Metadata alng

dc:language indicates "a language of the intellectual content of the 
resource". 

so you can specify language by

<meta name="dc:language" content="fr">



It does not say anything about the character set in use, e.g. I can 
write both English and French using the UTF-8 character set.

It is possible to indicate that a resource uses multiple languages by 
repeating dc:language, but in DC metadata descriptions there is no 
implied ordering of repeated properties, so if you have

<meta name="dc:language" content="fr">

<meta name="dc:language" content="en">

the two languages are treated equally.  SO that is not helpful if they 
have the same character set.
However if we have two languages on a page with different character sets 
we can say

<meta name="dc:language" content="fr">
<meta name="dc:language" content="he">

without any confusion and without requirement marking up every word from 
the second page language.

It will help WCAG adoption a lot in places like Israel if we do this

What do you think?
Lisa