· The Lucian Pronunciation of Ancient Greek ·

Ἡ Λουκιᾱνὴ Προφορὰ̄ τῆς Ἀρχαίᾱς Ἑλληνικῆς γλώσσης

a lovely & historically accurate pronunciation system for Ancient Greek

that is easy to learn & unifies all standards of pronunciation

by Luke Ranieri & Raphael Turrigiano

May 7, 2020


[companion article to the above video]

In this article we are going to present to you an optimal pronunciation convention for Ancient Greek, and the reasons why you might consider adopting it.

This article has two parts. In the first part, we’ll tell you about the sounds of the Lucian Koine Pronunciation and how they work in context. In the second part, we’ll explain how we developed this new, balanced convention, and why we recommed it for Ancient Greek. 

This article will have a lot of linguistic terminology and International Phonetic Alphabet symbols to help us clearly explain these sounds. If you’re unfamiliar, don’t worry too much about those! See the video above where we pronounce everything we’re talking about so you don’t feel lost.

PART I: The Sounds of Lucian Pronunciation

The phonology of Lucian Pronunciation is derived from the sounds found in Ancient Greek from about the 2nd Century AD.

Most Lucian Pronunciation vowel qualities are the same as in Modern Greek:

α /a/ ἄοκνος 
ε /e̞ / ἔχω 
ι /i/ ἴσως 
ο /o̞/ ὀπός  

Ancient Greek also has the sound:

υ /y/ λύκος  

Lucian Pronunciation retains the phonemic distinction between short and long vowels. The long vowels have the same quality, or basic sound, as the short vowels, but they have a longer duration:

ᾱ /aː/ ἄγᾱν άγααν
η /e̞ː/ ἔτη
ει /iː/ εἶπον
ῑ /iː/ αποκρί̄νεται 
ω /o̞ː/ κώμη 
ου /uː/ οὐδαμῶς 
ῡ /yː/ δείκνῡμι

N.B. ει /iː/ and ῑ /iː/ have merged completely.

Lucian Pronunciation also has two true diphthongs.

αι /ae̯̞/ κόραι
οι /øy̯/ οἴμοι 

The first of these you can think of as a transition between a and e.

The second of these is a transition from ø to y. Just as how y is the rounded version of i, ø is the rounded version of e.

L. Prior to the 1st Century BC, Ancient Greek had long diphthongs, marked by the iota-subscript, but these merged with the plain long monophthong vowels:

ᾳ /aː/ ᾄδω 
ῃ /e̞ː/ λῃστής 
ῳ /o̞ː/ κωμῳδίᾱ 

The “aspirate” is the name for the sound of English ‘h’ /h/. Pronounce this sound in the beginning of a word in isolation, or at the beginning of a sentence only. All other places in a sentence it should tend to go silent.

‘ /h~∅/ Ὁ παῖς τὸ ἅμα βλέπει.

As in modern Greek, σ and ζ are retracted because Greek of all periods lacks a post-alveolar fricative, the /ʃ/ sound.

σ /s̠~z̠/ συκίζω 
ζ /z̠ː/ ζητεῖ 

If your language contains both /s/ and /sh/, as in English or Italian you can find this sound by either retracting your tongue from your normal /s/, or pushing it forward from your normal /sh/. If you native language is Greek, European Spanish, Finnish, etc. then this sound should come naturally.

N.B. ζ is always geminated, meaning doubled in duration.

Letters δ and θ are dental fricatives, as in Modern Greek


Letters γ and χ are also as in Modern Greek. In front of back vowels, they are /ɣ/ and /x/ 



But before front vowels (αι ε ει η ι οι υ), they move forward into the palate, which is called palatalization, a key feature both of Modern Greek and Koine.

