Module talk:grc-pronunciation

Archives

• Archive of "Module talk:grc-pronunciation" (May 2014 – April 2018)
• Archive of "Template talk:grc-IPA" (July 2015 – October 2018)

ηυ

What do we know about how ηυ was pronounced? At μετηύδα (metēúda) it's being treated as a sequence of two vowels in hiatus, but considering it's just the augmented form of μεταυδάω (metaudáō), isn't it more likely to have been a diphthong? Shouldn't the 5th century BC pronunciation be /me.tɛ̌ːu̯.daː/ rather than /me.tɛː.ý.daː/, and the other chronolects altered accordingly? —Mahāgaja · talk 21:37, 8 February 2019 (UTC)

ηυ is definitely supposed to be transcribed as a diphthong, and I thought it was. It is tokenized as a diphthong by Module:grc-utilities. I'll look into what happened to change this. — Eru·tuon 20:33, 12 February 2019 (UTC)

@Erutuon: The accent placement in /hɛːú̯.ron/ for ηὗρον (hēûron) is not correct, right? —*i̯óh₁n̥C^[5] 22:11, 12 February 2019 (UTC)

@JohnC5: Whoops, fixed. — Eru·tuon 22:14, 12 February 2019 (UTC)

Circumflex notation

The circumflex pitch accent is marked with an acute accent instead of a circumflex accent over a diphthong. For example, the 5th century BC pronunciation of γαῖα (gaîa) is given as /ɡái̯.a/ rather than /ɡâi̯.a/. Other cases from Module:grc-pronunciation/documentation#Test cases include: Αὖλος /áu̯.los/, Θρᾷξ /tʰráːi̯ks/, ναῦς /náu̯s/, οἷαι /hói̯.ai̯/, ᾠδῇ /ɔːi̯.dɛ́ːi̯/. Over monophthongs the circumflex accent is shown correctly, e.g. πᾶς /pâːs/, πρᾶγμα /prâːŋ.ma/, σημεῖον /sɛː.mêː.on/, τμῆμα /tmɛ̂ː.ma/. Can this be fixed, please? —Mahāgaja · talk 13:39, 9 February 2019 (UTC)

@Mahagaja: That was intentional, but now I am inclined to think it is clearer to be consistent and use the circumflex in both cases. — Eru·tuon 22:16, 12 February 2019 (UTC)

Illustration of current scheme

	short vowel	long vowel	diphthong
acute	á	ǎː	aí̯
grave	à	a᷆ː	aì̯
circumflex	—	âː	ái̯

More consistent scheme

	short vowel	long vowel	diphthong
acute	á	ǎː	ǎi̯
grave	à (?)	àː (?)	ài̯ (?)
circumflex	—	âː	âi̯

That looks good. We could probably even get away with àː ài̯ for the grave accent on the long vowel and diphthong. —Mahāgaja · talk 14:56, 17 February 2019 (UTC)

Okay, done. I doubt the grave really represented low tone, but at least using just one symbol is less confusing. — Eru·tuon 21:38, 17 February 2019 (UTC)

anachronisms?

This website gives an extremely detailed progression of Greek sound changes. I'm not sure how accurate it is, but the differences with the current wiktionary pronunciation are:

• <ου> as [u:]
• <υι> as [yj]
• <ηυ> as [e:w]

Dngweh2s (talk) 20:07, 16 November 2020 (UTC)

@Dngweh2s:

w:Ancient_Greek_phonology#Vowels says

• <ου> as /uː/ (as diphthong /ou̯/, but "monophthongized in most cases")
• <υι> as /yː/ (as diphthong /yi̯/, but "monophthongized in most cases" )
• <ηυ> as /ɛːu̯/

According to w:Ancient_Greek_phonology#Monophthongization, sound shift /oː/ → /uː/ occurred "During the Classical period."

Having about one year and a half passed since this issue opened, I edited the module data so as to change classical pronunciation /oː/ → /uː/.

--Asibumi (talk) 00:04, 17 May 2022 (UTC)

Voice assimilation not shown in -ευσμ-, etc.

