[Ltru] Alemanic & Swiss German

[Martin Duerst]

At 03:10 06/12/02, Peter Constable wrote:
>Content-Class: urn:content-classes:message
>Content-Type: multipart/alternative;boundary="----_=_NextPart_001_01C71574.0CBEED23"
>
>The claim in Ethnologue is that the language $Bc`W4(Bchwyzerd$B%F%7(Btsch$Bc`Y
(B(aka $Bc`W4(Bwiss German$Bc`Y
(B is also spoken in Austria, France, Germany and Liechtenstein although the name $Bc`W4(Bchwyzerd$B%F%7(Btsch$Bc`Y
(Bis not generally used in those places; 

A definition like the above "'gsw' is the language named X
although also spoken in Y but not named X in that case"
is contradictory and doesn't make sense.

Schwyzerdu"tsch is not spoken in Austria, France, or Germany,
by definition (except by people who moved there from Switzerland).
[I don't want to say anything about Liechtenstein; if anybody
from Liechtenstein is on the list, please speak up.]

>that the name $Bc`W"(Bllemanisch$Bc`Y
(Bis sometimes used in reference to this language; and further, that this language is distinct from languages known as $Bc`W4(Bwabian$Bc`Y
(Band $Bc`W8(Balser$Bc`Y(B 
> 
>That is the semantic that ISO 639-3 and ISO 639-2 assume. I think this is consistent with your understanding and usage.
> 
>Now, I hear Mark saying that the name $Bc`W"(Blemanic$Bc`Y
(Bis not used to refer to Swiss German / Schwyzerd$B%F%7(Btsch but rather is only used to refer to a range of varieties of broader scope (a genetic sub-node, perhaps) of which Swiss German is but one language. (Ethnologue shows Schwyzerd$B%F%7(Btsch as being classified genetically in a branch of Germanic called $Bc`W"(Blemannic$Bc`Y(B) The ISO 639 identifier $Bc`WH(Bsw$Bc`Y
(Bis *not* intended to refer to such a collection of languages.

'language' or 'languages' are obviously in the eye of the beholder.
I assume it's linguistically possible to make a group or language
(whatever you want to call it) from Swiss German (including
the Wallis and the Walser dialects in Graubu"nden, as mentioned
in the Ethnologue page ("Graubenden-Grisons (Valserisch), Wallis";
btw 'Graubenden' is a missspelling of 'Graubu"nden'),
Elsa"ssisch (Alsace, France), the Voralberg Region of Austria,
and some pieces of southern/southwestern German, and to label
it gsw. That's consistent with the data in the Ethnologue, and
with the usage pattern that Karen is proposing.

[However, looking at http://www.ethnologue.com/show_language.asp?code=wae,
the Ethnologue seems to count Walser dialects in Graubu"nden and
probably also Wallis twice.]


>The only debate I see here is whether $Bc`W"(Blemanic$Bc`Y
(Bis ever used to refer to the specific language denoted by $Bc`WH(Bsw$Bc`Y(B

Looking at http://www.ethnologue.com/show_family.asp?subid=92157,
that seems to more or less make sense for Alemanic.


>On the one hand, we see Ethnologue use $Bc`W"(Blemannic$Bc`Y
(Bin reference to the genetic subnode. On the other, we see Ethnologue indicate that $Bc`W"(Blemannisch$Bc`Y
(Bis a name used to refer to the specific language

"also" -> "not actually"

>known as $Bc`W4(Bchwyzerd$B%F%7(Btsch$Bc`Y
(Bor $Bc`W4(Bwiss German$Bc`Y(B


>It appears to me that Mark$Bc`QT(B comments amount to one of the following:
> 
>-         questioning the claim that $Bc`W"(Blemannisch$Bc`Y
(Bis used as a name for the specific language denoted by $Bc`WH(Bsw$Bc`Y
(B or
> 
>-         a request to clarify that, in listing $Bc`W"(Blemanic$Bc`Y
(Bas an English name for $Bc`WH(Bsw$Bc`Y
(B the intended denotation does *not* encompass Alemannic languages other than the one known as $Bc`W4(Bchwyzerd$B%F%7(Btsch$Bc`Y
(Bor $Bc`W4(Bwiss German$Bc`Y
(B(and perhaps other names)

No, at least from my side, the request is not to use the name
Schwyzerdu"tsch/Swiss German for something it doesn't refer to.
If you want to refer to what the Ethnologue labels as 'gsw',
call it "high alemanic" or whatever you want, but not Swiss German.