γαῖα note that αι diphthong is a front vowel
γοῖ note that οι diphthong is a front vowel

χαῖρε note that αι diphthong is a front vowel
χοίρᾱ note that οι diphthong is a front vowel


γ /ɣ~ʝ/
χ /x~ç/

Palatalization also occurs for κ 


κ /k~c/

γ is also used sometimes for the velar nasal


And when geminated, “double γαμμα” γγ can be αγγα or 

Palatalized αγγε; the second component is not a fricative, as in Modern Greek:


These letters sound pretty much as expected:

τaw /t/ τέτταρα  - make sure not to aspirate initial tau
πi /p/ πάππος  - make sure not to aspirate initial pi
ρo /ɾ~r, r̥ʰ~r / single ρ is a alveolar tap, ἆρα, while double ρρ is a trill, ἔρρωσθο 
λ /l/ λάλλαι 
μῦ /m/ μαμμίᾱ 
νῦ /n/ ναί 

Double consonants are also geminated — lasting longer in duration than single consonants, as in Cypriot Greek.

Letters φ and β are fricatives, but not exactly as in modern Greek. In the Lucian pronunciation these are bilabial fricatives, meaning they are made with the lips alone as opposed to the lips and teeth.



Finally, we have the two false diphthongs:
These consist of a vowel followed by a bilabial fricative. Ideally the bilabial fricative component should be rounded, making this truly the intermediate point between the attic ‘au’ or ‘eu’ and the modern greek av or af and ev or ef. Thus, we should have /aβʷ/ or /aɸʷ/ and /e̞βʷ/ or /eɸʷ/.

αυ /aβʷ, aɸʷ/ αὐτός, αὔριον
ευ /e̞βʷ, e̞ɸʷ/ εὔτροπος, Εὐγενίᾱ

Main Prescribed Version of Lucian Pronunciation
γ /ɣ~ʝ/
δ /ð/
β /β/
χ /x~ç/
θ /θ/
φ /ɸ/

N.B. both the β and φ are bilabial fricatives /β/ and /ɸ/, respectively; if labiodental fricatives /v/ and /f/ are desired, then αυ /av, af/ and ευ /e̞v, e̞f/.

Evolved Variant
If you want to have
φ /f/
β /v/
then you have to have
αυ /av, af/
ευ /e̞v, e̞f/

Archaic Variant
If you want to have
αυ /au̯/
ευ /e̞u̯/
then you have to have
χ /kʰ~cʰ/
θ /tʰ/
φ /pʰ/

The Lucian Pronunciation of Ancient Greek is based on my research into the changing phonologies of Latin and Greek through the millenia, which I’ve conducted with the welcome assistance of Raphael Turrigiano and Stefano Vittori. I have summarized many of these changes on my spreadsheet, available online to anyone, called Ranieri’s Greek Pronunciation Chronology at bit.ly/ranierigreekpronunciation as well as my presentation at Living Latin & Ancient Greek in New York 2020.

The formally called Imperial Roman Koine system had a somewhat free attitute toward the various possibilities that could have existed, and was somewhat obsessed with permissiveness.

We’ve limited some of the variations, and grouped them together in a way that made more synchronic linguistic sense. For example, in Lucian Pronunciation, you can have labiodental fricative /f/ for φ, but then you also have to have /aftós̠/ αὐτός, since this sound shift would have happened for any given group of speakers at the same time. So if you want to say αὐτός /au̯tós̠/, that’s permissible in Lucian Pronunciation as a more archaic variant, but then you have to keep letter φ the aspirated occlusive /pʰ/.

Our goal for the Lucian Pronunciation of Ancient Greek, named after 2cAD author Lucian or Λουκιᾱνός who may have used a pronunciation very much like this one, was to prescribe a more useful Koine pronunciation system than that espoused by Randall Buth, for whom by the way we have the uptmost respect, as well as all those who use Buth’s system. Buth recommends a highly evolved phonology, with all diphthongs monophthongized, and phonemic vowel length and geminated consonants as well as pitch accent disregarded; yet, these features, although decidedly more conservative, are attested to exist well through the Roman Empire — and geminated consonants still exist in Cypriot Greek today. The Lucian Pronunciation retains these characteristics because they are crucial to the literature. Oddly, Buth does not prescribe the palatalization of velars before front vowels, a key feature of most historical Koine as well as Modern Greek. By having the palatalization of velars before front vowels, the Lucian Pronunciation often can sound much closer to Modern Greek than Buth’s system, and is also more historically accurate — a win-win!