@Erutuon: The module seems to erroneously show the final element of ευ, αυ as voiceless before a [z] that is the allophonic realization of underlying /s/ (e.g. κέλευσμα (kéleusma) as "/ˈce.lefz.ma/", ψαῦσμα (psaûsma) as "/ˈpsafz.ma/"). Can it be corrected so these will show /vzm/?--Urszag (talk) 17:29, 26 January 2022 (UTC)

Corrected the module to be more current with modern scholarship

• The proposed metathesis of an affricate d͡z to zd in Attic, most notably held by Allen, is not supported by all scholars. Much recent scholarship that could be cited to affirm this value usually takes cue from Allen without further investigation. Threatte, although telling, is not decisive in this matter. Given that the consensus view is that some speakers of Attic, at least by the end of the 5th century BC, had the voiced alveolar sibilant for ζ, I have consequentially modified the module.
• ο, ω, and especially ε and η have had their values better defined in view of Horrocks.
• Palatalization has been represented earlier, cf. Horrocks.
• ῥ is no longer represented as voiceless in Egyptian Koine due to the uncertain status of the aspirate in this period.

Some possible additional improvements:

• σ and ζ should be represented as retracted.
• λ and ρ after kʰ, pʰ, tʰ should be represented as voiceless.
• Geminated consonants seem to have been lost in most Greek dialects by the end of the first millennium, so 10th century Byzantine (and perhaps even 4th century Koine) should probably lack gemination.
• ζ intervocalically should be geminate until gemination is lost in most dialects.

Ἀττικός (talk) 05:24, 2 February 2023 (UTC)

I've undone your changes because they should be discussed first before being made. In particular, changing /e o/ to /e̞ o̞/ is a bad idea because it is contrary to IPA principles to use "fine-tuning" diacritics like that at the phonemic level unless there's evidence that (for example) /e e̞ ɛ/ are three distinct phonemes in a language. I don't have access to Horrocks, but we should be careful to reflect what the current mainstream view is rather than aligning ourselves too closely with a single author whose ideas may or may not enjoy widespread consensus. (BTW, the word is geminate/gemination, not germinate/germination, which means something quite different.) —Mahāgaja · talk 07:42, 2 February 2023 (UTC)

(New account, because I forgot the password of my old one.)

I've undone your changes because they should be discussed first before being made.

I see.

In particular, changing /e o/ to /e̞ o̞/ is a bad idea because it is contrary to IPA principles to use "fine-tuning" diacritics like that at the phonemic level unless there's evidence that (for example) /e e̞ ɛ/ are three distinct phonemes in a language.

As a matter of fact, such a tripartite distinction is indeed present in 1st century AD Koine. The system before my revision treats epsilon strangely, with a weird, entirely ahistorical shift from e (in Attic) to ɛ (in 1st century AD Koine) to e again (in 4th century AD Koine). Even considering the principles of broad transcription, this makes no historical or linguistic sense, as the module also gives the same value ε to αι in 1st century AD Koine, but doesn't update αι to e in 4th century AD Koine, which should happen if epsilon and αι had already merged.

I don't have access to Horrocks, but we should be careful to reflect what the current mainstream view is rather than aligning ourselves too closely with a single author whose ideas may or may not enjoy widespread consensus.

Actually, Horrocks excellently represents the consensus view on the phonological development of Greek. He doesn't have any exciting or groundbreaking new hypotheses to offer, but instead summarizes much of the best scholarship from the Mycenaean era to today. I for my part do not entirely agree with some of his premises (mainly zd for ζ, which, as I stated before, rests on not entirely certain conclusions - not a few scholars disagree with it, and Allen's proposed metathesis demands further linguistic scrutiny) or conclusions, but his work remains essential reading for anyone contributing to a module like this one - and indeed for anyone interested in the history of the Greek language as a whole.

(BTW, the word is geminate/gemination, not germinate/germination, which means something quite different.)

You can thank autocorrect for that. Fixed. Ἀττικός Β' (talk) 02:25, 3 February 2023 (UTC)

The changes sound reasonable to me. I will try to look into them further later, and I hope others will comment too so the discussion that Mahāgaja requested can occur. Perhaps it was too big a change to make all at once but I'm grateful for the contribution to the project.--Urszag (talk) 14:41, 2 February 2023 (UTC)

More possible improvements:

• Σαπφώ. 5th century BC Attic /sapʰ.pʰɔ̌ː/ should be corrected to /sap.pʰɔ̌ː/. Likewise ἄπφα.
• υἱός, υἱύς, and other words with prevocalic υι. 5th century BC Attic /hyː.ós/, /hyː.ýs/, etc. should be corrected to /hyi̯.ós/, /hyi̯.ýs/, etc.; cf. Allen p. 80 ff.

