Leif Halvard Silli xn--mlform-iua at
Wed Jun 16 16:03:31 CEST 2010

Milos Rancic, Wed, 16 Jun 2010 15:19:55 +0200:
> On Wed, Jun 16, 2010 at 13:30, Leif Halvard Silli wrote:
>> John Cowan, Tue, 15 Jun 2010 15:29:55 -0400:
>>> ISO639-3 scripsit:
> Treating Serbo-Croatian as a macrolanguage is not quite good solution.
> The term is ambiguous, there is no such thing as "Common
> Serbo-Croatian language" in the genetic (not prescribed) sense; as
> well as the majority of domestic population really don't like that
> term because of various reasons.

'sh' *is* treated as macrolanguage, in the Language Subtag Registry:

Type: language
Subtag: sh
Description: Serbo-Croatian
Added: 2005-10-16
Scope: macrolanguage
Comments: sr, hr, bs are preferred for most modern uses

You and I agree that it is questionable whether it is correct to 
consider 'sh' a macrolangauge. However, we should not ask the 'domestic 
population' about whether this is correct, but instead concentrate on 
'macrolanguage' as understood by BCP47 etc - in combination with a 
understanding of what 'sh' historically has referred to.

> It is ambiguous term because it may mean two things: (1) standards
> based on previous Serbo-Croatian standard, when it includes just
> standards based on Neo-Shtokavian language system; and (2) all
> language systems from the former Serbo-Croatian diasystem, when it
> includes: Neo-Shtokavian, Old-Shtokavian (which is also ambiguous
> term; see below), Chakavian and Kaykavian.

That it is an ambiguous term is not an argument against the very 
existence of the language subtag 'sh'. 

> If it means (1), then it is much better to call it Neo-Shtokavian or
> Shtokavian. And it is hard to say that it is a "macrolanguage", it is
> one language system with four varieties, very comparable with Spanish
> standard varieties.
> If it means (2), it is largely a political construct, which doesn't
> have political support anymore. Before WWII, Western Macedonian
> dialects were treated as a part of Serbo-Croatian diasystem. Purely
> linguistically speaking, it is similarly today with Kaykavian,
> Chakavian and Torlakian (as a part of "Old-Shtokavian" dialects).
> However, there are no separate ethnic identity among them (the first
> two groups are treating themselves as Croats, the second group is
> treating themselves as Serbs or Bulgarians, depending on heredity).

To be clear: the disappeared political support that you talk about, is 
not the political construct that made us all consider 
Bosnian/Serbian/Croatian as one language, but the political support 
which tried to include more than that/those language(s) under the 
Serbo-Croatian umbrella.

I believe that the more constructed it is, the more is is correct to 
consider it a macrolanguage. Thus, if 'sh' has been used to refer to 
(2), then it is correct to continue of the Language Subtag Registry to 
say that it is a macrolanguage.

However, the reason why I question whether it is correct to consider 
'sh' a macrolanguage, is based on the understanding that 
'Serbo-Croatian' refers to (1) - the standardized Neo-Shtokavian 
form(s). It may still be correct consider it a macrolanguage - I don't 
know - but it doesn't fit my understanding of how 'macrolanguage' 
should be used.
> To conclude, all four standards (Serbian, Croatian, Bosnian and
> Montenegrin) have one language system (paradoxically, with Montenegrin
> as the most distant of those) and, linguistically speaking, it is not
> a macrolanguage, but a language.

Hence, the the Language Subtag Registry is currently incorrect, it 
seems to me.

>  However, it is highly unrealistic to
> move back them to "sh-RS" etc., which would be the most precise
> description of the variants (actually, something like "hbs-srp" would
> be the most precise; while, again, it is better to put group
> abbreviation based on (Neo-)Shtokavian).
> Thus, the most realistic approach is to allow Montenegrins to take
> their ISO 639 codes, conclude one part of the history and mistakes and
> put on some future agenda to solve this issue more appropriately than
> it has been solved during the past two decades.

Part of John's justification for a separate language subtag for 
Montenegrin, was - as I understood him - that 'sh' would then cover 
them all - as a macrolanguage.

I don't think that it is necessarily is necessary to keep 'sh' as a 
macrolanguage in order to support a separate tag for Montenegrin. My 
point here, was only and solely to question whether 'sh' is a 
leif halvard silli

More information about the Ietf-languages mailing list