Now, not only is Ancient Greek a tremendously vast beast diachronically, stretching from epic Homer in the 8th century BC to the end of Antiquity around the 5the century AD, but also diatopically as the tongue of all the varied regions of the Eastern Roman Empire, and in Greek colonies in southern Italy and as far west as Spain. Koine is somewhat more strictly defined as the language of the Hellenistic period through the fall of the Western Roman Empire, but that’s still a huge period of at least six centuries.

The Lucian Pronunciation is an attempt to bridge the gap between the sound of Modern Greek and the systems of spoken Ancient Greek that others have become accustomed to, but are not historically valid — namely, Erasmian.


PART II: Methodology

Our friend Logan Kilpelä, whose YouTube channels Menelmacar & Menelvagor we recommend, wisely said that the reconstruction of the historical pronunciation of a language is a delicately balanced convention formed by the intersection of four factors:

science + art + pedagogy + politics

The first component, science, refers to the historical accuracy knowable based on as much direct evidence as possible, such as first-person accounts of ancient grammarians, spelling errors and other inscriptional data, as well as what can be inferred through the rigors of historical linguistics, working backwards from the modern dialects of Greek. However, the people using ancient Greek every day are by and large not historical linguists. The reconstruction of a historical pronunciation for regular use by modern speakers must start here, but it may not end here.

The problem can arise that the inherent uncertainties of the linguistic process offer more than one plausible outcome. Perhaps just one version is the only right one for a given time and place, or perhaps the others coexisted with it. The differences could be diachronic (variants through time), diatopic (variants over geographical areas), diastratic (variants according to different social groups), or diaphasic (variants of style or register). And with insufficient scientific data from which to draw a clear conclusion, as is the case with certain aspects of Ancient Greek, aesthetic concerns come into play: this is the shaping of our convention using art as a legitimate augment to the science; that is, a way to make decisions. We grant that aesthetic preferences are subjective, but they do matter

What we ultimately want is that aesthetic choices are made in reference to a universal default model, such that while there is diversity in pronunciation, that diversity centers around a shared standard that is scientifically and pedagogically sound, rather than a complete free for all. Additionally, we believe that this standard will make it easier for people to make the aesthetic choices that work best for them. For instance, many people may like aspects of the Attic reconstruction and aspects of the modern pronunciation but not know how to reconcile the two in a coherent manner. If you like our system, we encourage you to follow it, but we also grant you may want to make archaicizing or innovative modifications according to your preferences.

Pedagogical concerns are important as well. Certain pronunciation features may inhibit the ability to teach or learn the language, either due to the native language of the speakers or other factors. Moreover, the goals of most teachers and students are best served by a pronunciation that functions for the literature. Most learners will want to use their Greek to access a wide range of texts, and this means choosing a pronunciation which naturally fits the metre of poems, the ancient melodies, and thus also provide a chance to capture the authentic feel of the language as it may have existed in the minds of its native speakers.

And political aspects are necessary to think about too, such as potential acceptance by (or comprehensibity for) the established community of Ancient Greek speakers, as well as the living population of Modern Greek speakers. Our pronunciation system should, with a bit of adjustment, be just as comprehensible to a modern Greek as it is to an emulator of reconstructed Attic.

Together, these determine the viability of pronunciation conventions.

This quadrivium — science + art + pedagogy + politics — applies in the following manner to the prescriptions of the Lucian Pronunciation of Ancient Greek.