Ἀττικός Β' (talk) 00:55, 4 February 2023 (UTC)

Really the only thing I'm still sticking on is the tripartite /e e̞ ɛ/ distinction. Can you go into more detail about how that's supposed to have come into existence and what it became? Also, I hope @Erutuon will chime in since he knows as much about Ancient Greek phonology as anyone here. —Mahāgaja · talk 08:36, 4 February 2023 (UTC)

the tripartite /e e̞ ɛ/ distinction. Can you go into more detail about how that's supposed to have come into existence and what it became?

Sure.

This distinction has its origin in the development of the classical Attic vowel system. So let's start from there. We have the following phonemes in classical Attic:

• ε = e
• η = ε:
• ει = e:
• αι = ai̯

The relative infrequency of ῑ and the overcrowding of the front axis following the creation of e: via monophthongization of ei̯, contraction, and compensatory lengthening gradually raised ει, first preconsonantal and word-final, and then everywhere, to i:.

• ε = e
• ει = e:, i: -> i:
• η = ε:
• αι = ai̯ (perhaps ae̯)

The position occupied by ει is now vacant. As a result, η is raised to e:. Around the same time, αι is gradually monophthongized to æ:.

• ε = e
• ει = i:
• η = e:
• αι = æ:

In the Koine, the onset of the loss of phonemic vowel length (with corresponding shift to a stress accent) results in η, needing to be distinguished from ε, being raised to e̝; with αι following along to ε. This is what we have in 1st century AD Koine:

• ε = e̞
• ει = i
• η = e̝
• αι = ε

Now, by the 4th century AD, the e̞ e̝ ε distinction has been simplified by raising η to i, with a corresponding raising of αι to e̞. This remains the situation in Greek all the way to today:

• ε = e̞
• ει = i
• η = i
• αι = e̞

Ἀττικός Β' (talk) 04:33, 5 February 2023 (UTC)

@Ἀττικός Β': Great, thanks for the explanation! So for the 4th century, the diacritics definitely aren't necessary, as there's only a two-way distinction /i/ ~ /e/. For the 1st century, couldn't the phonemes be transcribed /i/ ~ /e/ ~ /ɛ/ ~ /æ/? —Mahāgaja · talk 09:12, 5 February 2023 (UTC)

That would show the value of Greek ⟨ε⟩ as evolving diachronically from /e/ to /ε/ to /e/, which seems misleading. I know the transcriptions are already pretty crowded, but maybe it would make more sense to use phonetic brackets instead of phonemic slashes. It's doubtful that /r̥/ was a distinct phoneme from /r/, since by the normal rules they're in complementary distribution.--Urszag (talk) 10:22, 5 February 2023 (UTC)

But no more misleading than showing it evolving from /e/ to /e̞/ to /e/ when it all probability it never actually moved at all, but the movement of other phonemes around it changed its relative position. Another possibility for the four-way contrast would be /i/ ~ /ɪ/ ~ /e/ ~ /ɛ/, which would at least keep ⟨ε⟩ at /e/ the whole time. As for /r̥/, even if its distribution was predictable, the distinction was apparently felt to be important enough to be reflected in the writing. I don't know the relative timing of the loss of gemination versus the voicing of /r̥/, but if degemination happened first, then there could have been a time when -αρρα- was /ar̥a/ and -αρα- was /ara/ and the two sounds would no longer be in complementary distribution. —Mahāgaja · talk 10:43, 5 February 2023 (UTC)

So, I don't really hold the principle of avoiding diacritics on IPA unless required to mark a phonemic contrast in very high regard. (I don't think it was a good decision in the first place to define [e] as close-mid and [ε] as open-mid, [a] as front open, etc rather than just defining the five basic vowel symbols in terms of a basic 5-vowel system, but given those are the official definitions of those letters' basic values, the official endorsement of vagueness as a workaround seems kind of hackish.) Using diacritics where they're not needed might look a bit amateurish or give a bit of a false impression of precision, but I don't think it's terrible to use them in a situation like this. That said, if we do adhere to the convention of not using /e̞/ when /e/ would do, I guess I would say it makes most sense to me to transcribe the sequence that Ἀττικός Β' gave as ⟨ε ει η αι⟩ = /e eː εː ai̯/ > /e iː eː æː/ (or /εː/ for the last?) > /e i e̝ ε/ > /e i i e/.--Urszag (talk) 11:30, 5 February 2023 (UTC)

