Re: Spanish and Italian _r_ and _rr_ -- for my Romlang #3

John Quijada skrev:
 > Benct Philip Jonsson wrote: I know that Spanish and
 > Italian /4/ and /r/ contrast only between vowels;
 > elsewhere the single spelling _r_ represents both
 > phonenes, distributed according to rule. I *seem* to
 > remember that the rule is [r] before vowels and [4] after
 > vowels, but I'm not sure, especially not about what
 > applies word-finally. Also I don't know whether the same
 > rules apply in both languages.
 >> And what about Catalan, Provençal and Portuguese? I
 >> think that in the
 > first two only /r/ has become /R/ while /4/ remains, and
 > IIRC the same is the case in some dialects of Portuguese.
 > How is the actual case with this, and again what is the
 > distribution?
 > _________________________________________________
 > Grrr! Wrote a big long answer for you, Benct, only to have
 > it erase itself when I posted it. Let’s try again:

FYI there was some trouble when I was responding: everything
below "Let’s try again:" was deleted until I turned on the
"view as HTML" option.  There was no HTML, thankfully, but
then the auto-quoting mechanism suddenly worked without
deleting anything.

 >
 > I speak Spanish, Italian and Portuguese, have spent my
 > whole life hearing Spanish almost daily (I live in
 > California and watch Spanish-language television about
 > twice weekly), have spent a cumulative total of about ten
 > weeks in Italy, have spent three weeks in Portugal (the
 > southern half), lived for about two years in a San
 > Francisco Bay Area neighborhood containing a large number
 > of Brazilians who spoke the Carioca dialect, and worked
 > with two Brazilians who spoke the Paulista dialect.

Wow, that beats my 6 weeks in Italy, 3 weeks in Spain 7
weeks in France (all cumulated) and 5 days in Portugal by
far! (At one point I had forgotten the word _gabinetto_ when
I went into a church to borrow one. In my desperation I
asked the only person I met inside in Latin _Ubine Latrina
est?_, and he politely answered in German. Apparently a
blond person is German by default in Italy! :-)

 > So based on that, my observations about Spanish, Italian,
 > and Portuguese /4/ and /r/ are as follows:
 >
 > Spanish: Word-initial _r_ is always /r/, pronounced [r].
 > _r_ is also /r/ after /n/ and, in some dialects, after
 > /l/. After all other consonants, it is /4/. In word-final
 > position, it is /4/ in normal speech, but /r/ in emphatic
 > or hyper-enunciated speech.
 >
 > Italian is the same as for Spanish except I’m uncertain
 > about the /r/ pronunciation after /n/ and /l/, and in word-
 > final position (not common other than “per” and elided
 > infinitives like “far niente”) it is /4/, not /r/.

What about pre-consonantal? I see I forgot to ask about
that, but as there seems to be some variation in word-final
_r_ it is probably important.

As for the difference between the position after /n/ or
/l/ and after other consonants, it might be significant
that Italian has assimilation in loans like _Corrado_ <
German _Konrad_. I don't know if there is anything
comparable in Spanish.

 > Portuguese: distribution same as Spanish. Actual
 > pronunciation of the phonemes depends on the dialect: In
 > Lisbon-area Continental Portuguese, /4/ is [4] while /r/
 > is in free variation between [r] and [R], often in the
 > same speaker. I’ve yet to discern a predictable pattern
 > to the variation, even in the recordings of fado singers I
 > collect. In Carioca dialect /4/ is [4] and /r/ is [h] in
 > normal speech, but voiceless uvular [X] in emphatic or
 > hyper-enunciated speech. In Paulista dialect, /4/ is [r]
 > but /r/ seems to be [h] for some speakers and [r] for
 > others; when emphatic or hyper-enunciated, Paulistas
 > pronounce /r/ as
 > [r], not [X].

I'd heard about the Brazilian [r] > [R] > [X] > [h] before,
and find it kinda cool, and a quite interesting way for a
language to acquire /h/ and/or lose /r/. It is interesting
to see that it is not uniform in Brazilian, especially that
Paulistas have [h] but lack [R] or [X]!

