At 8:03 pm -0500 13/6/01, Eric Christopherson wrote:
>> But in Italy & Romania the palatalized /t/ remained separate from the later
>> palatalized /k/; the former is /ts/ as in _nazione_ or _zio_ (uncle <<
>> /tiU/), and the latter is /tS/ as in _cinque_ [tSiNkwe] << VL *cinque
>> /kinkwe/.
>Hmm, how did /tiU/ become /tsio/, if (as I've always thought, and as you
>said) that the /t/ palatalization only happens before the *glide* /j/?

Latin /i/ before another vowel normally did become a glide in Romance.
Latin _natione(m)_ has four syllables, but Italian _nazione_ has only three
because /i/ is [j] there.

The Latin _theiu(m)_ [t_hiU] (<< Greek _theio-_) is interesting in that in
Italy the /i/ was treated as a glide, hence modern Italian _zio_ as I give
above.  But in the Iberian peninsular it retained its vocalic value, hence
Spanish _tio_.

>> >But now that I say that, I recall that
>> >/t/ before yod AND /k/ before front vowels came out identically in Spanish,
>> >so perhaps they did merge at some time to [t_j]. (But then mightn't
>>[k_j] be
>> >just as good a possibility? :) )
>> Not likely as in medieval Spanish soft-c and {cz} = [ts].  At the time {z}
>> = [dz], so that to represent [ts] before a back vowel they adopted the
>> convention of {cz}, eventually putting the {z} _beneath_ the {c} and hence
>> inventing the cedilla ('little zed)! The diacritic still has its Spanish
>> name in English.
>True enough, but I don't see how it's relevant; I was speaking only of the
>merger of [t_j] and [k_j], not the orthography.

It's certainly relevant as the orthography makes more sense if the onset is
dental; therefore it seems to me very unlikely from orthographical evidence
that [t_j] >> [k_j].  At some stage [k_j] certainly had changed to /ts/.
It seems odd to me that we'd have a change: [t_j] >> [k_j] >> [c] >> [tC]
>> [ts].

I still think on evidence it is more likely that [k_j] and [t_j] merged
either directly to [c] or, at some stage [k_j] merged with [t_j] and then
subsequently changed.

>> Later [dz] was devoiced to merge with [ts] which became the modern [T] in
>> Castillian and [s] in Andalucian and gave the Spaniards the opportunity to
>> tidy up their spelling and drop the cedilla entirely.
>Tidy up? But Portuguese looks so much nicer, with those cedillas everywhere

A matter of taste. I don't think any Spaniards would agree and its
orthography is certainly more regular.

>> >Related topic: Does anyone know why /c/ and /j\/ [...]
>> >seem to become affricates so frequently, where other stops don't?
>> But what other stops would go that way?  (I suppose aspirated /p/ became
>> /pf/ in High German.)
>Yeah, I was thinking of German. Presumably, there are other languages where
>plain stops went to affricates across the board, instead of just in one
>point of articulation.

I would think it not unlikely if the stops are aspirated.  It happens, in
fact, with final voiceless stops of the Scouse dialect of English where the
aspirate off-glide has become a homorganic fricative.

>> >I've wondered that for quite a while, and my guess would be that as a plain
>> >stop they sound to close to either the alveolars or the velars, but I'm
>> >really not sure.
>> Certainly [c] is close to both [t`] and [k`].
>I assume by [t`] and [k`] you mean here palatalized, instead of retroflex.

Yes, sorry - I should've written [t'] and [k']; I was getting the SAMPA
diacritics confused.

>Following that assumption, I agree with you there, but if [k_j] had already
>shifted to [c], there wouldn't be a [k_j] for it to be close to; and I don't
>know where the [t_j] would come from.

Sorry - I wasn't intending to be commenting on the Romance situation but
rather making a general comment.  Personally, I find [t'], [tS] and [c] all
very similar, but have no problem in distinguishing [k'] from those other
sounds.  But I know that others find [k'] almost indistinguishable from [c]
or [tS], but find [t'] significantly different.  It depends a good deal on
the speech habits of one's native language; I was brought in southern
England where _tune_ /tjun/ is more often than not pronounced /tSun/.


A mind which thinks at its own expense
will always interfere with language.
                   [J.G. Hamann 1760]