©2019 by KoineGreek.com


As someone with a specialty in the historical phonology of Koine Greek and Hebrew in the Late Second Temple Period, I often have to answer the following question: 'how do we know how Koine Greek was pronounced?'

While scholars and experts in the field have known the answer to this question for a long time, those of us just beginning to study Greek, or those of us who are asking questions about which pronunciation is the best for reading/speaking in ancient Greek (Erasmian, Restored Attic, Academic, Living Koine, etc.), it is a rather new question of interest.

Accordingly, in this short video, I attempt to explain methodologically just how exactly we determine how Koine Greek was pronounced based on the evidence from manuscripts, papyri, and inscriptions.

Feel free to show this to your students if it would be helpful!

A brief summary of the pronunciation of Roman Palestinian Koine Greek is outlined below. In a few places, it is noted where the "Living Koine" pronunciation (for pedagogical reasons) differs from that of the historical reconstructed pronunciation for Roman Palestinian Koine Greek.

For those interested in a much more in-depth look at the historical phonology of Koine Greek as well as a comprehensive application of this methodology to the data at hand, a number of resources are listed at the bottom of this page (note my own work [Kantor 2017] on the pronunciation of Palestinian Koine Greek).



* Note that in the Living Koine pronunciation, we pronounce δ, φ, θ, χ all as fricatives, even though these had not yet become fricatives in Roman Palestinian Koine.

In Classical Greek, the diphthongs αυ and ευ were pronounced as [au] and [eu]. During the Koine period, however, the second element of these dipthongs came to be pronounced with a consonantal value, first [w], then [β]/[ɸ], and then finally [v]/[f] in the Byzantine period. Thus, the sequences αυ and ευ came to be pronounced identically as αβ and εβ (see below for the pronunciation of this consonantal element).


See the following section.

In Classical Greek, the voiced stops were pronounced as [b] (as in boy), [d] (as in do), and [g] (as in go). In the Koine period, these stops became fricatives and were pronounced as [β] (like Spanish in trabajar), [ð] (like English th in this), and [ɣ]/[j] (a bit like gargling in some cases and like English y in year in other cases). However, when following nasal consonants—sounds you make with air coming out of your nose (μ and ν)—they were still pronounced as stops [b], [d], [g]. Generally speaking, it is likely that fricativization of γ happened first in the 2nd c. BCE, then β by the first century CE, and then δ by the third or fourth century CE. It is likely that these changes happened earlier in Asia Minor (Anatolia) than they did elsewhere. For a video and written explanation, click here.


Even though δ was not yet a fricative in Roman Palestine during the time of the New Testament—it would have became a fricative in the Late Roman or Byzantine period—we pronounce it as the fricative [ð] in the "Living Koine" pronunciation for the sake of consistency. One should also note the importance of pronouncing these consonants as regular stops after nasals (e.g., ἀνδρός [andros]) to students, since this has phonological significance. Finally, for students who have trouble pronouncing β or (α)υ/(ε)υ as [β], they may substitute [v].

In Classical Greek, the voiceless aspirated stops were pronounced as [pʰ] (as in pool), [tʰ] (as in to), and [kʰ] (as in kept or as in cool). Note that there is a little "h" after each of these letters, which indicates that each consonant was pronounced with "aspiration". Aspiration essentially means that a little puff of air comes out after the main consonant sound. You can tell if a sound is aspirated or not by putting your hand up to your mouth as you say these words and seeing if you feel a puff of air. In the Koine period, these consonants were pronounced in the same way in many regions. In other regions, such as Asia Minor (Anatolia), these consonants shifted to fricatives, so that φ was pronounced as [f] (as in fool), θ was pronounced as [θ] (th as in thin), and χ was pronounced as [x] (ch as in Bach). This eventually spread to most all regions of the Greek-speaking world in the Byzantine period. In Roman Palestine, however, at the time of the New Testament, these consonants were pronounced as [pʰ], [tʰ], [kʰ].



In the "Living Koine" pronunciation used in the materials on this website, the later fricative pronunciations of [f], [θ], and [x] are used, even though that would not have been the original pronunciation of Roman Palestinian Koine. Pedagogically, it makes more sense to use the slightly later pronunciation (or Anatolian Roman pronunciation).

In Classical Greek, what would later come to be signified by the spiritus asper (῾) indicated the consonant [h] (as in have). In the Koine period, this consonantal /h/ began to fall out of the language. Some speakers kept it and other eliminated it. The evidence from Roman Palestine suggests that it the consonant was present in some speakers' language and absent in others.


We do not observe spiritus asper in the "Living Koine" pronunciation just like in Modern Greek. However, because some (even if only a minority of) speakers in Roman Palestine likely still preserved it during the first century CE, one may choose to pronounce spiritus asper as /h/ for pedagogical reasons. Because it does not interfere with any other phonemes, this may be a pedagogically helpful decisions when teaching and it is not ahistorical.

