> -----Original Message-----
> From: Constructed Languages List [mailto:[log in to unmask]]On
> Behalf Of Nik Taylor
> Sent: Friday, November 5, 1999 6:04 PM
> To: Multiple recipients of list CONLANG
> Subject: Re: Allophony
> Rob Nierse wrote:
> > What is more likely?
> > /ti/ => [tsi] ot [ci]?
> Both are equally likely.  Consider Japanese, where /ti/ became [tSi],
> while in Spanish at one stage, /tj/ became /tsj/, as in /na'tjon/ -->
> /na'tsjon/ --> /nasjon/ or /naTjon/ (nacio'n)

Speaking of Japanese, does anyone know how /tu/ became [tsM] ([M] is an
unrounded [u])? I had thought maybe the /M/ was originally a front vowel,
but then how does one account for the presence of /i/?

As for Spanish, the change /tj/ > /tsj/ > /ts/ was universal in the Romance
languages, and apparently happened well before they broke apart. However, I
think the word <nacio'n> is a learned borrowing from Latin; words which
proceeded normally from Latin without learned influence seem to have lost
the [j] in [tsj], thus:

Latin <rationem> ("reason") /ra"tsjone/ > /ra"dzjone/ > /ra"dzon/ > /ra"zon/
> /ra"son/ (and in Castille, > /ra"Ton/) <razo'n>.

I just read that changes very similar to the ones in Romance also happened
in Greek prehistoricly, including the change in some dialects of /tj/ to