I can live with /i e̝ e ɛ/. It's preferable IMO to /i e̝ e̞ ɛ/, especially since there's no reason to transcribe the back vowels as anything other than /u o a/. —Mahāgaja · talk 12:04, 5 February 2023 (UTC)

I don't know the relative timing of the loss of gemination versus the voicing of /r̥/

Psilosis (which included r̥ -> r, e.g. the Aeolic orthography) preceded the loss of gemination. In fact, gemination still persists in a few modern Greek dialects, most notably Cypriot.

I can live with /i e̝ e ɛ/. It's preferable IMO to /i e̝ e̞ ɛ/

I agree. Ἀττικός Β' (talk) 03:53, 7 February 2023 (UTC)

@Ἀττικός Β': OK, I've restored your edits except I've replaced /e̞ o̞/ with /e o/. Please take a look and let me know if I did it right. —Mahāgaja · talk 08:51, 7 February 2023 (UTC)

It looks good. What do you think of /hyː.ós/, /hyː.ýs/, etc. -> /hyi̯.ós/, /hyi̯.ýs/? Monophthongization of prevocalic υι seems to have been universal only in the 4th century BC onwards, at least in Attic. Ἀττικός Β' (talk) 19:24, 7 February 2023 (UTC)

Seems reasonable, especially since the other diphthongs were still diphthongs at that stage. —Mahāgaja · talk 07:35, 8 February 2023 (UTC)

hi @Mahagaja please restore classic pronunciation of ζ to [zd] - 5th BC pronunciation for ζ is not [z] - as it is an indisputably double consonant according to ancient grammarians L0ngh3nry89 (talk) 20:42, 23 February 2023 (UTC)

ζ changes

hi as the scholarly consensus - and all ancient grammarians agree ζ is a double consonant. who decided against consensus that ζ now represents ungeminated [z] in 5th century attic? this is not the reconstructed sound for that period. [z] should be changed to [zd] for all entries containing ζ in 5th century attic pronunciation. i would do it myself but im not sure how. if the user who implemented this change is attikos 'b, i suggest gently that you dont try and correct a whole dictionary of entries to your personal liking because of a book or paper that you might have read. it is quite disconcerting to find such a radical change proliferated all over this site. L0ngh3nry89 (talk) 20:38, 23 February 2023 (UTC)

@L0ngh3nry89: Indeed, it was Ἀττικός Β' who changed the treatment of ζ. It's discussed in the thread above. In particular, that user says, "the consensus view is that some speakers of Attic, at least by the end of the 5th century BC, had the voiced alveolar sibilant for ζ" and "ζ intervocalically should be geminate until gemination is lost in most dialects". I'll leave it to the two of you to thrash out what is and is not scholarly consensus. —Mahāgaja · talk 21:25, 23 February 2023 (UTC)

@Mahagaja, L0ngh3nry89, Ἀττικός Β' I have reverted to 'zd' pending discussion. Since 'zd' is what was originally there and the change was disputed, it should not go in until the dispute is resolved. (Also IMO Attikos's statement about /z/ contains a lot of hedges in it that don't necessarily justify such a change.) Benwing2 (talk) 02:00, 24 February 2023 (UTC)

Is there really a scholarly consensus for the reconstruction of [zd] for Attic ζ? My impression is that a plurality or majority of informed sources assent to this hypothesis (in which case we should presumably present it) but that does not rise to the level of unanimity that I would associate with the word "consensus". I am not up to date on scholarship in this field, but the article "On the Pronunciation of Greek Zeta" by Sven-Tage Theodorsson (1978) reports some disagreement among reputable sources, and I am not sure this is such a fast-moving area of research that the controversy he records is now obsolete. There is consensus according to Theodorsson that ζ originated from earlier *[dj], *[gj], *[j], *[zd]. Obviously, it eventually came to have the fricative value [z]. As for its values in between those points, Theodorsson writes:

"The pronunciation [zd] has been assumed most generally in Ionic-Attic.⁷ Some scholars allow for two phonetic values, [zd] and [dz] (Schwyzer 1939: I, 179), while a few maintain the latter alternative (Curtius and Windisch 1879: 615; Matthews 1954/5). It has also been suggested that the pronunciation was [ʒ] (Witton 1898). Finally, a fifth solution has been proposed by Bailey (1968), who argues, on analogy with Old Church Slavic, that ⟨Ζ⟩ had the pronunciation [ʒdʒ] generally. Matthews (1954/5) has collected much evidence against the pronunciation [zd] in Ionic-Attic. It seems that there exists a further piece of counterevidence, which must be considered cogent: There is virtually no orthographic variation in archaic and classical Attic to show that ⟨Ζ⟩ corresponded to [zd]. In total, there exists only one, or at the very most four, instances that would suggest this" (pages 325-326).

Footnote 7 lists "⁷Blass 1888: 112-119; Brugmann 1913: 42; Buck 1928: 66; Brandenstein 1950/1: 45-46; Diver 1958: 18-19; Bartoněk 1961: 152; Lejeune 1972: 112-114, et al." as the sources that support [zd] in Ionic-Attic. Theodorsson himself disagrees with this general assumption, arguing that at the time of the adoption of the alphabet the value of the consonant was transiently [ʒdʒ], comparing Modern Bulgarian [ʒd] from Old Bulgarian *[dj] (page 327) and subsequently became [dz] in Ionic-Attic, then [zː], which Theodorsson says "is known to have spread in Attic by about 350 B.C." (page 330).--Urszag (talk) 03:11, 24 February 2023 (UTC)

hi urszag

there is no consensus on whether zeta represented zd or dz in the attic period, however the earliest grammarians state thusly, and the scholarly consensus is that it represents a doubled consonant. attikos changed zeta to be [z] UNGEMINATED for 5th C bc pronunciation, instead of [z:], [dz], [zd]. so whether zeta was [zd] or [dz] is immaterial. my contention is specifically ""who decided against consensus that ζ now represents ungeminated [z] in 5th century attic pronunciation. every text book on attic grammar ive read or consulted says that it is a doubled consonant. attikos' edit goes against this consensus. L0ngh3nry89 (talk) 13:20, 24 February 2023 (UTC)

as you point out those sources are quite old but not necessarily out of date. however new grammars continue to be reproduced, so either the research mentioned still holds up or there has been new research which backs up the pronunciation of zeta as [zd] since that is what is indicated in the newest grammars. the New Comparative Grammar of Greek and Latin by Andrew L. Sihler (1995, updated 2008) says: Greek ζ had the value of [zd] in the best period of Att.-Ion., but in late times became simple [z], as it still is in NG. Evidence for [zd] is as follows:

1. Statements of the ancient grammarians that the three 'diphthongs' ξ, ψ, and ζ were composed of κσ, πσ, and σδ, respectively.

2. Inscriptions and glosses that corroborate i, like Lesb. ΣΔΕΥΣ nom., ΣΔΕΥ voc. = Ζευς, Ζευ < *dy- (200).

3. Transcriptions like Ὠρομάζης for OPers. Auramazda.

4. Loss of a nasal before "ζ" precisely as before "σκ", "σπ", and "στ", as for example συν- 'together' in G σύ-ζυγος() 'yoked' just like σύ-στασις 'a putting together' and σύ-σπεύδω 'assist zealously'.

This evidence is cogent. From the purely phonetic point of view, however, the most straightforward reflex of [dy] and [gy] would have been something like [dž] or [dz], and such a pronunciation must in fact have been current in some parts; it was with this value that the letter 𐤆 was carried to Italy, where it was used to represent [ts], for example Osc. húrz [horts] 'garden' < *xortos. A development of stop + sibilant to sibilant + stop is not unprecedented: in several Slavic languages for example the presence vs. absence of such a metathesis is a distinguishing trait of dialect groups. L0ngh3nry89 (talk) 13:46, 24 February 2023 (UTC)

there is no consensus on whether zeta represented zd or dz in the attic period,

Which is why we should not blindly accept zd.

however the earliest grammarians state thusly, and the scholarly consensus is that it represents a doubled consonant.

