Diatopic and Diachronic Variations in the Pronunciation of Ancient Greek

“Ptolemy II Philadelphus talking with Jewish scholars in the library of Alexandria and having the Septuagint explained to him,” by Jean Baptiste de Champaigne

Any attempt to explain the historical pronunciation of Ancient Greek must take into account that the pronunciation varied both according to time and place and even within the same time and place. Luke Ranieri, in response to this problem, has published a much-circulated spreadsheet of the different variations. While Luke’s spreadsheet has been very helpful and justly lauded, more diatopic specificity would have, in my judgement, made more apparent the general phonological conservatism of Hellenistic and Roman-period Koine Greek. Below, therefore, is my own spreadsheet, which follows the general format of Luke’s while giving more attention to diatopic variety.

Because it was designed for would-be speakers of reconstructed historical pronunciations, I have avoided the inclusion of periods the pronunciation of which likely included difficult-to-pronounce transitionary sounds. Also excluded are alternative systems which seem to me not to rise to the level of majority systems (e.g. Teodorsson’s “innovative systems”). Moments of change are marked with bold typeface. In the case that one grapheme has two indicated pronunciations within the same system, the one listed first is thought to be the most common. The chart presupposes some familiarity with the IPA, but all sounds link to representative audio. * indicates that the linked audio is of a word containing the sound rather than the sound in isolation; ** indicates that the linked audio only approximates the correct sound. Audio links do not capture quantitative vowel distinctions. For more on the IPA, watch this and this.

For the sake of brevity, I have not cited sources for any claims presented in the spreadsheet. Those seeking this information may email me.

A color-coded version may also be downloaded here.


Classical AtticaClassical AtticaPtolemaic EgyptHellenistic AtticRoman Attic Roman Judaea1Early Roman EgyptLate Roman Egypt
Late 5c B.C.Mid. 4c B.C.350-150 B.C.150-50 A.D.50-150 A.D.1-200 A.D.1-100 A.D.100-400 A.D.
α[a], [aː][a][aː][a][aː][a][aː][a][aː][a][aː][a][a]
ι[i], [iː][i][iː][i][iː][i][iː][i][iː][i][iː][i][i]
υ[y], [yː][y][yː][y][yː][y][yː][y][yː][y][yː][y][y]
η[ɛː][eː][eː] [eː] [eː] [eː] [e][i]
ει /_{<C>, #}[eː] (raised)[iː][iː][iː][iː][iː][i][i]
ει / _<α>[eː][eː][eː][eː][iː][iː]
αυ[au]*, **[au]*, **[au]*, **[au]*, **[au]*, **[aw]*[aw]*[aw]*
ευ[eu]*[eu]*[eu]*[eu]*[eu]*[ew]*, **[ew]*, **[ew]*, **
οι[oi]*, **[oi]*, **[oi]*, **[oi]*, **[oi]*, **[oi]*, **[y][y]
υι[yi]*, **[yː][yi]*, **[yː][yː][yː][y][y]
[eː] (as morphological ending?)
ηυ[εːu]*, **[eːu]*, **[eːu]*, **[eːu]*, **[eːu]*, **[eːu]*, **[ew]*, **[ew]*, **
[h] (?)[h] (?)
σ[s] (retracted)[s] (retracted)[s] (retracted)[s] (retracted)[s] (retracted)[s] (retracted)[s] (retracted)[s] (retracted)
γ+back vowel[g][g][g][g][g][g][ɣ][ɣ]
γ+front vowel[g][g][g][g][g][g][ʝ][ʝ]
γγ+back vowel[ŋg]*[ŋg]*[ŋg]*[ŋg]*[ŋg]*[ŋg]*[ŋg]*[ŋg]*
γγ+front vowel[ŋg]*[ŋg]*[ŋg]*[ŋg]*[ŋg]*[ŋg]*[ŋɻ][ŋɻ]
δ[d]**[d]**[d]**[d]**[d]**[d]**[d]**; [ð] / _ι<α>Before ~400 A.D.: [d]**[ð] / _ι<α>

