Language for taxonomic names, redux

Doug Ewell doug at
Mon Feb 27 01:32:03 CET 2017

Arthur Reutenauer wrote:

>> [...] Aquitanian, like some other ancient languages, is known only
>> from a few hundred proper names embedded in texts written in another
>> language (Latin, in fact), and this is probably all the Aquitanian
>> there will ever be, yet we allow tagging these names with "xaq".
>> [...]
> But Aquitanian is not just that list of words, it is an actual
> language that is known today only through a few hundred words recorded
> in Latin-language sources.  There is no doubt that it was indeed a
> full-fledged language, even if its relationship to proto-Basque, or
> its status as a distinct language, is not fully understood (and may
> elude us forever).

This had occurred to me too. Evidently Linguist List, and therefore ISO 
639-3/RA, had decided that Aquitanian was a real language, and that's 
why it's in the Registry.

And that brings up the real difference to me, which is that when it 
comes to variants, we (the Reviewer and this list) are the arbiters of 
what qualifies and what does not; but when it comes to languages and 
regions and scripts, our job is to take what is on the standard code 
lists and adjust the Registry accordingly, deviating only in cases of 
conflict or code re-use. We don't pick and choose from among the ISO 
code elements; only when a code is truly "special" in the core standard 
(like [mis] or [Zxxx]) do we even attach a modified or secondary meaning 
to it.

So when we allow tagging of a list of names with "xaq", it's not because 
we decided the list of names was good enough, but because the RA did. 
For variants, on the other hand, It is up to us to decide what evidence 
and how much is sufficient for variants, irrespective of whether another 
body would have reached the same conclusion.

Doug Ewell | Thornton, CO, US | 

More information about the Ietf-languages mailing list