Is the Greek Cypriot and Cretan pronunciation kk = ts (zz) derived from Venetian, or is it archaic?

By: | Post date: 2016-10-23 | Comments: 4 Comments
Posted in categories: Linguistics, Modern Greek

The question and the question details are asking different things, and I’ll address them separately.

It is the doom of /k/ in front of a front vowel (i, e) to be palatalised, to be pronounced as [kʲ] > [c]. The palate is a notoriously difficult place to articulate a stop (too much surface area). So [c] more often than not ends up become (a) an affricate and (b) palatoalveolar: [tʃ] (moving forward in the tongue root, to where there is a more well defined articulator). It can move even further, and become an alveolar affricate: [ts].

That’s what’s happened to Latin <ci> throughout Romance. Caelum would have started as [cielu]; then [tʃielo] in Italian, then [tsiel] in French, which then simplified further to [siel].

The same change has happened in the Cretan and Cypriot dialects of Greek: /kokina/ ‘red’ > [kotʃina]. Standard Greek, on the other hand, stops at [c]: [kocina]. Both Crete and Cyprus spent time ruled by Venetians. Did they get [tʃ] from the Venetians?

No need to: this is a linguistic commonplace, and it has happened in dialects of Greek with no contact with Italian: Tsakonian, for instance (/kairos/ > [tɕere]), or Cappadocian (/kelyfos/ > /tʃefos/). In fact, Peloponnesian–Heptanesian, the base dialects of Standard Modern Greek, are outliers in not having palatoalveolars.


The question details throw some words:

Latin Aretium > Italian Arezzo

Darıca, a town in Turkey whose Greek name was Aretsou, and whose ancient name may have been Arethusa (though its classicising name was Rhysion).

Now, alveolars also palatalise cross-linguistically: Latin <ti> has indeed been through a bunch of changes, moving it towards the roof of the mouth, although it typically does not go further than affrication: [ti] > [tsi]. So Latin natione > Italian nazione [natsione] > French [natsion] > [nasion]. English in turn palatalised [sj] to [ʃ].

The corresponding palatalisation of /ti/ in Greek is rarer, but it has happened, and when it has happened, it’s been spectacular: /ti/ goes all the way back to [ci], in Tsakonian and Lesbian.

So. What about Aretsou?

I found out only today that [θθ] > [ts] was a thing in Finnish: Joonas Vakkilainen’s answer to What did your language sound like 500 years ago? (thanks, Joonas). But I’ve gotta say, I’ve never seen the equivalent in Greek. There is what <θ> pronounced as <σ> in Laconian, which appears backed up as rare instances of /θ/ > /s/ in Tsakonian. But that change was ancient, and likelier Laconian pronouncing /tʰ/ as /θ/ much earlier.

Aretsou is in Bithynia, so I was going to say “forget it, Bithynia was resettled by Greek-speakers in the 16th century from Epirus, there can’t have been any continuity from ancient Arethusa”. (Bithynia was the Ottoman heartland, so it was Turkicised early.) But Darıca is only 40km from Istanbul, so it’s plausible that it remained Greek–speaking after the Ottoman conquest.

Still, because I haven’t seen [θ] > [ts] elsewhere in Greece, I think Arethusa > Aretsou is unlikely. Any connection with Arezzo is also unlikely.

4 Comments

  • David Marjanović says:

    Standard Greek, on the other hand, stops at [c]: [kocina].

    In the IPA, c is a very convenient symbol for a quite rare sound. Consequently, it is commonly used for all kinds of more common sounds ranging from [kʲ] to [tʃ]. All the Greek I’ve heard, which isn’t much and was probably all modern standard, had exactly the same [kʲ] as Russian and Polish do. [c] instead sounds almost the same as [tʲ]; it occurs today in Hungarian and Latvian, and apparently in Basque diminutives.

    I found out only today that [θθ] > [ts] was a thing in Finnish:

    I’ve encountered this claim before, but, judging from what I’ve read of their writings, the Uralists don’t even consider the idea. They assume a sound change [ti] > [tsi] > [si]. From the purely graphic point of view, I don’t see a reason why Agricola’s z and tz shouldn’t simply be [ts] and [tːs] just as in German.

    There is what θ pronounced as σ in Laconian, which appears backed up as rare instances of /θ/ > /s/ in Tsakonian. But that change was ancient, and likelier Laconian pronouncing /tʰ/ as /θ/ much earlier.

    Why actually should [θ] have become involved at all? Isn’t it easier to interpret this as a [tʰ] > [ts] > [s] change? Apart from the High German Consonant Shift, Danish has recently completed the first half, and so has Scouse (Liverpudlian).

    • David Marjanović says:

      …Proto-Finnic *[ts] is also thought to have had a few other sources than *[t] before *[i], but I don’t know which ones.

    • [c]: in fact, the Russian linguist working on Mariupolitan Greek in the Ukraine, Zhuravliova, was adamant that the local realisation of palatalised /k/ was [tʲ]. “Nonsense,” said the Greek linguist Tombaidis when he went over; “it’s the same [c] we have.” [c] <κι> has also been claimed as identically realising /tj/ and /kj/ in Lesbos, and /tj/ in Tsakonian.

      All I can say from introspection is, my Standard Greek /kj/ contacts pretty full on the roof of the mouth.

    • /tʰ/ > Laconian <s> > Tsakonian /s/: I can see people assuming the <s> was a [θ] out of parsimony, avoiding /tʰ/ developing into two different phonemes in the history of Greek (and in fact of Tsakonian, since the normal tʰ > θ is regular). But this independent development of tʰ > s does indeed make a distinct intermediate stage like *[ts] more, not less likely. (We know that Arcadian had a [ts], but that was a reflex of PIE k′ʷ.)

Leave a Reply

  • Subscribe to Blog via Email

    Join 299 other subscribers

  • January 2018
    M T W T F S S
    « Dec    
    1234567
    891011121314
    15161718192021
    22232425262728
    293031  
%d bloggers like this: