Sten's message brings up a very important point: that the JAC needs to
carefully consider how it will deal with dialects. Undoubtedly we will
need to deal with many of these in the future (in fact, one other
discussed at our meeting was Alsatian). We noted in the meeting that it
was difficult to have a hard and fast rule about when a dialect is a
language, and that each would need to be considered on a case-by-case
basis. Before we commence with voting on Low German/Saxon we need to
There are two issues to consider.
1) We need to establish some guidelines to help us decide what should be
defined as a language. Millie Wewerka has kindly submitted some below.
2) If the dialect in question does not satisfy the criteria, how shall we
deal with it?
I will separate these two issues into two messages.
These are suggestions for evaluating the merit of including a separate
language code for a dialect.
1) Does the required number of documents exist? (This is, of course,
2) Does the dialect have a distinctive name? A name such as Parisian
French isn't particularly distinctive, being based on the combination of a
place name with the name of the language.
3) Does the dialect have a degree of standardization, such as a consistent
4) Is the dialect taught separately in schools?
5) Is there significant variation from the standard language?
6) Is the dialect sometimes considered a separate language?
7) Is there a tradition of literary usage or is the dialect used in media
such as radio, motion pictures, television?
The more "yes" answers that occur for a dialect, the more likely it is
that a separate code would be useful. (Would we want to say that more than
half of these need to be true?)
On Mon, 17 Apr 2000, Hedberg Sten wrote:
> It is with increasing respect and wonder that I have followed the last days'
> discussion of how to resolve the matter of the code for Plattdeutsch, a very
> realistic and sound proposal. I understand we will get material for a formal
> vote from Rebecca next week.
> We have not decided, within 639-2, on how to handle dialects in a more
> general way, although there are cases when a dialect is recognized in its
> own right (criterion 1.2). Furthermore, the only rule for the code is to
> base it on vernacular forms (crit. 2). Thus, I am surprised to see
> connections to various German- or Saxon-based local names for something that
> must be regarded as dialects brought forward to support "ndd" or similar
> codes. My second surprise is to see the amount of comparative linguistics
> brought forward.
> John's reminder of UKMarc "gml" for Low German was very comforting, as was
> his parallels with ger - gmh - goh - ...in 639-2B.
> As I see it, "gml" is, for the reasons John brings up, the only real
> alternative to the fundamental vernacular "pld" from Rebecca. To improve
> "gml", it might be argued to take away the "m", meaning "Middle" in "gmh",
> and establish the code "gel" meaning "German, Low" regardless of time. But
> the examples given in the proposal were, if I remember correctly, mainly
> fetched from early material (but, admittedly, not from medieval but from
> 16th and 17th centuries).
> Will the vote be more on what code than on the necessity to have a code? It
> seems so. Preferrably, we would need to get a final suggestion of two codes
> to choose between. Any more alternatives would be a nuisance to assess.
> Sten Hedberg
> Expert on cataloguing and standardisation
> Uppsala University Library
> POBox 510, S-751 20 UPPSALA Sweden
> Voice +46-(0)18 471 3970
> Fax +46-(0)18 471 3941