Phonology of the Koine

The phonology of the Koine (also Koinē , between 300 BC and 600 AD) developed, as linguists have established, during a period in which profound changes have taken place: While the pronunciation of the Koine at the beginning was almost similar to that of Ancient Greek, differed at the end of this it differs only in a few points from that of modern Greek .

The following applies to the local reconstructed pronunciation information in tabular form:

The spelling at that time is listed after the pronunciation in IPA .
The terms “rounded” and “unrounded” in vowels refer to the presence of rounding .
The abbreviations "stl." And "sth." For consonants mean "voiceless" or "voiced".
By clicking on a Greek character you get to the article about this character, by clicking on a sound you get to the article about this sound.

overview

Almost all important sound shifts during the koine concerned vowels: the distinction between long and short vowels was abandoned, the tonal accent and the ancient Greek diphthongs became monophthongs (with the exception of the diphthongs αυ, ευ and ηυ, which developed into vowel-consonant combinations) . In all likelihood, these shifts began in the 2nd century BC in Egyptian and ended in the 2nd century AD in Attic Greek.

Another series of changes concerned the shift from voiced and voiceless-aspirated plosives and the u-sounds of the diphthongs αυ, ευ and ηυ to fricatives, which most likely also began in Egyptian Greek of the 1st century BC and during the transition of the Koine to the Byzantine in the form of general independence came to an end.

Reconstruction problems

If you want to reconstruct the pronunciation of the koine, you have a relatively wide range of clues. Nevertheless, the size of the Greek-speaking area at that time is a problem: indications always indicate that changes occurred at different times in different places (e.g. differences between sociolects ). Therefore, one has to take into account that some of the sound changes that were supposedly so typical of the Koine had partly already occurred in other dialects before, but did not prevail before about 300 AD: If you bear in mind that Let later debates derive from earlier dialectals (which is, however, still discussed in discussion) and did not run independently in parallel, it would sometimes take up to a thousand years before they were generally accepted; and even assuming the opposite, there are gaps of up to two centuries. The problem is that not every Greek dialect could and cannot be reconstructed because there is too little evidence and evidence for this. Therefore, a pronunciation that you could reconstruct at a certain time in a certain place is not necessarily the generally valid one.

Many documents are simple typographical errors; however, the spelling may have remained conservative, and the first spelling mistakes may not have occurred until long after the actual coincidence of sounds.

If you have spelling mistakes from different sources, e.g. B. Egyptian papyri and Attic inscriptions, one often comes to different dates of similar sound changes. There are, however, many explanations for the emerging conservatism of the formal Attic inscriptions compared to the Egyptian papyri. A first would be dialectal deviations with influences from foreign languages; in this case the changes in Egyptian Greek would have occurred earlier than those in Attic. Another is that the upscale Attic language was more conservative than the Egyptian Greek; the formal language would then have retained linguistic peculiarities rather than the vulgar language. The third would be that the Attic orthography has remained rather more conservative than that of Egyptian Greek; in this case the sound changes would not have occurred at different times, but attention would have been paid to preserving the older orthography in Attic than in Egyptian Greek. All of these theories are plausible to a certain extent, but would lead to different dating of the changes.

Descriptions by grammarians and, to a lesser extent, transcriptions into other languages are further evidence, for they at least indicate what pronunciation was the standard of scholarship; however, it is also possible that it was more about conventional transliterations than phonetic transcriptions.

Pronunciation of speakers between the 1st and 4th centuries AD

Until late antiquity (from around 300 AD), learned Greek lecturers seem to have used a consciously conservative pronunciation based on the Attic dialect (which tends to be less evolving) (cf. Atticism ). In Attic itself, this is the "real" ancient Greek pronunciation until the early 2nd century AD, after which it declined relatively quickly.

The following pronunciation is consistently atticistic, with the exception of the diphthong [ yi ], which had already been monophthongized in Attic, and [ ɔːu̯ ], which did not exist at all in classical Attic, but was cultivated by some editors of the early Koine.

Vowels

Short vowels
	Front		Back
	Unrounded	Rounded	Rounded
Closed	[ i ] ι / ῐ	[ y ] υ / ῠ
medium	[ e ] ε		[ o ] ο
Open	[ a ] α / ᾰ

Long vowels
	Front		Back
	Unrounded	Rounded	Rounded
Closed	[ iː ] ι / ῑ , ει	[ yː ] υ / ῡ	[ uː ] ου
medium	[ ɛː ] η , ει		[ ɔː ] ω
Open	[ aː ] α / ᾱ

The pseudo-diphthong ει had the same sound value before vowels as the Eta, in all other cases like the long Iota.

