[Apologies if you got this, but it seems to be bouncing and so I'm sending
it again.]
On 09/12/2000 10:00:34 AM "Christopher J. Fynn" wrote:
[I just got this today. Email's not supposed to take 9 days to arrive... ]
>I think a clear distinction may need to be made between those languages
which
>are commonly written and those which are (largely) only spoken. Outside
the
>realm of specialised applications for linguists, most applications
currently
>only deal with written languages and scripts and it is only confusing (and
>storing up problems) to add codes for spoken languages and dialects to
that
>list of tags.
>
>It is quite easy to envision that a set of standard codes may also be
needed
>for spoken languages and dialects in for things like voice recognition
>applications and specialised linguistic tagging. My own feeling is that
there
>should be a separate set of codes for *spoken* languages and major
dialects.
This sounds like you are in agreement, then, with our particular suggestion
that alternate sub-namespaces* within a system should be allowed.
*Before now, I was saying "alternate namespaces", but yesterday Harald
Alverstrand made the comment that RFC 1766 defines *one* namespace. I
responded that it already makes use of at least three namespaces, but in
thinking further I realise that it is probably more helpful in the
discussion to say that RFC 1766 uses a single namespace composed of
distinct sub-namespaces. One aspect of what we proposed in our paper is
that additional sub-namespaces be allowed.
>Obviously may languages would fall in both lists.
Certainly, and in those cases, only one tag should be used. If it's in ISO
639-x, then that should be used.
>Looking over the Ethnolouge codes for "Bodhic" languages it seems quite
clear
>that most of the codes listed are for distinctive spoken languages and
>dialects - literate speakers of most of these languages have one
>written/literary language "Tibetan" which they share in common - though if
>they tried to speak to each other they might have a great deal of
difficulty
>understanding each other.
Exactly.
>Putting it simply we need one code "bo" for the written language but many
>distinct codes for the spoken languages and dialects which are often
*very*
>different from the common spoken language.
Our very point (apologies if I haven't adequately communicated it before
now).
>In short I favour inclusion of codes for written languages in the
Ethnolouge
>list which are currently missing in ISO 639 (and the requirement for a
certain
>number of publications does not seem too onerous) - but do not favour the
>adoption of all the languages in the Ethnolouge list wholesale
>at this time as many of these appear to be only spoken languages or
distinct
>dialects.
>
>I do think it would be useful to consider a separate set of codes for
spoken
>languages.
The goes into a level of implementation that is beyond what we discussed in
our paper - we didn't say anything about adding to ISO 639, but we did talk
about two options in relation to RFC 1766, one of which assumes that
adoption of our suggestion for additional namespaces, which you appear to
be agreeing with. So, I think, at this point, we are in complete agreement.
- Peter
---------------------------------------------------------------------------
Peter Constable
Non-Roman Script Initiative, SIL International
7500 W. Camp Wisdom Rd., Dallas, TX 75236, USA
Tel: +1 972 708 7485
E-mail: <[EMAIL PROTECTED]>