Macrolanguage (was: Forms for Montenegrin)

I was asked, offline, why the proposed form to add Montenegrin doesn't
include the field "Macrolanguage: sh" when the record for Serbian (and
Bosnian and Croatian) does.

Montenegrin is the same as Serbian, right? Even more so than Bosnian or
Croatian? So if Serbian has Serbo-Croatian as its macrolanguage, why
wouldn't we add that to Montenegrin as well?

"Macrolanguage" is completely an ISO 639-3 concept. The RA invented the
classification, and only they can decide which languages are
macrolanguages and which individual languages are encompassed by them.
We don't make those decisions. The Language Subtag Registry includes the
ISO 639-3 macrolanguage structure to help users choose language subtags,
but we do not determine the macrolanguage or encompassed-language
classifications.

Not only is this the law (RFC 5646, Section 3.1.10), it's also a good
idea. If we were able to apply our judgment to this case, we would open
the door to endless debates about other macro/encompassed language
decisions, and we would essentially be wrestling this responsibility
away from 639-3/RA. I for one wouldn't want to get embroiled in this,
especially considering how many people don't understand what a
macrolanguage really is.

As a side note, Montenegrin hasn't been added yet to 639-3, with its
annual cycle, but only to 639-2. It's likely that when 639-3 adds it,
they will make it an encompassed language of Serbo-Croatian, but that is
their decision and we can't jump the gun on it. 
 
--
Doug Ewell | Thornton, CO, US | ewellic.org