometimes have been for decades. So I 
suggested we give up, don't define a default form, and follow the 
rank-based codes in not treating the endings as part of the names.

Thus:

1) Yes;
2) no, make all agreement optional.