At 10:19 AM 11/5/2002 -0800, you wrote:

>For example, we could have a book in English but have tittles in different 
>    <title lang="en">Good Morning, New York</title>
>    <title lang="fr" type="translated">Bonjour New York</title>
>    <title lang="zh" type="alternative">早上好,纽约</title>
>     (do not worry, if your email system cannot display Chinese character 
> above)
>    <title lang="zh" type="transliteration">zhao shang hao, New York</title>

My understanding is that within Unicode you can detect the Unicode "range" 
and determine which script area any given byte belongs to. That is useful 
for individual scripts (Greek, Hangul, etc.) but the Western, Latin-based 
languages don't distinguish themselves in that way. So the marking, above, 
of languages like French and English could be useful, and I really like the 
making of the transliterated field, with both the language it represents 
(Chinese) plus the transliteration, although I would add the 
transliteration scheme to it because it may change how you process or 
interpret the data.

First let me say that I would like to see the transliterated field related 
structurally to the field for the original language, so we would have 
something like:

<title type="original">Original title</title>
<title type="transliterated" scheme="pinyin">Transliterated title</title>
<title type="translated" lang="fr">Translated title</title>

This is what we cannot do easily in the flat structure of MARC but we can 
do in XML, which is to create relationships between fields.

That said, as we've discussed either here or elsewhere (and I lose track of 
these things), it is often very hard to determine the language of the 
original title. Book titles can pull words from any language they like, and 
can even use characters from other scripts. I think we can define the 
language of any *created* titles (translations, transliterations), however, 
and that in itself would be valuable.

>The languagetype elements of MODS is to identify the language used for the 
>entire book/journal, it should allow any cataloger to use their native 
>language to describe the object, therefore we need to allow different 
>language to be used at element level.

Yes, we should distinguish between the language of the item being described 
and the language of the cataloging. So when a librarian in China describes 
a book published in England, the book is cataloged using Chinese, although 
the book itself is in English. As more metadata crosses national and 
language boundaries, this will be an important piece of information about 
the metadata itself.

Karen Coyle           [log in to unmask]