Re: OT: Origin of Catalan reinforced weak pronouns

Eric Christopherson wrote:
>Does anyone know where the Catalan reinforced weak pronouns come  from?
>Many of them appear to be the reverse of the corresponding  plain form weak
>pronouns, e.g. plain forms me/te/se : reinforced  forms em/et/es. Did these
>come about via metathesis, or by some other  route (e.g. by prefixation of
>e- followed by deletion of final -e)?
Can't comment knowledgeably on this, but I'd suspect it might have something
to do (historically at least) with environment--- ...V#me C.... vs. ...C#me
V... > [vowel insertion/deletion rules] ...C#em V...
>
>Also tangentially related is something I've wondered about for a  while: do
>linguists make a distinction between the *process* of  metathesis and *end
>results* which appear to show metathesis (whether  in fact those results
>came about through said process)?
I do, at least. I think true metathesis is sporadic-- cf. Engl. wart :: Du.
wrat, or OE hros :: later horse; there may be some phonological motivation
("difficult" clusters??). Spanish speakers are _said to_ have difficulty
with the word "atlántico", mispronouncing it as "altántico" though I've
never heard that........

Then there's a non-sporadic kind. At least two widely separated Austronesian
langs. have grammatically motivated metathesis-- Rotuman (in the Pacific)
and Timorese (in Indonesia); in these, CVCV bases > CVVC in various
environments, often with changes to the nature of the VV cluster.  This is
not well understood, to say the least :-)))

Another group of Indonesian languages (spoken near Timor and eastward) have
what I've termed "pseudo-metathesis", where bases may have one of two forms,
e.g. Leti ulti ~ulit 'skin'. Historically this has an explanation: canonic
AN shape is CVCV(C), and original final C were preserved  by adding an echo
vowel, so *kulit > kúlit-i, and the post-tonic original final-syl. V was
then deleted, > **kúlti. This could well have started with "fast-speech"
rules. Other workers in this area seem to have accepted my historical
explanation, but still go to great and complicated lengths to explain it
synchronically.  I guess it depends on how willing one is to accept the
"reality" of underlying forms, and the persistence of historical rules.

Another type of met. occurs in these langs., whereby compounds,
nominalizations and conjugated verbs are "bound" together by metathesis:
Leti compd. /pipi+duma/ pipdiuma 'sheep'
Leti noml. sora 'sew' + infix -in- > sniora 'needle'
Leti vb. au 'I' + laa 'go" > aluaa 'I go'
(the i/u reduce to [j, w] respectively)

These can be explained with rules of anticipatory assimilation of the i/u
element (other vowels do not take part). And probably (originally) another
form of fast-speech phenomena.