Diphthongs
	Front end	Back end
Short first vowel	[ai̯], [oi̯], [yi̯] αι, οι, υι	[au̯], [eu̯] αυ, ευ
Long first vowel	[aːi̯], [ɔːi̯] ᾳ, ῳ	[aːu̯], [ɛːu̯] [ɔːu̯] αυ, ευ, ωυ

Diphthongs, in which the first vowel is short, were usually written in brackets, as they have been diphthongized gradually since the beginning of the koine. Speakers from the higher classes of the early koine probably tried to preserve these diphthongs, but by the 1st century BC this process of monophthongization was complete.

Consonants

Plosives
	Bilabial	Alveolar	Velar
Voiced	[ b ] β	[ d ] δ	[ ɡ ] γ
aspirates voiceless	[ pʰ ] φ	[ tʰ ] θ	[ kʰ ] χ
Unvoiced	[ p ] π	[ t ] τ	[ k ] κ

According to information from ancient grammarians and transcriptions, the plosives remained until around 300 AD.

Other consonants
	Bilabial	Alveolar		Velar	Glottal
	sth.	stl.	sth.	sth.	stl.
Nasals	[ m ] μ		[ n ] ν	([ ŋ ]) ( γ )
Vibrants			[ r ] ρ , [rʰ] ῥ
Fricatives		[ s ] σ	[ z ] ζ , σ		[ H ]
lateral approximants			[ l ] λ

The fact that [ŋ] is in brackets stems from the fact that only some scholars regard it as a phoneme in its own right, while others regard it as an allophone of [n].

At the beginning of the word, the ρ carried the spirit asper , ῥ; the sound value of this combination is unclear: it was probably an allophone of / r /, but it is also not known which allophone it was. The most likely are the voiceless alveolar vibrant [ r̥ ] or an aspirated sound [ rʰ ].

The zeta is used to represent the gemination [zː].

Boeotian, 4th century BC

Although it belongs to the late classical period rather than to the Koine, Boeotic is called here because it was relatively progressive and corresponds well to a first intermediate stage between the classical and today's Greek.

In the 4th century BC almost all diphthongs in Boeotian were monophthongized and the plosive-to-fricative shift had already begun with the gamma.

It is also important to mention that, in contrast to the Ionian-Attic and the Koine, the Ypsilon remained a back-tongue vowel, so it was pronounced rather "Upsilon" (even before the Classical period, [u (ː)] became [y (ː)] become).

There are no indications as to whether the vowel lengths were still differentiated in Boeotian or not; it is not unlikely that the task of distinguishing between long and short vowels was on the same level as that of monophthonging the diphthongs, but it may just as well not be the case.

First theory: vowel length was no longer differentiated

Short vowels
	Front		Back
	Unrounded	Rounded	Rounded
Closed	[ i ] ι , ει	[ y ] (?) οι , υι	[ u ] υ , ου
Almost closed	[ ɪ ] η
Half closed	[ e ] ε		[ o ] ο
Half open	[ ɛ ] αι		[ ɔ ] ω
Open	[ a ] α

It is questionable whether the sounds given as y-sounds in this table were really such.

Second theory: vowel lengths were still differentiated

Short vowels
	unrounded front	rounded at the back
Closed	[ i ] ι / ῐ	[ u ] υ / ῠ
medium	[ e ] ε	[ o ] ο
Open	[ a ] α / ᾰ

Long vowels
	Front		Back
	Unrounded	Rounded	Rounded
Closed	[ iː ] ι , ει	[ yː ] (?) οι , υι	[ uː ] υ , ου
Half open	[ eː ] η
Half closed	[ ɛː ] αι		[ ɔː ] ω
Open	[ aː ] α

Consonants

Plosives
	Bilabial	Alveolar	Velar
Voiced	[ b ] (?) β	[ d ] (?) δ	[ ɣ ] γ
aspirates voiceless	[ pʰ ] (?) φ	[ tʰ ] (?) θ	[ kʰ ] (?) χ
Unvoiced	[ p ] π	[ t ] τ	[ k ] κ

Fricative pronunciation of beta, delta, phi, theta and chi are not documented for Boeotic, but would not be atypical, as the fricative pronunciation of theta in Laconic is documented as early as the 5th century BC.

