Subscribe to Blog via Email
February 2023 M T W T F S S « Nov 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
The phonology of “Sitia”
No hyperlinks for this post, as my internet time is rationed while I’m on holidays.
Sitia, which is my hometown in Crete, does not figure prominently in history. The guidebooks say that in antiquity it was Eteia, and gave birth to Myson, one of the Seven Sages of Antiquity. The only Sage out of the set anyone’s heard of is Solon, and the Seven Sages were some sort of mythic convenience anyway, so that doesn’t count.
Sitia does turn up in the Middle Ages, with its S. It turns up in the Notitiae Episcopatum, the lists of bishoprics of the Empire. The lists went through a dozen generations, and were increasing garbled and fictionalised, as more and more of the sees they listed had not seen a Christian clergyman in centuries. The lists lived on in the West, and were used to generate the titular sees of the Catholic church: a way of granting someone a bishop’s title without a bishop’s responsibilities (unless they were minded to go over and convert Nisibis back to the True Faith).
Because the Catholic lists are available on the Intertubes, and the Orthodox lists are on dead tree I don’t have, the Catholic lists have often come to my rescue when I was adding placenames to the TLG lemmatiser. Particularly as the Catholic lists include the Latin nominative, while the Greek lists stop at the morphologically ambiguous genitive. So the Latin gave me as good a guess as any as to the gender of the towns listed.
Sitia turns out in two Notitiae, as Σιτία and Σιτεία. I don’t have dates on the Notitiae, since I don’t have the dead tree full text.
There are other mentions of Sitia here and there, some Byzantine, some Venetian, few Ottoman (the town was abandoned from the Ottoman conquest until around 1870). But the mention I’m going to dwell on is from around 1460; and I’m going to dwell on it because it has an oddity of historical phonology.
First, the phonology that the oddity perturbs. In Cretan, /t/ is lenited to /θ/ before [j]. So the plural of [ˈmati] in Standard Greek is [ˈmatja] “eyes” (from Ancient ὀμμάτια); in Crete it’s [ˈmaθja]. “Fire” is [foˈtja] in Standard Greek, and [foˈθja] in Crete. The singular of “eye” is still [ˈmati] thoughː the rule only applies when /i/ is not a syllable—that is, when it precedes another vowel. The rule is general enough for Cretan songsmiths to have taken advantage ofː
Δεν εκυνήγα λαγούς κ’ ελάφια
μόν’ εκυνήγα δυο μαύρα μάθια.
ðen etʃiniɣa laɣus tʃ elafja
mon etʃiniɣa ðjo mavra maθja
He would not hunt for hares and deer;
he’d rather hunt for two brown eyes.
This particular song has travelled far; it’s travelled as far as Mani, and from there to Corsica, without the phonetic rule that /t/ goes to /θ/. So when Maniats sing it, not only does [elafja] not fully rhyme with [maθja], it doesn’t even partially rhyme with [matja].
The song has travelled even further than Corsica: it’s made it all the way to the eastern tip of Crete, Sitia, which is the only part of the island that the rule does not apply. I heard the song several times while I lived in Sitia; I never heard the song rhyme, because the singers would sing it the way they spoke the dialect, and not the way the other seven eighths of the island did.
The name of the town, like I said, is Sitia, [si.ˈti.a]. The name has [i] as a separate syllable from the following [a], and that should tell you it is not a colloquial form: the vernacular rule is that [i] before a vowel goes to [j], and if it doesn’t, the form isn’t vernacular. That’s why the word for fire is [foˈtja], though it used to be [foˈti.a]. That’s why άδεια is pronounced two ways, without the spelling telling you which: [ˈa.ðja] for the colloquial adjective “empty”, [ˈa.ði.a] for the learnèd noun “leave, day off” (which is, literally, an empty day). The word for Turkey is now Τουρκία [tur.ˈki.a]; but folksong and previous centuries knew the Ottoman Empire as [tur.ˈkja] > [tur.ˈca], Τουρκιά.