There are Swedish dialects south of the apical-uvular /r/
isogloss which have loss not only of post-vocalic, but also
of pre-vocalic /r/. I have never seen if they have /h/ loss
too, though.

 >
 > I don’t speak much Catalan (although I read it
 > passably), but the three times I’ve been in Barcelona,
 > I’ve never heard any Catalan pronounce /r/ uvularly,
 > only as [r].

Ah, then my sources were in error, or more likely I confused
statements about Catalan and Provençal. Do you know
anything about the loss of post-vocalic _r_ in Catalan? Is
it geographically or socially localized?

I'm not sure yet what all this will lead to in R#3. I
guess some dialects will have an /4/ <> /r/ distinction
finally, while the standard won't. I keep being influenced
by those Swedish dialect which have /r/ as [R] when
initial or geminate, but [4] elsewhere, including a
distinction in final position, since /r/ and /rr/ contrast
there in Swedish.

In my own Swedish I seem to have moved from a [4] ~ [r]
allophony to a [4] ~ [r\_a] allophony in my lifetime, and I
still use [r] when enunciating. That shouldn't influence R#3
of course...

caeruleancentaur skrev:
 >> Spanish:
 > Two additional notes:
 >
 > 1) Word initial <r> /r/ is spelled <rr> when the word
 >    becomes a part of a compound word. "Guardarropa" (coat
 >    room) comes to mind immediately, formed from "guardar"
 >    (keep, guard, protect) and "ropa" (clothing).

Interesting, rhough _guardarropa_ wouldn't seem to be an
instance of it, if the first part of the compound is really
the infinitive.

 >        The <r> is doubled to indicate that the /r/ is
 >        retained, since it would be /4/ otherwise. Cf.
 >        "perro" and "pero."

I seem to remember that Monterey, California and Monterrey
Mexico are spelled differently, though!

This is a feature, or spelling convention, I'll probably
borrow in R#3.

 >
 > 2) When I lived in Honduras I was introduced to /r_0/, the
 >    voiceless alveolar trill. Allophonic, of course, it
 >    occurs when <s> is final. My American contacts often
 >    tell me they hear <sh> /S/. I'm told that this is part
 >    of a Caribbean regional accent.

I don't know if this is significant, but in Hlasa
Tibetan, where /r/ is a fricative /r\`/ the spelling _hr_
is pronounced /s`/. Are you sure Carribean Spanish /r7
isn't fricative? However that is American English /r/
*is* fricative, which may cause Americans to perceive
[r_0] as [S].

 >        I have no experience with that. I own several CD's
 >        of Nana Mouskouri singing in Spanish. She can be
 >        heard using this allophone often.

It may be a carryover from her Greek accent, though I don't
know if this is a feature of Modern Greek pronunciation.
FWIW geminated /rr/ was alophonically voiceless in
hellenistic Greek.

 >        I understand that /r_0/ is phonemic in Welsh,
 >        spelled <rh> to distinguish it from <r> /r/. It is
 >        a phoneme in my conlang Senjecas.

Also in Icelandic, where it is spelled _hr_, where it is
derived from Germanic */xr/. In Icelandic [r_0] also appears
in medial clusters of /r/ + /hp ht hc hk/, realized [r_0t]
etc. and written _rt_ etc. with [rt_h] etc. as a possible
alternative pronunciation.

AFMOC Kidilib has _hr_ /r_0/, which in some cases
corresponds to Classical Sohlob _hl_ /K/ and in some cases
to CS _çr_ /s`/. There were several Kijeb combinations
_*sr, *sry, *rs *xr_ etc. which merged differently in the
different dialects. As you may expect _r_ is [r] in Kidilib
and [r\] in CS and Heleb.

r#3 may get devoicing of final /r/, as it has devoicing of
final obstruents.  It may not probably get /r/ > [S] or
[s`], as it wont likely have fricative apical /r/ allophones,
and already has /S/ from other sources.


/BP 8^)>
--
Benct Philip Jonsson -- melroch at melroch dot se

    a shprakh iz a dialekt mit an armey un flot

                                 (Max Weinreich)