In Classical Greek, the voiced unaspirated stops were pronounced as [p] (as in happy), [t] (as in stop), and [k] (as in close or as in ski). Note that none of these consonants are pronounced with a puff of air after them in the example English words. This is not a distinction English speakers are typically aware of, but if you stop and pay attention, you will notice it. Note that aspiration is the only difference between π, τ, κ and φ, θ, χ in ancient Greek (before φ, θ, χ became fricatives). These consonants remained unchanged from the Classical period to the Koine period. It should be noted that after nasal consonants these stops were pronounced voiced (i.e., μπ = [mb], ντ = [nd], γκ = [ŋg]).


One should encourage their students to pronounce π, τ, κ without aspiration even though this may be a bit unnatural for English speakers. Even though the aspirated-unaspirated distinction between φ, θ, χ and π, τ, κ is not preserved in the "Living Koine" pronunciation—since we pronounce φ, θ, χ as the fricatives [f], [θ], [x]—it will still be more "Greek" to pronounce π, τ, κ without aspiration. One should also note the importance of pronouncing these consonants with voicing after nasals (e.g., ἀντί [andi]) to students, since this has phonological significance.

In Classical Greek, the nasals μ and ν were pronounced as [m] (as in man) and [n] (as in now) and the liquids λ and ρ were pronounced as [l] (as in lot) and [r] (as in right). During the Koine period, all of these consonants undergo weakening to some degree, with the greatest weakening occurring with the nasals. The result was that sometimes a nasal would either be left out of the word entirely or assimilate and double the following consonant: e.g., λαμβάνω [labano] or [labbano].


One might experiment occasionally with weakening nasals in fast speech in a controlled exercise to make students aware of such a phenomenon. However, it is more pedagogically appropriate to always pronounce these consonants clearly, even though they were weakened in the Koine period.

In Classical Greek, the sibilant σ was pronounced as [s] (as in soon) generally but as [z] (as in zoo) before voiced consonants: e.g., σοφός [sopʰos], but κόσμος [kozmos]. This has remained unchanged during the Koine period, although it has undergone some weakening in some environments. The consonantal grapheme ζ originally indicated two consecutive consonants, whether [zd], [dz], [sd], or [ds]. In the Koine period, this was simplified to [z]. There is some evidence in regional varieties of Koine that, after its simplification to [z], ζ could also be pronounced as σ before voiceless consonants.


It is important to emphasize to students that σ had a voiced allophone of [z] before voiced consonants. this has phonological significance and should be considered an integral part of the pronunciation.

In Classical Greek, the grapheme ξ represented a combination of κ and σ (as in execute) and ψ represented a combination of π and σ (ps as in lapse). In the Koine period, however, specifically in Palestine, there is evidence that ξ actually represented the combination χσ and that ψ represented the combination φσ. The presence of an aspirated stop rather than an aspirated one in these combinations in Palestine is possibly due to bilingual interference of Hebrew/Aramaic, in which the voiceless stops were pronounced with aspiration.


Despite the peculiarity of Roman Palestinian Koine with respect to ξ and ψ, it is pedagogically appropriate to encourage students to treat ξ as κσ and ψ as πσ.

In Classical Greek, there existed a phonemic contrast between single consonants and double consonants. A double consonant was probably held from anywhere between 1.5–2.5x the length of a single consonant. Early on in the Koine period, this distinction was eliminated from the language.


There is no reason to pronounce gemination when teaching Koine Greek. It would be artificial.




In addition to providing an overview of the spelling changes in Koine Greek papyri and epigraphy during the Koine period, Buth also deals with a number of pertinent questions regarding the pronunciation and its relationship to pedagogy:

ἡ Κοινὴ Προφορά | Koiné Pronunciation


While a number of works on Hellenistic/Koine Greek pronunciation based on the Egyptian papyri are helpful (e.g., Edwin Mayser, Grammatik der griechischen Papyri aus der Ptolemaerzeit, 8 vols. [1906]; Sven-Tage Teodorsson, The phonology of Ptolemaic Koine [1977]; Geoffrey Horrocks, Greek: A History of the Language and Its Speakers, 2nd ed., pages 160–72 [2014]), I recommend the following as the most helpful and comprehensive:

Gignac, Francis Thomas. 1976. A Grammar of the Greek Papyri of the Roman and Byzantine Periods. Milan: Istituto Editoriale Cisalpino - La Goliardica.

Painting of the Pyramids


In addition to Randall Buth's paper mentioned above, a fairly extensive analysis of Palestinian Koine Greek phonology from the Hellenistic period to the Byzantine period based on the Palestinian Greek epigraphy and documentary texts is found in chapter 4 (pages 97–131) of the following Ph.D. thesis (note also the appendix of spelling interchanges on pages 363–85):

Kantor, Benjamin. 2017. “The Second Column (Secunda) of Origen’s Hexapla in Light of Greek Pronunciation.” diss., The University of Texas.