On 5/17/07, Eric Christopherson <[log in to unmask]> wrote:
> On May 17, 2007, at 7:40 PM, Alex Fink wrote:
> > On Thu, 17 May 2007 19:15:02 -0500, Eric Christopherson
> > <[log in to unmask]>
> > wrote:
> >> Is that a mainstream interpretation of the allophones? As I
> >> understand it, etymologically, "se lo" etc. come from ILLI ILLUM etc.
> >> by mostly regular sound changes, without avoidance of cacophony as a
> >> "motivation". On the other hand, if at some point both *"le lo" and
> >> "se lo" were available to speakers, they might commonly choose "se
> >> lo" for that reason.
> >
> > Just sound change?  That's surprising.  How does ILLI ILLUM
> > generate an /s/?
>
> Roughly like this:
> ILLI ILLUM [eljelo] > [ljelo] > [Lelo] > gelo [Zelo] > [Selo] > se lo
> [selo]
>
> Normally, of course, [S] > [x] instead of [s]. I suppose maybe the
> [s] outcome was influenced by the existence of the pronoun "se".
> (It's been a while since I read about the development of Spanish. I
> probably should have consulted my books before asking here, but oh
> well.)
>

Ralph Penny says that it the substitution of gelo by se lo is partly
phonological between /S/ (</S/, /Z/) and /s/, and also syntactical. Se
lo already existed in Old Spanish and Middle Spanish. He says that the
se in se lo was quasi-reflexive. He gives two examples:

Su amigo se lo tomó - His friend took it (for himself)

Su amigo gelo tomó - His friend took it from him.


In other words, the two constructions had close enough meanings, along
with confusion of phonemes to cause "se lo" to replace gelo entirely.