Other consonants
	Bilabial	Alveolar		Velar	Glottal (?)
	sth.	stl.	sth.	sth.	stl.
Nasals	[ m ] μ		[ n ] ν	([ ŋ ]) ( γ )
Vibrants			[ r ] ρ ,? ῥ
Fricatives		[ s ] σ	[ z ] ζ , σ		[ H ] (?)
lateral approximants			[ l ] λ

It is questionable whether the Spiritus asper was still pronounced; the same remarks apply to the other consonants as to the attic pronunciation.

Egyptian Greek, around the turn of the times

From the 2nd century BC on, the diphthongs in Egyptian Greek had been monophthonged and the distinction between vowel lengths had been abandoned.

Vowels

	Front		Back
	Unrounded	Rounded	Rounded
Closed	[ i ] ι , ει	[ y ] υ , οι , υι	[ u ] ου
Half open	[ e ] η		[ o ] ο , ω
Half open	[ ɛ ] ε , αι
Open	[ a ] α

It is controversial whether there were still closing diphthongs in Egyptian Greek at this point in time, possibly intermediate stages with [aw], [ɛw] and [iw] were reached.

Consonants

Plosives
	Bilabial	Alveolar	Velar
Voiced	[ b ] β	[ d ] δ	[ ɡ ] γ
aspirates voiceless	[ pʰ ] φ	[ tʰ ] θ	[ kʰ ] χ
Unvoiced	[ p ] π	[ t ] τ	[ k ] κ

There is slight evidence that beta and gamma could already be fricatives, but they are not concrete. The fricative pronunciation of the aspirated sounds probably only became established later.

Other consonants
	Bilabial	Alveolar		Velar	Glottal (?)
	sth.	stl.	sth.	sth.	stl.
Nasals	[ m ] μ		[ n ] ν	([ ŋ ]) ( γ )
Vibrants			[ r ] ρ ,? ῥ
Fricatives		[ s ] σ	[ z ] ζ , σ		[ H ] (?)
lateral approximants			[ l ] λ

It is questionable whether the Spiritus asper was still pronounced; the same remarks apply to the other consonants as to the attic pronunciation.

4th century AD

In the 4th century, the non-differentiation of the diphthongs had already prevailed. The Eta has often been confused with Iota, but sometimes with Epsilon. The fricative pronunciation of the former plosives had already established itself, only in a few dialects the plosive pronunciation seems to have lasted until the turn of the millennium.

Vowels

	Front		Back
	Unrounded	Rounded	Rounded
Closed	[ i ] ι , ει , η	[ y ] υ , οι , υι	[ u ] ου
Half open	[ e ] ε , partly η		[ o ] ο , ω
Half open	[ ɛ ] αι
Open	[ a ] α

The confusion of the y and i sounds had already begun in the 2nd century, but was probably not yet generally accepted.

The collapse of the alpha-iota digraph with epsilon was undoubtedly under way, but there is still no evidence of its completion.

Consonants

Plosives and former plosives
	Bilabial / Labiodental	Alveolar / Dental	Velar
Voiced	[ β ] or [ v ] β (υ)	[ δ ] δ	[ ɡ ] γ
aspirates voiceless	[ φ ] or [ f ] φ	[ θ ] θ	[ x, ç ] χ
Unvoiced	[ p ] π	[ t ] τ	[ k ] κ

Other consonants
	Bilabial	Alveolar		Velar
	sth.	stl.	sth.	sth.
Nasals	[ m ] μ		[ n ] ν	([ ŋ ]) ( γ )
Vibrants			[ r ] ρ
Fricatives		[ s ] σ	[ z ] ζ , σ
lateral approximants			[ l ] λ

Phonetic explanation of the above

Loss of vowel length

While in the classical period a consistent distinction was made between long and short vowels, this was gradually abandoned so that all vowels were pronounced the same length over time.

From the 2nd century BC onwards, spelling mistakes in Egyptian papyri indicate the loss of this distinction and the tonal accent in favor of a pure stress accent . The widespread confusion between omicron and omega in Attic inscriptions since the 2nd century AD also indicates a loss of the distinction between the quantity of vowels; However, it can also be that the sounds were qualitatively coincident (i.e. that Omega continued [ɔː], but Omikron was already pronounced [ɔ]).