Which means the town should have ended up called [siˈtja], and anyone from outside the town should have ended up calling it [siˈθja], just as they say [foˈθja].
That’s not what’s happened. The Renaissance poet Vincenzo Cornaro was from Sitia: the most renowned son of the city until contemporary Greek pop sensation Giorgos Mazonakis. And in the epilogue to his great romance Erotokritos, he introduces himself, and his home town:
Βιτσέντζος είναι ο ποιητής, και στη γενιά Κορνάρος,
που να βρεθεί ακριμάτιστος όντε τον πάρει ο Χάρος.
Στη Στείαν εγεννήθηκε, στη Στείαν ενεθράφη,
εκεί ‘κανε κι εκόπιασε ετούτα που σας γράφει.
The poet is Vincent, of Cornaro stock:
may he be sinless when he’s Death’s to take.
In S’tia was he born, in S’tia bred;
there did he write and labour what you’ve read.
So the town did not end up as [siˈtja], but as [ˈsti.a]. Cornaro’s quite careful about hiatus in his verse: vernacular verse avoids two consecutive vowels in separate syllables, just as the vernacular phonology did. In fact, he made a point of nativising Classical loanwords like περικεφαλαία [pe.ri.ke.fa.ˈle.a] “helmet” to the more vernacular-sounding περικεφαλιά [pe.ri.ke.fa.ˈʎa]. So if Cornaro is writing [ˈsti.a] (and we know the syllable count because of the metre), then that’s what the place was called in the vernacular.
As indeed it is called by the locals to this day: the town is Στεία, the adjective is Στειακός. The surrounding villages preferred to call it Το Λιμάνι, “the harbour”, because it’s where their produce ended up. And the local name S’tia still has that non-vernacular hiatus in it. It isn’t supposed to: the Ancient word for hearth, ἑστία [hestía], has ended up in other Modern dialects as their word for fire, στιά [stja], and phonologically that should be the same as S’tia: the reduction to one syllable should not have made a difference.
So the town name is more archaic than it’s supposed be. But that’s not in fact the conundrum I’m leading to. The conundrum I’m leading to is how the town gets named by Michael Apostolius.
You will occasionally, in accounts of Greek lexicography, come across plans to span all of Ancient Greek literature from Homer to Apostolius. Michael Apostolius (or Apostolis, the more vernacular form he used himself), was one of the many scholars who fled Constantinople after its Fall to the Ottomans. Greek was written before Michael Apostolius, and Greek was written after Michael Apostolius; but even though much of that Greek strove for compatibility with the Ancient norms, it has usually been beneath the notice of Classicists.
Apostolius marks an endpoint for Classicists, not because he wrote any Greek himself that they care about, but because he was the last Byzantine editor of a text they care about. There are three major collections of Ancient Greek proverbs: Diogenian’s and Zenobius’ from Roman times, and Apostolius’, who fills in some of their gaps. So lexicographers need to go through Apostolius’ proverbs, just as they need to go through the scholia to Ancient texts—the latest of which were written not long before Apostolius: to get as complete a picture of antiquity as can be recovered from the Byzantines.
So there’s one text by Apostolius that Classicists care about. There’s another text that less people care about: a collection of his correspondence. The collection was edited in 1888 by a scholar called Hippolyte Noiret, who died in his twenties before he could see it in print. I sighted a copy at the ANU library when I was there, looking somewhat forlorn; but it has also been added to the TLG.
I haven’t got around to reading the preface to Noiret’s edition entirely, but I was amused to read Noiret found Apostolius to be childish, both in his energy and his impetuousness. Whereas other Greek scholars ended up in Italy, taking up jobs at universities or printing presses, Apostolius ended up in a lower-rent version of Italy: Crete, which had been a Venetian colony the past two hundred years. Apostolius was always struggling to make ends meet as a scholar for hire, but he did travel widely through the island, and report back to his many correspondents. And one of the places he mentions is Sithia.