By far the majority of our prescriptions are based on linguistic and philological science. The evolving fricative nature of β during the 1cAD, for example, has been well established by Horrocks and Buth, among others — but what kind of fricative? /β/ or /v/ ? Earlier we showed you a few alternative systems. For example, if you want to pronounce φ as /f/, that’s fine, and reasonable for early Roman Imperial Koine, but you have to then pronounce β as /v/, and the false diphthongs as /av/ and /ev/ as well. And if you want to pronounce αυ and ευ in their older forms /au̯/ and /eu̯/, then φ must be /pʰ/. You can’t have both /au̯/ and /eu̯/ as well as fricative /f/ for φ. The linguistic constraints demand one of the sets be chosen.

That’s why we have prescribed φ and β as the bilabial fricatives /ɸ/ and /β/, and αυ /aβʷ~aɸʷ/ and ευ /eβʷ~eɸʷ/ — not only is this historically plausible, even likely in this period, but we find these sounds aesthetically pleasing. Additionally, the combination of rounding and frication makes it comprehensible to users of both true diphthongs and the modern fricative values. Another aspect where art has a profound influence on our decision making, is in Ancient Greek literature, namely poetry. Buth has adequately demonstrated that many Greek speakers were losing phonemic vowel length as well as pitch accent during the latter Koine period. However, “losing” does not mean “completely lost” for all speakers, particularly when these changes could be diastratic or diatopic. Koine poetry and songs unto the end of antiquity demonstrate beyond a shadow of a doubt that pitch accent and phonemic vowel length are as essential as ever. Therefore, in order to appreciate the aesthetic beauty of these works, intuitive knowledge of long and short vowels as well as pitch accent is most vital — and that intuition is best cultivated through everyday speech. We believe the challenges of learning pitch accent, however, may be mitigated by allowing stress accent to accompany the pitch accent, which was doubtless a transitional pronunciation system of Koine.

This decision has a pedagogical element as well. By going through the effort to learn phonemic vowel length, along with geminated consonants, one becomes a better speller and notices conjugation and declension more readily. Additionaly, correct recitation of Ancient Greek poetic metre is naturally a skill many would like to acquire. Accurately pronouncing long and short consonants and vowels will make the vast majority of metre effortless. Pedagogical concerns also temper certain linguistically sound concepts; for instance, in this period, the dominant pronunciation for χ θ φ may well have been an affricates — but these have the disadvantage of not being easily pronounced with the necessary precision by most people interested in Ancient Greek — namely speakers of Western European languages, although we by no means discourage attempting them. This is also true of the aspirate pronunciation of χ θ φ, which, while we believe it is a plausible variety extant through all of Koine, perhaps even the dominant pronunciation, the aspirated consonants are not sufficiently clearly understood by European language speakers today, and most are very bad at distinguishing them from unaspirated consonants, and so we deemed them as not worth prescribing for our main system for Lucian Koine, and just as an alternative.

Another pedagogical choice has to do with whether the aspirate /h/ is retained. This sound behaves quite the same in the contemporary Latin of the period as it does in Ancient Greek, and grammarians assert that this sound exists in both languages well into the Roman Empire. Yet we have plenty of inscriptional and evolutionary evidence that it was vanishing everywhere. We recommend it be pronounced at the beginning of a word in isolation, or at the beginning of a sentence, but internal to a sentence it should tend to go silent. This is the same practice we use when we speak Latin, and the results have been quite good. [N.B. Instead of this, Raphael prefers not to pronounce the apirate /h/ in any location, as he believes that speakers with fricative /x/ would probably have already lost this sound. Luke disagrees.]

Politics are unavoidable even in historical phonological debates. There are two main political concerns relevant to this discussion. The first has to do with whether or not it is legitimate to revive an ancient language like Koine Greek to be spoken on a daily basis. As Ancient Greek has already been revived successfully and continues to enjoy the Sixth Renaissance, as Christophe Rico calls it, we’ll leave this aside.

