ATTENDED: Workshop on Modern Greek Koineisation

By: | Post date: 2021-11-08 | Comments: 2 Comments
Posted in categories: Linguistics, Modern Greek

So, over the past weekend, I’ve attended two linguistic workshops in Greece. (I’ve already complained about the scheduling conflict on Nick Nicholas’s answer to Why in Australia do people of Greek & Italian origin just say they are Australian, but in the US Greek/Italian American? Being closer to Europe, do they still wish to keep the old world ties in tact? How does that work also with Chinese Australians? .) The workshops were in the evening into the wee hours for me in far distant Australia, and my attention did fade as I went. But the ability to dial in to workshops and conferences, rather than have to get on a plane and take off two weeks and $5k, has been a boon of COVID to your antipodean correspondent, and he appreciates that it is now possible.

Even if that means he doesn’t get to go out to a taverna for chops with the participants afterwards.

Herewith, reactions to the bits of the koineisation workshop I attended and can remember. Because of the scheduling conflict, I wasn’t able to attend as much of the koineisation workshop as I’d have preferred; I only caught the first morning session.


As background: Contemporary Modern Greek as we know it emerged in Athens as the capital of the Modern Greek state; it owes only a couple of words to the native dialect of Athens, which was extinct by the 20th century, and a lot to the dialects that converged in the capital, settled from elsewhere in the Greek-speaking world. That makes it a koine, but a koine with a poorly understood history: we don’t have a lot of written records of how the dialects converged in Athens in the 1840s and 1850s.

We also don’t have a lot of records of Peloponnesian, the dialect group widely held to be the basis of that koine: dialectologists have not bothered to record it historically, precisely because they assumed it was identical to the Modern koine. (And as Nikolaos Pantelidis has been saying for decades, they were wrong.)

We do know that Puristic Greek did contribute to the formation of that koine, as a template to fall back on when dialects disagreed (or even when they didn’t): that’s something Brian Newton worked out in the 70s, in explaining why Standard Greek has gone with the archaic –ome instead of –ume as a verb inflection. (Yes, that is also the inflection used in the Ionian islands; but no, the Ionian islands really did not have as much influence on Standard Greek as literature scholars like to think.)

And because the Greek Language Question was and remains so ideologically loaded, pointing out that Puristic contributed to the Modern koine is still contentious. I have anecdotally heard of conference papers being rejected for suggesting so.

So, what papers do I remember?


I’m going to start with a paper from the time-clashing workshop, on post-antiquity Greece, by Pantelidis himself. Thematically, I think it’s a better fit with this group of papers, and I have already mentioned it.

Pantelidis’ current interest is the survival of the Ancient Greek pronunciation of /y/ as [ʏ, ʉ] in Modern Greek dialect (as opposed to Standard /i/, with a few dialects doing [ju])—notably in the dialect group of Old Athenian, which also includes nearby Aegina and Megara. There are lots of quite clear statements in the 19th century that people in Athens were pronouncing it as <ü>, but they weren’t being made by Greek speakers (who could not hear the sound, because they were Greek speakers); those statements were being made by Germans and French, and they were either ignored by subsequent scholarship, or disbelieved.

Pantelidis concedes that in most places, the old pronunciation has vanished with nobody remembering it was ever otherwise. But he has been digging up more recent recordings of Aegina and Megara, and he’s got the instrumental phonetics to prove that people are still producing [ʏ, ʉ] in 2016—a millennium after a poem in 1030 mocked a priest as coming from a village “where people’s intellect is not better than oxen”, because they were using the new-fangled [i] pronunciation. And he’s been the first to notice it.

Roger Lass once wrote a paper called “When is a language change”, noting that Southern English started dropping its r’s in the 17th century, but a few people were still pronouncing them in London as late as 1870. It is convenient to think of language change as a brief switch-over from form A to form B, especially when looking back to a change that happened centuries past; but that is a convenience. Language change takes a very long time to settle.

I’ll also note that there’s another instance of the cloth ears of Greek dialectologists, that Pantelidis bemoans, in the eccentric dialect Samothrace. That dialect happens to have high central vowels. No ethnic Greek dialectologist has ever noticed them. August Heisenberg (Werner’s dad) didn’t notice them either, when he recorded the dialect of a Samothracian POW. And that’s because high central vowels are not a feature of either Standard Greek, or Standard German. If you’ve got a poor ear, you’d have to be a native speaker of a language like Romanian to notice them.

… Or Aromanian. Which is why the Vlach linguist Nikos Katsanis was the first to record them in Samothracian, in the 1980s.

Pantelidis is indignant about the poor job Greek dialectologists past have done with the data of the Greek language, and concluded by saying that “I’m afraid we’re going to have to redo everything from scratch”. I find it terrifying to hear him say that, because he may well be right.


Eleni Karantzola’s paper was an overview of the five grammars of Modern Greek written in the 16th and 17th century, the time period she specialises in and keeps doing wonderful work on. I’m proud to say that I was already familiar with those grammars (I wanted all the data I could get; unfortunately I was working on syntax and not morphology, so those grammars had nothing much to tell me.) Her conclusion was that those grammars were trying to set up a standard form of the language; they do point out dialectal variation (Romanus’ more than others, but I remember Portius’ doing so too), and they do pick forms as the preferred forms.

She also found that the literary works of the time completely ignore those grammars’ advice, and that’s to be expected: unlike France and Italy and the other countries who were standardising their languages at the same time period, those grammars did not have the authority of the State or of organised writers behind them.

Christos Karvounis’ paper on the formation of Modern written standards in the 18th and 19th century, I am told, was eagerly attended, but Karvounis was a no-show, so can’t speak to that.


Just as Karantzola is working on the 16th-17th century, Peter Mackridge is working on the Phanariot texts of the 18th century. The 18th century is the last frontier for Greek historical linguistics, and has had very little attention paid to it to date; even the texts we have were mostly published in the last two or three decades—and there are plenty of texts still unpublished. That lack of attention has mostly been ideologically motivated: the Phanariots, as loyal Ottomans, were regarded as traitors in the new Greek state, and 19th century language activists dissociated their initiatives from them.

The Phanariot texts are grammatically consistent in their version of the vernacular. (At least, the vernacular of their authorial voice and their Constantinopolitan characters. To the Phanariots, all their islander servants sounded the same, which is why the low-class characters in their literature mix up the dialects of the Ionian islands with those of Crete and the Aegean.) And whatever koine had formed in Constantinople must have contributed to what happened later on in Athens, even if it didn’t contribute much. (The imperfect –usa inflection is the only element of the Modern Standard we can be sure is Constantinopolitan, although Triantafyllidis did notice that the Northern Greek accusative indirect objects used to be fashionable in Athens; now, they are what Athenians mock Salonicans for.)

But much like the other literary “koines” that preceded it—that of Early Modern Greek literature, of Cretan literature, and in a sense of folk song—the Phanariot literary “koine” seems to have been a passive thing, rather than an active thing. They dismissed islander dialect as peasant-talk, but they didn’t have a rigorous norm of what mainland dialect features were in or out.

Rather, they had a vague and varying sense of what dialect features particular to their own region were not used more widely, and sometimes they avoided them, sometimes they didn’t bother. They de-Thracianised their language, but they didn’t de-Thracianise it assiduously. Mackridge has confirmed that they avoid the characteristically Constantinopolitan broad use of pu as a complementiser, which Cavafy and (in the first edition of My Voyage) Psichari did not manage to—although that could just be because it was a later innovation. But some authors use the Thracian subjunctive ðjo “see”, while some avoid it and use the more common form ðo.

Mackridge finds literary Phanariot charming, as an evolutionary dead end of koineisation—what might have been Standard Modern Greek, if history took a different path. People have long said the same about literary Cretan—and, because of Language Question politics, more people have felt they had license to.


Christos Tzitzilis’ talk…

… ok, this would normally be a career-limiting move, but this isn’t my career to begin with.

Tzitzilis is coming from a place of militant demoticism, which has always been Salonica’s preserve. But to refight battles that were settled in 1910 is just embarrassing. And the prescriptivism is no more welcome in linguistics from someone who thinks I should be talking like an unsullied 1900s peasant, than it is from someone who thinks I should be talking like St Paul.

The ire of Tzitzilis was drawn in this instance from the current confusion around the 3rd person aorist passive. Puristic Greek is officially dead as of 1976; unofficially, Puristic morphology is still around, and was reasserted in the public sphere from the 90s on. As a result, there’s a vernacular standard inflection (-θike), there’s an archaic variant which is now showing up with high register verbs in 3sg (-θi), and there’s a wide range of hybrid forms, as speakers are confused as to how archaic their learnèd aorists should be: should they also be doing an internal augment? A second aorist? Should they be levelling the stems in the vernacular direction instead?