Seeing Σιθία in print took me by surprise. Was this supposed to be Σητεία? Noiret doesn’t address the question, and I wouldn’t have expected an answer from Google; but Google does now index old journals, even if the publisher is only making the back issues available through institutional subscriptions; and a snippet of a Byzantinische Zeitschrift article from the 1950s looks like confirming that yes, Σιθία is the town in the North-East of Crete.
There’s further confirmation from one of the letters where Sithia is brought up: the person carrying the letter was seeking to be appointed bishop of Sithia, and Apostolius trusts that his correspondent will help the letter-carrier in his efforts:
XCII: To the admirable Bessarion.
Since I am in receipt, wisest of cardinals, I have reported to you in the last writing, which three pious peasant priests have brought, that without your generous and gratifying contribution, namely the benefaction of twenty gold coins, I have no other income from any other source—as I am currently feeding five people, and now moreover another newborn child by the name of Aristobulus, whose father I have been called literally; and I now have an heir not in money, but in family. And if I ever happen to broach the topic of the scarceness of children (?) , I also owe some rent; the priests are my witnesses as to whether I speak the truth. Therefore, bearing in mind how I will turn out in my old age, and whence I will have resources to marry off my dearest children, I pray to you, prudent and generous, and would exact from you advice as to what I should do: whether I should come there and serve you with the others, or whether I should allow myself the solution of inquiring about Germany and England, and the surrounding districts as well, in the manner about which I shall come and kneel before your lordship and report to you.
But so much for that; and as I am friend to God, so may I be as well to you, my lord. Now, the one bearing this letter is a Cretan from good Cretan parents, who has lived saintly and justly. And being thus worthy of priesthood, he has kept company as if born and reared, first with a certain holy bishop of the Gortynians, then after a short time with the bishop of Cydonia as well—who was your attendant and who has also followed you in Venice and elsewhere. Being such and born of such, it is necessary—for you and many others, and through myself—that he become bishop of Sithia. He is who all Sithians want, and seek as their bishop because of the impeccability of his life, and because he is their family and familiar to them. For he too is Sithian, born and reared and living there. If you should help this man (and I hope you will help him promptly, emulating your innate generosity in this matter too), then you shall do what is pleasing to God, and natural to your forebears, aiding the poor and the good; and what is hoped for and dear to the Sithians. And you shall show me too to be good, a pauper helping a pauper, submitting myself to reciprocity.
If Sithia is a town in Crete big enough to have a bishop, and Sitia was already known from the Notititae to have a bishop, that does narrow things down a lot. It narrows the timeframe too: Sitia no longer has its own bishop, and is under the authority of Hieraptyna (now Ierapetra, “Holy Rock”, through folk etymology, and in the local pronunciation [ʝeˈrapetra]). That may be because Sitia was depopulated for a couple of centuries; dunno.
And if the letter was written in the 1460s, that’s not a Greek Orthodox bishop’s position he’s talking about either. The lobbying needs to happen in Italy, because Venetian Crete had Catholic bishops; the Orthodox islanders had to look to Modon (Methoni) for their spiritual guidance. Apostolius was Catholic himself, and several of his letters are about how to combat the schismatic Orthodox on their turf. The addressee, Cardinal Bessarion, was himself Catholic (which tends to be a prerequisite for being a Cardinal), and a good target for lobbying; he too was a convert from Orthodoxy, and one of the more prominent Greek intellectuals of his day.
But I’m still stuck on what Apostolius calls the place. [siˈθia] is not [ˈsti.a]. The /i/ hasn’t gone to [j], so the lenition doesn’t make sense by Modern Cretan standards. Moreover, as Chatzidakis already wrote a century back, there’s no evidence for the lenition from Renaissance literature, so the change must date from after 1669 anyway. When exactly I don’t happen to know, because the readiest evidence for that would lie in Ottoman archives, and Greeks until fairly recently liked to pretend the Ottomans had no archives. (Turci sunt, non leguntur.) But the change can’t date from the 1460s, and even if it did, it wouldn’t be as [siˈθia].
There’s another placename in Apostolius which probably confirms the absence of the sound change, but has problems of its own. The three western prefectures of Crete are named after their major cities: Hania, Rethymnon, Iraklion (formerly Kastro, “The Fortress”, and Candia before that). Sitia never made it to major anything-at-all status; so its prefecture is Lasithi, named after the mountain plateau to its west. Lasithi is also mentioned by Apostolius, as Lasition. That would be consistent with /t/ > [θ] / _i, and the plateau is far enough from Sitia for the change to be allowed there.
The problem this time is, the /i/ is not [j], because the -on ending had long been deleted in the vernacular. So the sound change can’t have happened: a learnèd Lasition corresponds to a vernacular Lasiti, not a *Lasitjo that would have gone to *Lasithjo. (Just as ὀμμάτιον /ommátion/ “eye” went to [mati(n)], not [matjo].)
So in Apostolius, we have a th in Sitia and a t in Lasithi, neither of which make sense by how the modern dialect works. The form may not just be Apostolius’ imaginary friend either: you will on occasion see the spelling Sethia in Venetian documents, although I’d assumed it was just the random insertion of h‘s that transliterations of Greek are susceptible to.
So how can we explain Sithia? I don’t know that we can; maybe it is just one of those things, a local variant without rhyme or reason. Which is the kind of defeat an historical linguist is never meant to admit to. There is a possibility that the vernacular form was actually Sithia all along; if Sithia was shortened to S’thia, S’thia would have to become Cornaro’s S’tia by vernacular phonology, which does not tolerate the /sθ/ cluster either. But the Notitiae have Sitia from earlier on, so I’m not convinced that’s what happened.
There’s one more feature of Apostolius’ letter which comes as a surprise. Ethnonyms, the names of people derived from where they’ve come from, have a multiplicity of suffixes to choose from in Greek. There’s a similar diversity in English (which is not completely unrelated): New Yorker but Washingtonian, Muscovite but Milanese. For the Byzantines, this multiplicity meant abundant confusion about which was the right suffix to use when writing Classical Greek. For that reason, Stephen of Byzantium wrote the Ethnica, a dictionary of ethnonyms; and because it is chock full of towns and ethnonyms never heard of before or since, it’s one of the texts the TLG lemmatiser has the most difficulty dealing with.
For Modern Greek, that multiplicity means an excuse for parochialism: every village picks its own suffix to names its people. Zakros has Zakrites; 20 km up the mountain, Ziros has Ziriotes. And the people of S’tia are S’tiaki, Στειακοί.
In Apostolius’ letter, they aren’t S’tiaki though; they’re Σιθιανοί, Sithiani.
This came to me as even more of a shock than the [θ]. Surely the S’tiaki have always been S’tiaki; surely this proves that Apostolius was making ethnonyms up.
Of course it doesn’t, though. Ethnonyms do change in time, and assuming the vernacular of 1460 is the same as that of 2010 is a mistake students of Early Modern Greek are all too prone to. Cypriot is the ethnonym used in English because Κυπριώτης [kipriotis] was the ethnonym that used to be used in Greek. In fact it’s the form used in Erotokritos. But Standard Greek only knows the form Κύπριος [ciprios], and Cypriot itself uses Τζυπραίος [dʒipreos].
And after all, there was a time before S’tiaki and Sithiani, when inhabitants of Sitia were called something different again. When Diogenes Laertes wrote about Myson of Chen, he brought up the conundrum of why he was also termed Ἠτεῖος [ɛːteîos]. Diogenes speculates that maybe one of his parents was from Eteia instead, “for Eteia is a city in Crete”.
Assuming Eteia is Sithia is Sitia; and the assumptions in this line of work are always more tenuous than you might think. (The assumption that Eteia is Sitia is why the town is now spelled with an eta, though: Σητεία.) And I still don’t know where that [θ] in the name came from. Or that initial [s] for that matter (although the metanalysis of /eis ɛːteían/ “to Eteia” to /eis sɛːteían/ is as good a cause as any).