For phonological reasons, however, this transition probably went hand in hand with the change from the tonal to the pure accent, which was generally accepted from the 3rd century AD.

Shift to the stress accent

The type of accentuation of words changed from a musical to a stress accent, which means that the accented syllable is no longer pronounced with different high tones, but louder and / or stronger.

In addition to the above-mentioned references in Egyptian Greek, there are others in poetry from the 2nd century AD.

Diphthongs

Pseudo-diphthongs

Before consonants, the diphthong <ει> in Attic was probably monophthonged from the 6th century BC and was probably pronounced as a long < ε̄ > [ eː ]. From the 4th century BC, this pseudo-diphthong , which was now also used for words that etymologically did not contain <ει>, shifted to ῑ , probably as [ iː ], which already had the same quality as today.

Before vowels, the diphthong <ει> developed differently: One theory to explain this is that the diphthong remained a diphthong of the form [ej] until the [ j ] as a smooth transition of the e- Loud was felt to the next word. From the late 4th century BC on, the diphthong ει was also confused with the simple Eta, which indicates that, as not before a consonant, the Eta had assumed a kind [ eː ] as a sound value, which in turn indicates that the Eta was gradually closed; see below for details.

From the 6th century BC, the diphthong <ου> was gradually monophthongized and sometimes confused with the simple omicron ο̄. Although its original sound value was probably [ oː ], it developed relatively quickly after [ uː ] (probably around 350 BC). At least as far as the vowel quality is concerned, this is the same sound value as it is today.

Diphthongs with a short first component

The diphthong <αι> was probably first monophthongized as [ ɛː ]. This indicates that Boeotian is consistently written as Eta. Confusion of <αι> and simple epsilon indicates a shift in Egyptian Greek during the 2nd century BC. However, at least in the language taught there must still have been an <αι>, since it is transcribed into Latin with <ae> and this in turn with <αι> into Greek. Further confusions between <αι> and <ε> are in Palestine in the early 2nd century BC and from around 125 BC. BC in the Attic, which indicates that this shift occurred from the late 2nd century BC in the Attic. Allen assumes that the transition to [ eː ] happened later; he is not very precise on this point, but it seems that his theory is based on the confusion of <αι> with epsilon and Eta with epsilon, but not <αι> and Eta. However, not all scientists agree.

The diphthong <οι> monophthongized to [ y (ː) ]. This is evidenced by the confusion of <οι> and <υ> in Boeotian, but probably only a dialectal shift. However, at least in the language taught, there must still have been an <οι>, since it was transcribed into Latin with <oe> and this in turn into Greek with <οι>. Further evidence of monophthongization has come down to us in the 1st century BC in Egyptian Greek and in the 2nd century AD in Palestine. It is definitely documented in the second century AD, when <υ> is written for <οι>.

In the Koine, the diphthong <υι>, which was probably first monophthonged to [ yː ] in the 6th century BC in the Attic and until the 4th century BC in all other dialects , seems to have been preferred.

Closing diphthongs with a short first vowel

The diphthongs <αυ> and <ευ> lost their ancient sound values of [ au ] and [ eu ] and were given a fricative pronunciation like [ aβ ] or [ eβ ] or [ av ] or [ ev ]. Confusion of <αυ> or <ευ> with <αβ> or <εβ> is documented in Egyptian papyri from the turn of the times, which indicates the fricative pronunciation. However, it took some time for this debate to prevail; for example, Jewish catacomb inscriptions still show diphthong pronunciation in the second and third centuries AD. Confusion of diphthongs with <αβ> or <εβ> becomes normal from the transition from the sixth to the 7th century.

Opening diphthongs with a long first vowel

The diphthong ῃ was monophthonged in Attic from the 4th century AD, as it was often written <ει> and probably pronounced [ eː ]. Therefore they developed further in the Koine and became [ iː ]. However, these changes had probably not taken place in some inflected endings and there was further pronounced and written [ eː ].

The other closing diphthongs with a long first vowel, ᾳ and ῳ , became monophthongs from the 2nd century BC, when they were written as simple alpha and omega, respectively. They were probably pronounced [ aː ] or [ ɔː ] (if the [ ɔː ] had not become [ oː ] for a long time ; see also the treatise on the further development of Omikron and Omega).

Closing diphthongs with a long first vowel

If it was an extension of the <ευ> in verbs, the diphthong <ηυ> became <ευ> from the 4th century BC.

The other closing diphthongs with a long first vowel (< ᾱυ >, <ηυ> and <ωυ>, which did not exist in classical Attic) became monophthongs from the 1st century BC and became <α>, <η> or <ω> written; the first was probably pronounced [ aː ], the other two [ ɛː ] or [ ɔː ], if the openness had not already been lost (otherwise [ eː ] or [ oː ]), and possibly later to [ iː ] or [ oː ] (see also the treatise on the further development of Eta and Omega).

Quality of the monophthongs

If you disregard the Eta, the sound value of the monophthongs has remained closer to the ancient "original" than that of the diphthongs.

As mentioned above, the pseudo-diphthongs <ει> and <ου> at the beginning of the koine had sound values of [ iː ] (only before consonants) and [ uː ], respectively , which have not improved qualitatively until today . The diphthong <ει> had already become [ e war ] as an intermediate step , like the simple eta, which then shared their further development.

The qualities of the vowels alpha and iota, apart from the eventual loss of their length, have not developed any further and have remained up to the present day [ a ] and [ i ]. The simple epsilon opened up a bit over time and changed from [ e ] to [ ε ].

Omicron and omega were regularly confused in Attic inscriptions from the 2nd century AD, which could indicate that the qualitative distinction was lost from around this time. An alternative interpretation, however, is that the vowel quantity coincided at this point in time and the qualitative distinction was perhaps even omitted earlier. In fact, some Attic inscriptions from the 4th century BC show certain, albeit less frequent, confusions of the two vowels, which could possibly indicate that the qualitative distinction was no longer applicable at this point in time and the quantitative distinction was not made until the 2nd century AD.

The qualitative distinction between Eta and Epsilon was probably abandoned in Attic in the 4th century AD, since from this point in time the pseudo-diphthong <ει> was confused pre-consonant with Iota and pre-vowel with Eta. From around 150 AD, Eta and Iota are confused with each other in Attic inscriptions, which is known as a shift of the Etas after [ i (ː) ] (depending on whether a distinction was still made between long and short vowels), which is still used today Standard is, can be interpreted. However, at least some educated people seem to have pronounced the Eta further as [ eː ] or [ e ], since Eta is confused with Epsilon in some Attic inscriptions and is rendered with “e” in transcriptions into Gothic or Armenian. At least in the case of Gothic and Armenian, however, a regional dialectal difference or influence can also be used as an explanation: Even in modern Pontic Greek , <η> is pronounced as / e /.

The Koine had adopted the pronunciation [ y (ː) ] from the Ionian-Attic for the Ypsilon . It was first confused with Iota in the 2nd century AD in Egyptian papyri, which suggests the pronunciation as [ i (ː) ], but this seems to have been a regional development. Transcriptions into Gothic still indicate the ü pronunciation; it is believed that the i-pronunciation did not prevail until the turn of the millennium.

Loss of aspiration

The aspiration (the phoneme / h / , which is usually marked with the Spiritus asper ), which had already been partially lost in the Ionian of Asia Minor and the Aeolian of Lesbos (psilosis), later also disappeared from the koine. Spelling mistakes in Egyptian Greek indicate that this loss was already underway in the 1st century BC. Transcriptions into other languages and consonant changes before aspirations suggest that it had not yet established itself in the 2nd century AD, but was at least very much on the way in the 4th century AD.

Consonants

Probably beta, gamma, phi, theta, and zeta were the only consonants that changed since the classical period. The delta and (less likely) the chi will likely have changed too; however, there is no concrete evidence that this happened in the time of the Koine.

The zeta, which was probably pronounced [ zd ] in Classical Attic (although some scholars assume it was more like [ dz ] and the sound value differed from dialect to dialect), over time became [ z ], like it is still pronounced today; however, it seems that it was pronounced , at least intervowelly, geminated as [ zː ]. Attic inscriptions indicate ^{[explicitly?] That} this pronunciation was already general towards the end of the 4th century BC.

The digraph -σσ- corresponds to the Attic -τ τ- in the koine.