You can look at that as a language community member, or indeed a language activist, and find it laughable. I’ve done it with the bizarre revival of feminine –us genitives (Σαπφώ, Σαπφούς). You can think that the country has lost its marbles by reviving those forms. You can agitate for them, and you can agitate against them.

But you’re not doing linguistics when you do that. You’re doing something different.

And to turn around and say that the Greek people is dishonouring its history, by not following the vernacular idealised norm, is as silly it is to say that it is dishonouring its history, by not speaking like St Paul. To go on to say that fully productive inflection paradigms are a fundamental entitlement of speakers (!) — that somehow the 3sg-only archaisms are a violation of human rights—is several steps beyond silly. And to defend it with a teleological argument is just embarrassing. Sure, the Standard vernacular inflection resolved some problems that the earlier inflection had in a Modern context.

That doesn’t mean it was inevitable and optimal: what language change is? The island dialects have been slow to adopt the innovative inflection: what does that make them, brain-damaged? Less mentally agile to see a problem than their mainland cousins? No. It just means teleology is a poor way to explain language change. None of it is inevitable, or just about problem-solving.

Greek has always been subject to learnèd influence. Greek has always had people hesitate on how to inflect learnèd forms. Greece is neither the first nor the last language to have been impacted by diglossia, or indeed by clashing registers. Yes, it’s messy if you’re a neogrammarian. Yes, it violates the simple rules that the mainland dialects followed in 1800. And that just means you’re going to have to rewrite your rules, because that’s how people are now speaking. The Common Greek Language is what Greeks commonly speak. Standard Modern Greek is the language of texts considered to be standard in Modern Greek. It’s not the other way around, however distasteful that might be politically.

That term “Common Modern Greek”, Koinē Neoellēnikē, was coined by George Babiniotis, when he proclaimed that the new standard of 1976 was not just the Demotic of old that Psichari had defended: that it was indelibly marked by the heritage of Puristic. He said so in a polemical paper, “Beyond Puristic and Demotic”, a title consciously patterned after Nietzsche’s “Beyond Good and Evil”.

Yeah, you can say any number of bad things about Babiniotis, from his alleged linguistic tub-thumping in support of the junta, through his eccentric pedantic orthography, to his promotion of antiquarian inflections, and just recently his dismissal of abuse survivors in schools he was on the board of.

… As a dialectologist told me years ago: “As a human being, he is a turd (σκατάς). But linguistically, he was completely right about Demotic.” And he’s more right than Tzitzilis is.

Tzitzilis in fact knew that he was going too far in his decrying of those inflections. He did concede that there are learnèd elements that have been accepted into the Modern Standard, and that there are formal models of language variation that can be used to make sense of it. He cannot but: Psichari is still being mocked today for rejecting those learnèd elements 130 years ago. But, he insists, there are limits! There are features that the people’s linguistic intuition shall never accept!

Well, no. There are no limits. Nobody blinks now at hiatus—the learnèd violation of the universal rule, that /i, e/ before a vowel was reduced to [j]. Everyone accepts that, everyone knows that vjazo means “to rush” and viazo means “to rape”, that aðja means “empty” and aðia means “day off”, even if they are spelled identically. Because people have adjusted their internal grammars.

But that’s been no less chaotic and artificial a change to Greek than the passives Tzitzilis is decrying. It’s actually added an entirely new phoneme to the Greek lanɡuaɡe (/i/ distinct from /j/), a distinction neither Learnèd nor Vernacular Greek made, and which is all about register. itʼs thrown the phonoloɡical analysis of Modern Greek into disarray for decades (because people really did not want to accept that /j/ is a new phoneme.) It’s made for confusion diachronically: Vamvakaris in the 1930s sang “divorce” as five syllables, ði.a.zi.ɣi.o, since at least the 90s songs scan it as four syllables, ðja.zi.ɣi.o, and I don’t doubt there are people out there who pronounce it as three, ðja.zi.ɣjo. (The hiatus of διαζύγιο “divorce”)

That confusion is OK, and the passives confusion is not? That’s mere hair-splitting. And note that people’s linguistic intuition is playing a role in how that archaic passive is being revived, and how it is being integrated into the vernacular: they’re only using the 3sg –θi, and not the 1sg –θin or the 2pl –θite, because the 3sg suffix fits much more readily into the vernacular than the other persons.

