Antoine Leca joked:
> Neither you nor I would accept that our national language are tagged,
> respectively, la-ital and la-fran... ;-)
> Similarly, I believe Norwegians and Danes will not accept to
> have their
> present 2-letter codes replaced with cascaded ones in the form
> "Norse"-n? or "Norse"-da (replace "Norse" with the adequate code for
> the common root language of both language). And again and again.
> Also, what is the root language of Urdu? Is it Sanskrit, or
> Arabic, or Persian?
> Can of worms, can of worms...
OK, then we are playing paradoxes?
So, if go *walking* (not driving or flying) from Palermo through Rome,
Florence, Milan, Marseille, Paris, Lyon, Barcelona, Madrid, Valladolid, to
arrive in Lisbon, and chat with all the people you meet on your road.
At the end of your long journey you'd be totally unable to tell how and
where "it" became "fr", then "ca", then "es", to end up with "pt".
If you consider this, language tags should look like the RGB(x, y, z)
notation for colors. For a (oversimplified!) representation of Romance
dialects you'd need at least a CGA(x, y, z) notation to specify the quote of
Celtic, Germanic and Arabic components that blend with the Latin base to
form each village's speech...
But we all know that determining which language is spoken where is the task
of politics, not of linguistics; and Real numbers are too much for the
culture of the average politician. So, counting with Integer numbers,
Serbo-Croatian (or Hindi-Urdu) are either 1 or 2 languages. :-)
This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:21:13 EDT