Of course it was! Read above my proposed improvement to the template: "ζ intervocalically should be geminate until gemination is lost in most dialects."

attikos changed zeta to be [z] UNGEMINATED for 5th C bc pronunciation, instead of [z:], [dz], [zd]. so whether zeta was [zd] or [dz] is immaterial.

I actually strongly encourage z:, not an ungeminated z.

my contention is specifically ""who decided against consensus that ζ now represents ungeminated [z] in 5th century attic pronunciation. every text book on attic grammar ive read or consulted says that it is a doubled consonant. attikos' edit goes against this consensus.

Once again: "ζ intervocalically should be geminate until gemination is lost in most dialects."

as you point out those sources are quite old but not necessarily out of date. however new grammars continue to be reproduced, so either the research mentioned still holds up or there has been new research which backs up the pronunciation of zeta as [zd] since that is what is indicated in the newest grammars.

Not necessarily.

the New Comparative Grammar of Greek and Latin by Andrew L. Sihler (1995, updated 2008) says: Greek ζ had the value of [zd] in the best period of Att.-Ion., but in late times became simple [z], as it still is in NG. Evidence for [zd] is as follows:

So let's look at Sihler's "evidence" (which is basically Allen's).

1. Statements of the ancient grammarians that the three 'diphthongs' ξ, ψ, and ζ were composed of κσ, πσ, and σδ, respectively.

That fails for one important reason: not one grammarian describes ζ as equivalent to σδ. For example, take Aristotle, Metaphysics I.933a (https://www.perseus.tufts.edu/hopper/text?doc=Perseus%3Atext%3A1999.01.0051%3Abook%3D1%3Asection%3D993a): οἱ μὲν γὰρ τὸ ζα ἐκ τοῦ ς καὶ δ καὶ α φασὶν εἶναι, οἱ δέ τινες ἕτερον φθόγγον φασὶν εἶναι καὶ οὐθένα τῶν γνωρίμων. For some say that ζα is composed of σ, δ and α, while others say that it is a distinct sound and not any one of those which are familiar to us.

Notice φασίν ("they say"). The idea that there is a σ and a δ in ζ is treated as a theory, not a fact. Had ζ been equivalent to σδ, there should be no doubt, since the constituents of such a cluster would be easily identifiable. It cannot be automatically assumed that the confusion was due to a contemporaneous simplification, as it would more likely be reported that some just pronounce ζ differently.

Furthermore, Velius Longus rejects the theory that ζ is σδ and teaches that ζ was pronounced with a uniform sound; stating that ἀζηχής does not sound like ἀσδηχής, but ἀζζηχής does sound like ἀσσηχής.

Inscriptions and glosses that corroborate i, like Lesb. ΣΔΕΥΣ nom., ΣΔΕΥ voc. = Ζευς, Ζευ < *dy- (200).

Such spellings do occur in the texts of Lesbian poetry, but they are not found in early Lesbian inscriptions. Cf. Allen p. 59: "These spellings almost certainly represent a later editing", and thus bear little, if any worth as evidence of the value of ζ in classical Attic.

Transcriptions like Ὠρομάζης for OPers. Auramazda.

Ὠρομάζης has a number of variants. A very early reference to Ahura Mazda occurs in an Assyrian text, circa the 8th century BC, in the form As-sa-ra Ma-za-aš. It is entirely within the realm of possibility that the Ὠρομάζης originates from an Assyrian or descendant form. The evidence is in any case quite sketchy - the name Ὠρομάζης is only encountered twice in classical literature: in pseudo-Plato (far from a standard author) and Plutarch, who wrote so late that the use of ζ cannot be due to equivalence to zd. We encounter similar problems with other apparently Iranian transcriptions.

4. Loss of a nasal before "ζ" precisely as before "σκ", "σπ", and "στ", as for example συν- 'together' in G σύ-ζυγος() 'yoked' just like σύ-στασις 'a putting together' and σύ-σπεύδω 'assist zealously'.

This is not as cogent as it may seem at first, as such behavior is peculiar to prefixed σύν. No other type of composition appears to lose ν before ζ.