Mertyris & Sampanis

The framework Dion Mertyris and Konstantinos Sampanis presented on Standard vs Koine Modern Greek was altogether more satisfactory, and I’m not just saying that because I was the unofficial supervisor of Dion’s thesis. Their framework draws on an unfortunate conflation Greek was doomed to make: the “common” language was how Greeks always referred to its vernacular—”Demotic” was a 19th century term, itself a creation of Puristic Greek. When Babiniotis rejected that the new Standard was Demotic, he fell back on the old word, Koine. But “Common Greek” is not the same thing as “Standard Greek”; and English-language scholarship has been careful to use the latter. Greek-language scholarship has not, and virtually never refers to Standard (Protypē), as opposed to Common (Koinē) Greek.

This allows them to differentiate between ±Standard and ±Koine elements of Modern Greek, without getting as stuck on the Learnèd/Vernacular, Puristic/Demotic dipole. There are elements of the Athenian Koine that are not part of the Standard language: Athens slang, and informal inflections like the southern Greek imperfect –aɣa (displaced in the Standard by Constantinople’s –usa), or the pronominal ending –onon (e.g. alunon “of others”, ðjonon “of two of them”). There are elements which are both Koine and Standard. There are elements which are Standard, but not Koine: features which are not of vernacular origin, but which are accepted in the Standard. And there are elements which are neither Koine nor Standard—Puristic archaisms which have been left out of the current Standard.

As Mertyris and Sampanis concede, the Standard has changed: the Standard of 1976 allowed more learnèd elements than the Standard of 1941. I’d add that the Koine has shifted too, if we define it, not as what is of learnèd origin, but what Greeks’ linguistic intuition does not protest. (Mertyris and Sampanis themselves refer to learnèd elements treated as compatible with the vernacular—as indeed did Tzitzilis.) The hiatus of /i/ vs /j/ is clearly learnèd, is clearly now regarded as +Koine, and just as clearly didn’t used to be. People have developed more of a tolerance for archaic elements than they had two centuries ago. Even if there are still elements they have not accepted.

(Are there limits, after all? Yes there are. But they are contingent, not predetermined. A new passive is as eligible as a new phoneme.)


The last talk I got to attend was Dionysis Goutsos’, on what the corpora of Greek he’s been involved with for the last three decades tell him about the spread of vernacular features in Greek. The curve we expect to see, from Labov’s work on sociolinguistics, is an S-curve: a feature starts infrequent, then accelerates in take-up, then levels off as it becomes universal.

In a context of language-planning, where features are promoted or suppressed in official discourse according to the politics of the day, we expect to see some wiggles in the data that don’t go in one direction, as a monotonic change. But Goutsos’ data, tracking various learnèd shibboleths through the 20th century… was weird. There was a consistent trend of lots of vernacular features in the 1900s-1910s, retreating through to the 1960s, and then coming back up—particularly in the 1970s, once the official rejection of Puristic took place (and even more so in the 1980s, when the Socialists mainstreamed vernacular choices in grammar—something we have seen a backlash to in decades since. The 1980s is when the –eōs singular of Classical –is nouns died out: poleos > polis “of the city”.) Goutsos finds a U-curve.

There were political changes one could attribute to the timeline: the political reaction to early Demotic activism, culminating in the Gospel and Orestiad riots of 1901; the relaxation on Demotic in schools in the 1920s; the Greek Civil War, and the association of Demotic with Communism. But the curve doesn’t seem to fit those dates, and I admit that I cannot understand how the U-curve happened: I would want to have a closer look at the data.

Thankfully, Goutsos did provide a closer look, breaking the corpus down by genre. The broken down results did make much more sense; in particular, movie dialogue and newsreels were level in the degree of their vernacular, reflecting consistent, vernacular-leaning prescription. The most interesting genre, with the most wiggles in frequency, were private letter-writing: private citizens took in learnèd influence, and the ups and downs in prestige of the vernacular—and they adjusted how they wrote decade by decade, without any copyediting to push them consistently one direction or the other, the way that happened in either Puristic, officialese discourse, or conversely Demotic, movie and literary discourse.


Leave a Reply

  • Subscribe to Blog via Email

    Join 319 other subscribers

  • December 2021
    M T W T F S S
    « Nov    
%d bloggers like this: