Re: rough sketch of a potential solution

I fear my first reply was too flippant, so let me try again:

1. the ability to send a message in multiple languages to multiple 
recipients may be valuable even if it is not used today very often.  
value is not necessarily related to frequency of use.

2. even if a message is sent in a language that all of the recipients 
can read, the senders and/or some of the recipients may have multiple 
spellings of names in various languages.  for maximum address 
recognition, transcribability, etc., each recipient would like to see 
the addresses of the senders and recipients in the recipient's native 
language/script, if possible, and if this is not possible then the 
recipient would like to see the addresses in a form which can be 
remembered or transcribed, at least.  The senders may not even know 
each of the recipients personally, and may not be aware of the 
recipients' preferences.  The senders may also not know how to spell 
recipient A's address in a format which recipient B can easily 
transcribe or remember.

This is, in my view, the real problem we need to solve.  The reason we 
want email addresses other than pure ASCII is so that recipients can 
recognize and transcribe our addresses in order to use them later.  For 
this to work in a world where people with different native languages 
correspond with one another, we need some way not only to represent 
different versions of an address in an email message but also a way to 
find out what other representations exist for an address.

>> The problem is that there may be multiple recipients and the sender 
>> may not know which recipients can read which languages, or even if 
>> this is known, the sender may not know alternative addresses for some 
>> of the recipients.  Consider a reply to a message that was originally 
>> sent to a list of recipients, some of whom the sender knows and some 
>> of whom the sender doesn't know.  Either the original message or the 
>> reply could be in multiple languages (say with multipart/alternative)
>
> Do you have information on how often multiple languages (with 
> multipart/
> alternative or plaintext, but nevertheless parallel texts) are used?
> My guess is that they are very rare. Preparing messages in multiple
> languages is everything but cheap.
>
> Regards,    Martin.