This evidence is cogent. From the purely phonetic point of view, however, the most straightforward reflex of [dy] and [gy] would have been something like [dž] or [dz], and such a pronunciation must in fact have been current in some parts; it was with this value that the letter 𐤆 was carried to Italy, where it was used to represent [ts], for example Osc. húrz [horts] 'garden' < *xortos. A development of stop + sibilant to sibilant + stop is not unprecedented: in several Slavic languages for example the presence vs. absence of such a metathesis is a distinguishing trait of dialect groups.

There we see Allen's Old Church Slavonic "medža" to "mežda". But there's a major problem with this theory: dz to zd is a metathesis of an affricate. Mežda came from an occlusive followed by a fricative, not an affricate. In fact, a dental affricate simplifying to a fricative is a far more likely to happen than a metathesis of such (which translinguistically is extremely rare). Ἀττικός Β' (talk) 04:19, 8 March 2023 (UTC)

Are you saying early Slavic dž was literally IPA [dž]? I've always assumed that the /d/ was assimilated in pronunciation to the following fricative, as it is in all languages that I've come across. —Soap— 17:02, 11 March 2023 (UTC)

I'm not sure why the distinction between a stop and fricative is a hard barrier either. I'd figured that in the scenario where Attic ζ is /sd/, there would have been an affricate /d͡z/ in earlier Greek that was reanalyzed as a consonant cluster /ds/ and then metathesized. But I don't really have a strong opinion on the pronunciation of pre-Koine ζ as all the evidence seems pretty speculative. — Eru·tuon 22:11, 11 March 2023 (UTC)

attikos your dismissal and distrust of the sources not withstanding, you continue to perpetuate such arguments such as "ζ intervocalically should be geminate until gemination is lost in most dialects." there is even less evidence for ζ being [z:], and Dionysius thrax and other ancient sources indicate ζ as a double consonant - which means here a "cluster", "geminiation" is not a kind of consonant cluster. ξ, ψ, and ζ are not "geminated" consonants. so now you are trying to argue that ζ is of a different species to the other two "double consonants" in part because you dont buy what the other researchers are saying? can you even explain how word initial [z:] is supposed to be pronounced? where else in greek do we find a geminate in "word initial" position? because ζ be [z] word intially but somehow [z:] intervocalically if it is in fact a "double consonant", because there is no evidence that shows that ζ was only "contextually" double. where is the proof? however i also need to remind you that wiktionary like wikipedia is NOT the platform for your own research and it is NOT the place to settle your grudges. the ancient grammarians knew what gemination was and yet they made a distinction between ξ, ψ, and ζ and all the other consonants. the quoute "This evidence is cogent. From the purely phonetic point of view, however, the most straightforward reflex of [dy] and [gy] would have been something like [dž] or [dz]" New Comparative Grammar of Greek and Latin by sihler, and does NOT contain a reference to "medža" or "mežda". and if we look for research done on metathesis - what do we find? that stop-sibilant metathesis is possible and does happen. the most obvious example is hebrew in the hithpael conjugation. tš or ts metathesise to št and st. i dont know why you would claim that because the metathesis of stop-sibilant is rare that this would not in fact be the case in greek, it is less likely but we do have evidence from sources which indicate the pronunciation [sd] in some parts, and to ignore the fact that ζ comes from gy and dy knowing that yod regularly triggers palatalization and knowing that ντς resolves as ς not σς, and that σσ and ττ are both reflexes of proto hellenic t' [ts] with ττ being the attic pronunciation in the 5th c, why is it so hard to believe that ζ was [sd] or [ds] even if later it was [z:/z], knowing that attic here of which we are speaking is very early and conservative.

see metathesis of stop-sibilant clusters in modern hebrew for more cross linguistic examples of stop-sibilant metathesis.

