Serbo-Croatian continuum: the top level

John Cowan cowan at
Mon Mar 3 18:31:32 CET 2014

Michael Everson scripsit:

> So? If a text is in Kajkavian, then the subtag will apply. Otherwise
> it will not.

The point is that Kajkavian is spoken in Croatia, but that does not make
it a variety of Standard Croatian, any more than Occitan is a variety
of Standard French.  The _Abstand_ is substantial.  Note that we do not
tag fr-occitan, much less fr-breton: we treat them as separate languages
as they deserve to be.

> Why do we have to “do” anything about “neo-Shtokavian”? 

You wanted a comprehensive tagging scheme for the SCC.  Neo-Shtokavian is
the most widely spoken variety of it.  Not all those who speak neo-Shto.,
speak the standard languages.

> Why would this be tempting? If we registered a three-letter code,
> the RA could assign those three letters to something else later.

We would assign a 5-to-8 letter subtag, as we are permitted to do
if the RA has refused to assign a 3-letter subtag.

> This brings to mind another question though. Do we have the power to
> create our own primary script tags?


> > The worst case is that we need five tags, for Kajkavian, Chakavian,
> > neo-Shtokavian, palaeo-Shtokavian, and Torlakian.
> Why do we need five? 

That's how many top-level varieties of the SCC there are.  Or at any
rate there are no more than that.

> Where is [the distinction between neo- and palaeo-Shtokavian]
> written up?

See <> and the
references attached there.

XQuery Blueberry DOM                            John Cowan
Entity parser dot-com                           cowan at
    Abstract schemata                 
    XPointer errata
Infoset Unicode BOM                                 --Richard Tobin

More information about the Ietf-languages mailing list