Phi and Theta, which were originally pronounced aspirated [ pʰ ] and [ tʰ ], became fricatives [ f ] and [ θ ] , respectively . On the other hand, there is no concrete evidence that the chi changed from [ kʰ ] to [ x ] or [ ç ] during the koine . There is evidence of a [ θ ] pronunciation of theta in Laconic in the 5th century BC, but it is unlikely that this influenced the koine, which was largely based on the Ionian-Attic. The first clear evidence of the fricative pronunciation of the theta and phis in the koine comes from Pompeian inscriptions from the 1st century BC. Nevertheless, inscriptions from Palestine in the early 2nd century tend to place the aspirated pronunciation of the theta and Jewish catacomb inscriptions from the 2nd and 3rd centuries AD [ f ] for the phi, [ tʰ ] for the theta and [ kʰ ] for the Chi is close, which indicates that the fricative pronunciation of theta was not yet general at that time and the shift in phi occurred faster than that of the other two letters. Armenian transcriptions reproduce the Chi as [ kʰ ] until the 10th century AD , which can be interpreted to mean that it was at least partially (dialectically?) Pronounced as such until then.

It is unknown when beta, delta and gamma, which were originally pronounced as [ b ], [ d ] and [ ɡ ], respectively, became [ v ], [ ð ] and [ ɣ ] as they are pronounced today . Although some references to the fricative pronunciation of the gamma behind front tongue vowels have been documented as far back as the 4th century BC, this does not seem to have been the standard at that time. Ancient grammarians describe these letters as plosives, the beta is represented in Latin with b instead of v, and Cicero also clearly assigns the Latin b to the letter. Evidence from non-literary papyri indicates a fricative pronunciation in some contexts (mostly intervowel) from around the turn of the century, which, however, did not prevail. Confusion of <αυ> and <ευ> with <αβ> or <εβ> becomes normal from the late fifth and early sixth centuries, so that one can assume that the fricative beta had prevailed at this point in time. Nevertheless, in Armenian transcriptions up to the 10th century there are renditions of the beta as [ b ]; in other words, there may have been some conservative (or dialect) speakers to date who have pronounced the beta as [ b ]; however, it may also be that this rendering was a learned rule.

literature

Sidney Allen: Vox Graeca: the pronunciation of Classical Greek . 3. Edition. University Press, Cambridge 1987. ISBN 0-521-33555-8 .
Michel Lejeune : Phonétique historique du mycénien et du grec ancien . 2nd Edition. Éditions Klincksieck, Paris 1972.

Individual evidence

↑ The second seems to be correct in most cases, but the other two may have occurred in other cases as well.

↑ Allen, Vox Graeca . P. 94, note 9.

↑ ^a ^b Allen, Vox Graeca . P. 94.

↑ Allen, Vox Graeca . P. 130.

↑ Allen: op. Cit. Pp. 69-72. The diphthong <ει> was already in the fifth z. B. in Argos or in the 4th century BC in Corinth coincided with <ι>.

↑ ^a ^b Allen: op. Cit. Pp. 72-73.

↑ This sliding transition theory would explain why, although there was no pre-vowel < ε̄ > to confuse it with, it was often written as a simple epsilon. Allen: op. Cit. Pp. 83-84.

↑ Allen: op. Cit. Pp. 75-78.

↑ Such spellings as “πης” instead of “παις” indicate that this shift occurred in Boeotian, but not yet in Attic. (Allen: op.cit. P. 74.)

↑ ^a ^b ^c ^d Randall Buth: Ἡ Κοινὴ Προφορά . P. 3.

↑ ^a ^b Allen: op. Cit. P. 79.

↑ Allen: op. Cit. P. 79. The transition would then have occurred after the shift of the Etas to [ i (ː) ] in most of the other varieties of Greek, but this cannot be assumed before about 550 AD.

↑ This may have happened with the intermediate levels [ øj ] and [ øː ], which could at least partially explain the German spellings oe and ö.

↑ ^a ^b Allen: op. Cit . P. 81.

↑ Allen: op. Cit. P. 81, note 54.

↑ Comparable to the modern pronunciation [ av ] or [ ev ] (the partial assimilation to [ af ] or [ ef ] before the voiceless consonants theta, kappa, xi, pi, sigma, tau, phi, chi, psi is still undated).

