[Ltru] Alemanic & Swiss German

Peter Constable petercon at microsoft.com
Wed Dec 6 17:43:21 CET 2006

> From: Martin Duerst [mailto:duerst at it.aoyama.ac.jp]

> In my understanding, gsw as described in the Ethnologue refers
> to "Alemanic" in the narrow sense. Alemanic can be used both for
> a wider group that includes Swabian (Schwa"bisch), as well as
> for a narrower group that doesn't include Swabian. See e.g.
> http://de.wikipedia.org/wiki/Schwaben#Schwaben_und_Alemannen. 
> This is close to what Peter suggested in one of his mails.

Just so.

Btw, I went to that Wikipedia page, and since I can't read German, I followed the link to the corresponding English article. Now, unfortunately, the English article is a pretty different article and doesn't have a section corresponding to " Schwaben und Alemannen", but I did note this: if you're looking at the English article "Swabia", then one of the other-language links is "Alemannisch": http://als.wikipedia.org/wiki/Schwaben 

Now, the linguistic variety used in that page is coherent -- one language -- and not from some collection of languages. I presume this content would appropriately be tagged "gsw".

> So if a correction is necessary, it should be changing
> "Alemanic" to "Alemanic (in the narrow sense)" or something
> similar.

(Well, probably with two 'n's, right?)

The problem is how that should be conveyed e.g. in ISO 639-2 in which all that has ever been listed are names.

> >The term "Alemanic" in English ("Alemanisch" in German) encompasses a wider
> scope (eg including Swabian, which is clearly not Swiss German*), and should not be
> used.
> No. Using the word "Swiss German" as the label for something that
> covers parts of Austria, France, and Germany is clearly wrong.
> I'll explain it in more detail in an answer to a mail from Peter.

I believe Mark was misled by the fact that one of the machine-readable (draft) code tables for ISO 639-3 (the file for the code set) had "Alemanic" as the reference name for "gsw", which to Mark suggested the collection (Alemannic in the broader scope) rather than the (posited) individual language (Alemannic in the narrow scope). To the extent that the reference name in ISO 639-3 has a somewhat similar effect as the representative glyph in Unicode -- it's a one-shot impression -- then one might contend that, for English speakers, "Swiss German" gives a better first-pass impression of what's intended than does "Alemanic". (Of course, it's important to note that the reference name is not the whole story, just as Unicode's representative glyph is not the whole story.)

Probably the best solution -- and I think Mark indicated he was OK with this -- would be to change the reference name in ISO 639-3 to "Swiss German", list "Alemannic" as another name, and include a note (in the entry in the LST registry) clarifying that "Alemannic" here is meant in the narrow sense.

Peter Constable

More information about the Ietf-languages mailing list