[Taxacom] ICBN (orthography of geographical epithets)

[Paul van Rijckevorsel]

----- Original Message ----- 
From: "Guido Mathieu" <guido.mathieu at taxa.be>

> This makes sense to me. May I rephrase it this way:

> When Latinization of a geographical name results in two Latin words, the
> use of a hyphen is to be accepted. 'Accepted' means that a hyphen is not
> mandatory and that neither its presence nor its absence is a correctable
> error according to the current rules. [It might be questionable however
> whether the acceptability of two spellings is ideal...].
 Ex. flumen-viridis, sancti-johannis, sanctae-ritae.
>
> When a geographical name is not Latinized but taken from its original
> language and only given a Latin ending it is considered as one word and
> the presence of a hyphen is a correctable error, Ex. rioverdensis,
> sanjuanensis, santaritana.

> According to this the epithet costaricensis (the Spanish name Costa Rica
> plus a Latin ending 'ensis'), has to be written without hyphen, hence
> costa-ricensis is a correctable error (60.9. Ex.20). When the epithet
> would consist of the Latin words for coast and rich it would result in
> ora-dives in which a hyphen is  to be accepted.

> p.s.: In IPNI there are 355 entries with rio- epithets. Many were written
 with hyphen but I've noticed that some 6 hours after I launched this matter
on TAXACOM, the hyphens have been removed from all rio- and san- epithets in 
the IPNI database.

***
Well, I have reservations on two points. The "... taken from its original
language and only given a Latin ending" is not universally applicable. If
this were to be applied to personal names it might be argued that if Otto
Schulz is to be commemorated in a two-word epithet this must become
ottoschulzii (not otto-schulzii). However, Otto is a word accepted in Latin,
and there is an epithet ottonis-schulzii (both words taking the genitive) to
prove it. Clearly Otto is to be accepted as a distinct word, so in
otto-huberi the hyphen is to be accepted also. This makes a hyphen
acceptable in epithets consisting of a given name and a surname. There may
be more exceptions ...

A second reservation I have is that I lack hard data: how many epithets
(i.e. what proportion) were originally published with rio- and san-? How has
the hyphen fared in usage in such cases?

Just as the fact that effective 10 February 2007 all epithets in IPNI
starting with rio were written without a hyphen does not prove that the
original spelling was without a hyphen, so equally the fact that a few hours
earlier many such epithets did use a hyphen does not prove these were
originally published with a hyphen.

As an aside: this change in IPNI raises the specter of "ruling by database".
To the world at large reality is that what can be found in an authoritatve
database. A few keystrokes can alter the content of a database. Just as to
many web inhabitants the website kept by Peter Stevens represents THE ONE
AND ONLY APG, so TROPICOS and IPNI represent the ultimate truth about
spelling to many web inhabitants. Certainly, this once again stresses the
importance of databases consistently recording the original spelling in all
cases ("as 'rio-negrensis' ").

Paul