Morphologised Initial palatalizations?

[alice]

Are there any instances of palatalised and nonpalatalised consonants alternating in initial position within the inflection of a word? This could happen, for example, in some Indo-European varieties when an initial /k/ or /g/ is variously followed immediately by palatalising /e/ alternating with nonpalatalising /o/ as a result of ablaut.

[Linguoboy]

Are there any instances of palatalised and nonpalatalised consonants alternating in initial position within the inflection of a word?

Irish has at least one example:
bean /bʲan/ "woman" nom.s
ban /ban/ "woman" gen.pl

[missals]

There are some Bantu languages that do this. I can't recall an example from a specific language, but it generally went something like this:

okana > okana > okana > kana
ikana > ikʲana > it͜ʃana > t͜ʃana

[akam chinjir]

I took a quick look at this: http://linguistics.berkeley.edu/~merril ... tation.pdf. The only mention it makes of palatalisation is in Chaha (Semitic, Ethiopia), where it affects words from the right but you can get cases like "k’am (eat)" vs "kʲ’am (eat something small)." For what it's worth.

(Am I right to think that it's relatively rare for something to trigger palatalisation to its right?)

[Linguoboy]

(Am I right to think that it's relatively rare for something to trigger palatalisation to its right?)

Isn't this essentially what happened with Latin /ct/ in Spanish, Occitan, and Romansh? E.g. VL noctem > Sp noche.

I believe progressive palatalisation is also found in the history of Slavic, as well as in Old English (e.g. wīc > -wich).

[akam chinjir]

Wouldn't that be the "e" triggering palatalisation to its left?

[Linguoboy]

Wouldn't that be the "e" triggering palatalisation to its left?

No, because /e/ didn't generally trigger palatalisation of /t/ in Romance. Cf. VL partem > Sp parte.

Moreover, the palatalisation is observed even before back vowels, e.g. VL dictu(m) > Sp dicho.

[akam chinjir]

Ah, okay. Anyway the wīc > wich case, and the Bantu case that missals remembers, should work.

[KathTheDragon]

Latin /ct/ more directly turned into /jt/ - this is what caused palatalisation.

[Salmoneus]

Interestingly, the outcome in Mozerabic is "ht". Nobody knows what this means exactly*, and whether it's a different outcome, or an intermediate stage on the way to /jt/.

*this is true of everything in Mozerabic. It turns out that recording an early romance language in the medium of a regional variation on arabic orthography as practised by people who were second-language speakers of arabic is perhaps not a great idea...


ANYWAY: on the OP, this must surely have happened somewhere in Italy?

[Linguoboy]

ANYWAY: on the OP, this must surely have happened somewhere in Italy?

I can't think of any examples from Italy proper, but you reminded me that Sursilvan Romansh has umlaut in some masculine nouns and adjectives which, when combined with palatalisation, leads to some alternations like:

tgiern "horn", pl. corns (Cf. piertg "pig", pl. pors)
tgietschen "red" m.sg, cotschna f.sg

[Acid Badger]

I did a lot of reading about Mixe-Zoque languages ages ago and can't remember which language in specific it was, but it has a person/possession prefix /j-/ that is realized as palatalization of the following consonant. I think it was a Mixe language, I recall them using /j-/ for third person, but I can't seem to find the reference grammar in my folders. Edit: Apparently it's a Zoque thing: “the prefix j- indicates 3rd person in the imperfective aspect. This prefix is metathesized, presumably due to its high sonority."

[akam chinjir]

Here's a slightly longer discussion of the Zoque case, that treats it as (just) palatalisation rather than metathesis: https://books.google.com.hk/books?id=3_ ... lpg=PA1950.

[Frislander]

Actually it's not just Zoque, Southern Highland Mixe has it too.

[miekko]

You find some strong verbs in Swedish have alternations between /g/ and /j/, e.g.
gå - gick (where ''gå' has /g/ as onset and gick' has /j/ as onset)
ge - gav (where 'ge' has /j/ and 'gav' has /g/)

skära - skar ('cut') likewise alternates between S and sk (and Swedes from Sweden of course have some weirdo S-like fricative there, but ignore them)
stjäla - stal (steal, stole) alternates between StS and st IMD, and S- vs. st- in most other dialects in Finland

also, at least one noun has such an alteration:
gås - gäss (g / j) /goose - geese)

IMD you also find further examples where 'breaking' has occurred in contexts where it hasn't occurred in standard Swedish, giving pairs of Cj clusters initially that alternate with single instances of C, such as
bjär - bar (carry) (c.f. bära - bar)
jät - ut (eat) (c.f. äta - åt)
we also find an odd one out where a -j- that is present throughout the Swedish paradigm has been lost:
bjuda - bjuder - bjöd - bjudit < standard
bjud - bjudär - böud - budi < dialect
I guess the diphthongization of ö there (iirc a secondary diphthongization) got too much and the semi-vowel onset was lost to compensate.

you also find more complicated derivative pairs where alternations exist, e.g. standard Swedish
kåt - kättja (horny, horniness) where the latter has tS in Finland but whatever weird Swedish fricative in Sweden
kall - kyla & köld (cold (adj), cold (noun)*), where the latter has the same as in the previous
* distinguished by kyla being more general and köld being more often specifically weather-related cold)
skälla - skall (to bark, a bark), S vs sk
skämmas, skam (to be ashamed, shame)

This is of course not palatalization in the strict, secondary articulation sense, but as far as sound changes goes it can be called palatalization, and is close enough to demonstrate that at the very least, it seems very possible.