[Ltru] Collection tags considered problematic (was: "mis" update review request)

[John Cowan]

Mark Davis scripsit:

> What is truly unfortunate is that ISO defined these (Other) items
> by exclusion. If they had defined the tag "ger" to simply meant
> Germanic, it would remain valid indefinitely. They would then behave
> like macrolanguages.

Well, unless a language were to be reclassified.  Very unlikely for
Germanic, but not so unlikely for some of the others.

In any case, the "(Other)" is, IIRC, in the process of being removed.

> In LTRU what we need to decide is whether we bail on stability, or
> we impose stability. We can do the latter by deciding, like we do
> with other unstable ISO codes, that the meaning of the tag "ger" is
> set at the time we add it to the registry, and can only be narrowed,
> not broadened.

The trouble is that we don't really know the extensional definition
of these collective language tags:  they are defined purely intensionally.
So they change when the literature says they change.

> Then if ISO decides to add a new code for a Germanic language "gxx",
> it then simply overlaps with "ger". It is then a strong SHOULD to use
> "gxx" in preference to "ger" where it applies, but it doesn't invalidate
> previous usage.

Again, what happens if a language is found *not* to belong to some
collection in which it had been included until now?  How do we even know
about such an event?

-- 
John Cowan                                cowan at ccil.org
I amar prestar aen, han mathon ne nen,    http://www.ccil.org/~cowan
han mathon ne chae, a han noston ne 'wilith.  --Galadriel, LOTR:FOTR