The other political factor has to do with broad acceptance and usage, especially for the modern inhabitants of the countries of origin. We have found that the Restored Classical Pronunciation of Latin is, not suprisingly, generally unpopular with Italians, who are most familiar with the Ecclesiastical Italian Pronunciation of Latin. However, this appears to be part euphony, part cultural. The Restored Classical Pronunciation of Latin is most often associated with Germans and Anglophones who speak Latin — the vast majority of Germanic language speakers have very un-Latin intonation, vowel placement, and attention to consonant articulation, generally using their native phonologies with little regard for the finer details.

Since we both speak several romance languages and attempt to use much of romance phonology and intonation, where applicable, in our spoken Latin, the Italians we speak with tend to find our pronunciations sufficiently pleasant as to be acceptable to their ears. So the main reason that Italians tend to disdain what they think is “Restored Classical” Latin pronunciation, is not because it’s /kɛːna/ instead of /t͡ʃeːna/, but because an English speaker might tend to say /kʰeɪ̯nə/.

To that end, we encourage Ancient Greek speakers to at least study basic Modern Greek phonology if not just learn the language to fluency, and to listen to Modern Greeks speaking Greek, imitating their intonation, prosody, and whatever sounds that are common to both Lucian Pronunciation and Modern Greek. We’ll provide a playlist of videos of Modern Greek speakers in the description for you to enjoy. 

We also recommend the same for Latin: listen to Romance speakers, and imitate whatever is appropriate for Classical Latin. [Luke and Stefano Vittori had a nice long talk about that in Latin in a six-episode podcast series.]

We have observed that many involved in the theory and practice of historical phonology often, perhaps even unknowingly, rest upon one of the latter three of the quadrivium (namely art, pedagogy, or politics) while spuriously invoking linguistic science as justification. For example, we have known some Greeks who assert that many, if not all aspects of Modern Greek phonology, date back even as far as Homer! And these folks seem to want to use a tortured version of historical data to make their aesthetic preference more legimitate. But that’s just not necessary — a pronunciation doesn’t have to be historically verifiable to be a legitimate practice. There’s nothing wrong with a Greek person today preferring to recite Ancient Greek literature with the same sounds heard in today’s δημοτική — just as their is nothing wrong with an Italian wanting to recite Latin in the Ecclesiastical Pronunciation — if for no other than it’s more pedagogically convenient for them, and aesthetically what they’re used to. So speaking about these motivations honestly, rather than hiding behind a guise of facts, is key. As another example, the Erasmian Pronunciation is demonstrably inappropriate as an historical reconstrution of any era of Ancient Greek due to its imbalanced restoration of various phonemes. However, we fully accept that someone may prefer it, or any other standard, and we urge those people to say honestly that this is simply because they are accustomed to it.

The Lucian pronunciation system, then, can act as a binding glue between the most common systems in use — Classical, Erasmian, late Koine and Modern. Since it has similarities with all of these systems, it increases the mutual comprehensibility of everyone, while still managing to be historically plausible. It also promotes an atmosphere of tolerance, since it has both internal variation and is compatible with other pronunciation systems.

We hope for the Lucian Pronunciation to grow be a conventional pronunciation through which Ancient Greek literature of any century can be enjoyed.


Main Prescribed Version, All Letters

α /a/
ε /e̞ /
ι /i/
ο /o̞/
υ /y/
ᾱ /aː/
η /e̞ː/ = ῃ
ει /iː/
ῑ /iː/
ω /o̞ː/ = ῳ
ου /uː/
ῡ /yː/
αι /ae̯̞/
αυ /aβʷ, aɸʷ/
ευ /e̞βʷ, e̞ɸʷ/
οι /øy̯/
υι /yː/
ᾳ /aː/
ῃ /e̞ː/
ῳ /o̞ː/
‘ /h~∅/
ζ /z̠ː/
σ /s̠~z̠/
γ /ɣ~ʝ/
δ /ð/
β /β/
χ /x~ç/
θ /θ/
φ /ɸ/
κ /k~c/
τ /t/
π /p/
ρ /ɾ~r, r̥ʰ~r/
λ /l/
μ /m/
ν /n/




© 2014-2020 LukeRanieri.com