Post 400 A.D.: [ð]
χ+back vowel[kh]*[kh]*[kh]*[kh]*[kh]*[kh]*[kh]*[kh]*
χ+front vowel[kh]*[kh]*[kh]*[kh]*[kh]*[kh]*[ch]**[ch]**
θ[th]*, ** (dental) [th]*, ** (dental)[th]*, ** (dental)[th]*, ** (dental)[th]*, ** (dental)[th]*, ** (dental)[th]*, ** (dental)[th]*, ** (dental)
κ+back vowel[k][k][k][k][k][k][k][k]
κ+front vowel[k][k][k][k][k][k][c][c]
ρ[r̥], [r], [ɾ][r̥][r][ɾ][r̥][r][ɾ][r̥][r][ɾ][r̥][r][ɾ][r̥][r][ɾ][r̥][r][ɾ][r̥][r][ɾ]
νκ, γκ[ŋk]*[ŋk]*[ŋk]*[ŋk]*[ŋk]*[ŋk]*[ŋk]*[ŋk]*
<σσ>~<σ> [C:][C:][C:]>[C] /[V]_[V][C:][C:][C:]>[C][C:]>[C][C:]>[C]
[C:][C:]>[C] /[V]_[V][C:]>[C] /[V]_[V]

Other information:

  • [y] was often confused with [u] and (less frequently) [i] in Egyptian Koine. This may be due to the absence of [y] from the Egyptian vowel system, or because ου = [u] was more common outside of the Ionic-Attic dialects at the beginning of the Hellenistic period.
  • The ι component of long diphthongs tended to become a soft glide when in intervocalic position.
  • σ was pronounced as ζ before voiced consonants. Final σ was dropped in the speech of many later Egyptian writers.
  • According to Gignac (1976: 102), for many writers in Roman and Byzantine Egypt there may have been only one liquid, [l]. Liquids may also have not been pronounced by some before or after stop consonants.
  • Gignac (1976: 86), “χ, θ, φ are frequently replaced by κ, τ, π after σ, before another aspirate, and before or after a liquid or nasal, with the converse occurring occasionally. This indicates that aspirated stops tended to lose their aspiration in these posi­tions, as attested at least dialectally elsewhere in Greek.”
  • Gignac (1976: 64), “In Egypt, the occasional unconditioned interchange of φ, χ, θ with π, κ, τ indicates the identification of aspirated stops with their corresponding voiceless stops in the speech of individual writers.”
  • Gignac (1976: 95), “The unconditioned interchange of aspirated and voiceless stops is caused by bilingual interference. Only in the Bohairic dialect of Coptic, spoken in the Delta area, were there aspirated stop phonemes. Even in this dialect the op­position between voiceless and aspirate occurred only in accented syllables, and the aspirates were lost in late Byzantine times.”
  • [e] sounds were likely all slightly lowered.
  • Due to bilingual interference, many speakers in Judaea and Egypt likely pronounced σ=[s].
  • νγ, νκ, νξ, νχ = γγ, γκ, γξ, γχ should apply across word-boundaries (e.g. τὸγ γραμματέα for τὸν γραμματέα) (see Allen (1987: 34).
  • The collapse of pitch accentuation and its replacement with a stress-based system seems to have been linked to the collapse in quantitative vowel distinctions (Allen 1989: 130). It is likely, however, that slight stress was used to emphasize the pitch accent in the Hellenistic and early-Roman periods, especially by non-native speakers whose first language made use of stress accentuation (esp. in Egypt).


Allen, W.S. (1987) Vox Graeca, Cambridge.
Allen, W.S. (1976) “Long and Short Diphthongs: Phonological Analogies and Phonetic Anomalies,” in Anna Morpurgo Davies and Wolfgang Meid (eds.) (1976) Studies in Greek, Italic, and Indo-European Linguistics Offered to Leonard R. Palmer, Innsbruck, 9-16.
Evans, D. Emrys (1918) “Notes on the Consonants in the Greek of Asia Minor” in Classical Quarterly 12, 162-70.
Gignac, Francis T. (1976) A Grammar of the Greek Papyri of the Roman and Byzantine Periods: Vol. 1 Phonology, Milan.
Horrocks, Geoffrey (2014) Greek: A History of the Language and its Speakers, Oxford.
Kantor, Benjamin (2017) “The Second Column (Secunda) of Origen’s Hexapla in Light of Greek Pronunciation” Ph.D dissertation, University of Texas at Austin.
Ruijgh, C.J. (1978) Review of S.-T. Teodorsson (1974) in Mnemosyne 31, 79-89.
Teodorsson, S.-T. (1974) The Phonemic System of the Attic Dialect 400-340 BC, Göteborg.
Teodorsson, S.-T. (1978) The Phonology of Attic in the Hellenistic Period, Göteborg.
Teodorsson, S.-T. (1977) The Phonology of Ptolemaic Koine, Göteborg.
Threatte, Leslie (1980) The Grammar of Attic Inscriptions I: Phonology, Berlin.


  1. While I have used the data of Kantor (2017), I have not followed his analysis.

Leave a Reply