John Cowan cowan at
Wed Jun 16 19:32:56 CEST 2010

Leif Halvard Silli scripsit:

> So, if we rule out that 'sh' is a macrolanguage, according to your and
> mine perception of 'macrolanguage' (which may or may not be in tune
> with Paul and John's understanding of 'macrolanguage'), then what does
> 'sh' represent now - versus then?

"Macrolanguage" is not a Platonic category, but a rough and ready
accommodation to the design of different coding standards, which reflect
the different needs of various groups (very roughly, bibliographers
vs. linguists).

> The very term 'macrolanguage' is used about languages that can be very
> distant. My point is that the tag we can ignore what Serbo-Croat has
> been used about and whether it it was correct or not. The important
> thing is that the language tag - sh - has never been used about anything
> other than Neo-Shtokavian. At the very least, this is the truth about
> the Language Subtag Registry.

Nobody knows that.  'Sh' and its equivalent 'hbs' may well have been
used for recordings or transcriptions of speech in other dialects;
it would have been, and (I argue) still is, rational to so use it,
there being no better tag available.

> So, there we can identify a proposal for 'sh': Add 'Neo-Shtokavian'
> as a name synonymous with Serbo-Croatian.

That would be tantamount to narrowing the semantic scope of a subtag,
which we are not allowed to do.

John Cowan    cowan at
The whole of Gaul is quartered into three halves.
        --Julius Caesar

More information about the Ietf-languages mailing list