↑ Buth: op. Cit. P. 4, p. 8, where Francis Thomas Gignac: A Grammar of the Greek Papyri of the Roman and Byzantine Periods. Volume One: Phonology. Milan 1976 . P. 68, note 1, and page 70. is cited.

↑ Allen: op. Cit. P. 80, note 47.

↑ Buth: op. Cit. P. 4, note 8, where Geoffrey Horrocks : Greek, A History of the Language and its Speakers. P. 111. cited.

↑ ^a ^b Note that the iota subscriptum is a medieval invention; The Iota adscriptum is used consistently in ancient texts.

↑ Allen: op. Cit. Pp. 85-86.

↑ Allen: op. Cit. P. 86. However, when arguing on the basis of <οι> in verbs, <ῳ> became <οι> instead (Allen: op. Cit. P. 87, note 70).

↑ Allen: op. Cit. P. 87, note 70.

↑ Allen: op. Cit. P. 87.

↑ Allen: op. Cit. P. 94.

↑ Allen: op. Cit. P. 73. At that time, this development had probably taken place in Boeotian, but definitely not in Attic. B. can be seen from the fact that in Boeotian "πατειρ" ( "pateir" ) in Attic, on the other hand, <πατήρ> ( "patér" ) was written (Allen: op. Cit. P. 74).

↑ Allen: op. Cit. Pp. 74-75.

↑ Allen: op. Cit. P. 68.

↑ Allen: op. Cit. P. 68, note 14.

^ Lejeune: Phonétique historique du mycénien et du grec ancien . Pp. 281-282.

↑ Randall Buth: op. Cit. Pp. 5-6, where Gignac: op. Cit. Pp. 137-138. is quoted.

↑ Allen: op. Cit. P. 53.

↑ Allen: op. Cit. P. 56 and 58, note 115.

↑ Allen: op. Cit. P. 58.

↑ Allen: op. Cit. P. 13 f.

↑ Some assume that this happened with the intermediate stage [ ɸ ], but there is no concrete evidence for this. (Allen: op.cit. P. 25.)

↑ Eg Aristophanes: " Εἰρήνη ", line 214, " σιώ " instead of " θεώ ". (Allen: op.cit. P. 26.)

↑ Here the transcription of “ λάσ θ η ” as “las f e” should be mentioned. (Allen: op.cit. P. 23.)

↑ Randall Buth: op. Cit. P. 4

↑ Allen: op. Cit. P. 24.

↑ Allen: op. Cit. P. 25.

↑ In some cases, [ β ] is assumed to be an intermediate stage (no further source citation)

↑ Except when preceded by a nasal consonant (gamma, My or Ny); in this case they keep their original phonetic value (e.g. γαμβρός = [ ɣambr'ɔs ], άνδρας = [ 'andras ], άγγελος = [ ' aŋgelos ]).

↑ Allen: op cit . P. 31 f.

↑ Allen: op. Cit. P. 31.

↑ Allen: op. Cit. P. 32, note 46.

↑ Randall Buth: op. Cit. P. 4, note 8, where Horrocks: op. Cit. P. 111. is quoted

↑ Allen: op. Cit. P. 32, note 45.

[1] The second seems to be correct in most cases, but the other two may have occurred in other cases as well.

[2] Allen, Vox Graeca . P. 94, note 9.

[allen94-3] Allen, Vox Graeca . P. 94.

[4] Allen, Vox Graeca . P. 130.

[5] Allen: op. Cit. Pp. 69-72. The diphthong <ει> was already in the fifth z. B. in Argos or in the 4th century BC in Corinth coincided with <ι>.

[allen72-73-6] Allen: op. Cit. Pp. 72-73.

[7] This sliding transition theory would explain why, although there was no pre-vowel < ε̄ > to confuse it with, it was often written as a simple epsilon. Allen: op. Cit. Pp. 83-84.

[8] Allen: op. Cit. Pp. 75-78.

[9] Such spellings as “πης” instead of “παις” indicate that this shift occurred in Boeotian, but not yet in Attic. (Allen: op.cit. P. 74.)

[buth3-10] Randall Buth: Ἡ Κοινὴ Προφορά . P. 3.

[allen79-11] Allen: op. Cit. P. 79.

[12] Allen: op. Cit. P. 79. The transition would then have occurred after the shift of the Etas to [ i (ː) ] in most of the other varieties of Greek, but this cannot be assumed before about 550 AD.