>If it is the case that $Bc`W"(Blemannisch$Bc`Y
(Bis used in some places to refer to the specific, individual language denoted by $Bc`WH(Bsw$Bc`Y
(B

"specific, individual language" is in the eye of the beholder.
Linguistically, 'gsw' may qualify as a 'language' according to
some criteria, but that's not how people think. It's much more
like a vast area with a common language background, where you
can find differences more or less between every two neighbour
villages you look at. What researchers in Switzerland do is
not to try to cut up that continuum into pieces, but to document
where the boundaries are. I.e. you take a traditional concept,
and draw a map to show where one word is used and where another,
or you take a certain syllable or ending and draw a map with
a boundary line between different forms. 

>then this is a good example for a problem that Gary Simons and I identified in a paper back in 2000: the problem of clearly documenting what is the intended denotation of a given language identifier.

Yes. At least in the case at hand, where you want to create or
identify subsections of a language continuum such as German/Germanic,
the best way to describe what you mean would be not in terms of
regions (which of course can help), but in terms of grammatical
or vocabulary features. As an example, Austians east of the Arlberg
pass call their countrymen west of the Arlberg (Vorarlberg)
"Gsiberger", because they use "gsi" as the past participle of
the verb "to be", whereas the rest of the Austians use some
form of "gewesen", similar to standard German.

So at least with respect to Austria, this could be listed as
a criterion (probably one of several) of what distinguishes
gsw from other German dialects/variants in Austria.


>For, if $Bc`W"(Blemannisch$Bc`Y
(Bis used in this way, then it would seem that $Bc`W"(Blemanic$Bc`Y
(Bmight actually be used in two different senses: one referring to an individual language, and one referring to a collection of languages corresponding to a genetic subnode of Germanic. But, $Bc`WH(Bsw$Bc`Y
(Bdoes not denote either of those semantics according to the whims of a given user; it denotes just one of those semantics, viz. the former. 
> 
>(This is also a good example to show that the objects of codification are not merely names but rather specific concepts that are referenced by a given name $Bc`E
(Bterms, not lexemes.)
> 
>So, then, how can the intended semantics of the identifier $Bc`WH(Bsw$Bc`Y
(Bbe clearly documented in ISO 639?

See above.


>In the past, for parts 1 and 2, all that was provided was a list of names. In developing part 3, the RA and I both felt that this was not adequate. Formally, we could not include the various kinds of encyclopedic information that might actually be needed to make it clear what the intended denotation of a given identifier actually is (machine-readable tables need to have a well-formed structure with clear semantics for each data category). What we could add formally are attributes, of which we added two: scope and type. So, the data tables for 639-3 show in particular that $Bc`WH(Bsw$Bc`Y
(Bhas a scope of individual language. That at least makes clear that the intended semantic for $Bc`WH(Bsw$Bc`Y
(Bis not a collection of languages. 

On the ground, that doesn't help at all, because what's an individual
language and what not is in the eye of the classifier. While I don't
want to criticise the classification of the Ethnologue, others are
certainly possible.


>But even that is not sufficient: perhaps the range of language varieties under the Germanic/../Alemannic classification are being treated as a single, individual language. 

The 'perhaps' is probably true, but still as a 'perhaps'.
The fact is that people's minds are much more flexible than
classifications; people can speak about German (fairly high
up in your hierarchy) as a language in one sentence, and
can speak about the dialect/variant used in their village
as a language in the next sentence (and understand each
other very well, too).


>To overcome such limitations and to provide clear documentation of the intended denotation for each identifier, the ISO 639-3 Web site gives, for each entry, a $Bc`WN(Bore$Bc`%rc`Y
(Blink to a page that shows additional information for the given identifier, and that additional info includes not only the formal attributes of scope and type but also links to external sources, including sources of an encyclopedic nature such as Ethnologue, specifically for the purpose of documenting the denotation.

That's good. But only in as far as those sources are clear and correct.
Using "Swiss German" for anything other than the German dialects spoken
in Switzerland, and only in Switzerland, means just misusing a term
and confusing more than documenting.


>So, if you want to know an ID in ISO 639-3 is intended to mean, don$Bc`QU(B just look at the list of names associated with that ID since names alone cannot guarantee that the intended semantic is clearly communicated. Rather, go to the ISO 639-3 Web site, get a report listing the given ID, follow the $Bc`WN(Bore$Bc`%rc`Y
(Blink to the documentation for that particular ID and look at *all* the information provided there, including the links to external sources. (You can get to these documentation pages directly using URLs such as the following example, for $Bc`WH(Bsw$Bc`Y(B <http://www.sil.org/iso639-3/documentation.asp?id=gsw>http://www.sil.org/iso639-3/documentation.asp?id=gsw.)

Well, yes, except that this should be fixed.

Regards,    Martin.



#-#-#  Martin J. Du"rst, Assoc. Professor, Aoyama Gakuin University
#-#-#  http://www.sw.it.aoyama.ac.jp       mailto:duerst at it.aoyama.ac.jp