also dionysius thrax an ancient grammarian says: "ζ ξ ψ. διπλᾶ δὲ εἴρηται, ὅτι ἓν ἕκαστον αὐτῶν ἐκ δύο συμφώνων σύγκειται, τὸ μὲν ζ ἐκ τοῦ σ καὶ δ". so there is an ancient source that says what i have claimed. i dont know how you can claim such sources do not exist - more precisely you said "not one grammarian describes ζ as equivalent to σδ"

i am not sure why you are taking Velius Longus 2C AD over dionysius 2C BC - noting that Longus was a Latin grammarian, and while that doesnt mean he is uneducated when it comes to greek i am not sure why a latin source 300-400 after the sound change should be trusted over a greek source. and again on wikipedias entry on ZETA it says: "The Roman grammarian Verrius Flaccus believed in the opposite sequence, δ + σ (in Velius Longus, De orthogr. 51), and Aristotle says that it was a matter of dispute (Metaph. 993a) (though Aristotle might as well be referring to a [zː] pronunciation). It is even possible that the letter sometimes and for some speakers varied in pronunciation depending upon word position, i.e., like the letter X in English, which is (usually) pronounced [z] initially but [gz] or [ks] elsewhere (cf. Xerxes)." again why only post part of Longus' and not the full account? is it because longus own opinion is favourable to you but Verrius Flaccus is to be ignored?

/z/ was not native to Classical Latin. It appeared in Greek loanwords starting around the first century BC, and was pronounced [z] initially and doubled [zz] between vowels - however by this time greek had lost gemination and vowel length. so how do we know that Velius Longus isnt applying the rules of his language to that of 5th BC greek? we know that latin didnt have "stop-sibilants" except in a few compound words and this cluster was regularly pronounced as [s:] not [ts]. ζ in the earliest borrowings was pronounced as [s] not [z], for more evidence see Ancient Greek Dialects and Early Authors by Gary miller.

how can you say: "No other type of composition appears to lose ν before ζ." for upon searching wiktionary itself we find: Τίρυνς (Tíruns) is one of very few names of native Ancient Greek origin that contains the ΝΣ digraph. Furthermore, the nominative case has a long acute-accented penultimate vowel and an apparently short final vowel (ΤῙΡῨΝΣ). However, as this situation can normally only yield a circumflex accent in Classical Attic, the fact that an acute accent remains shows that the final vowel is scanned as long. ΝΣ and ΝΖ are not normally viable in Classical Attic and are normally respelt in inflection as just Σ or Ζ while compensatorily lengthening preceding Ᾰ Ε Ῐ Ο Ῠ to Ᾱ ΕΙ Ῑ ΟΥ Ῡ. Therefore, evidence and established grammar suggest that although Classical Attic retained the spelling as Τίρυνς instead of respelling it Τίρῡς (Tírūs), the standard Classical Attic pronunciation of Τίρυνς was to treat it as Τίρῡς. wherefore then can you say No other type of composition appears to lose ν before ζ? i dont know of any other word that has νς other than ἕλμινς and πείρινς which may be dialectal pronunciations and certainly no word with νζ, and as another entry points out that νς νζ ντς are not viable in attic greek, and that in those cases the ν or ντ are silent and lengthen the previous vowel. https://en.wiktionary.org/wiki/Wiktionary:Ancient_Greek_transliteration. thus the συν rule still as an example still holds water.

i do hope you not the author of this site - http://attic.kanlis.com/zeta.html

or are copying from that site. because that site is not trust worthy because it is the work of one determined individual (greek nationalist overtones are detected on said site).

i say that because a google search for "a metathesis of an affricate" shows only four results one which is this page, two are from attic.kanlis and one from reddit which can only be accessed via cache. L0ngh3nry89 (talk) 21:36, 14 March 2023 (UTC)

attikos your dismissal and distrust of the sources not withstanding,

I do not dismiss any sources, but I must distrust unreliable conclusions. By the way, you should take some time to learn proper capitalization and punctuation or utilize a browser extension to assist you, your posts are honestly painful to read.

you continue to perpetuate such arguments such as "ζ intervocalically should be geminate until gemination is lost in most dialects." there is even less evidence for ζ being [z:], and Dionysius thrax and other ancient sources indicate ζ as a double consonant - which means here a "cluster", "geminiation" is not a kind of consonant cluster.

Actually there's nothing preventing "double" from meaning gemination.

ξ, ψ, and ζ are not "geminated" consonants.

For the first two, you're correct.

so now you are trying to argue that ζ is of a different species to the other two "double consonants" in part because you dont buy what the other researchers are saying? can you even explain how word initial [z:] is supposed to be pronounced? where else in greek do we find a geminate in "word initial" position? because ζ be [z] word intially but somehow [z:] intervocalically if it is in fact a "double consonant", because there is no evidence that shows that ζ was only "contextually" double. where is the proof?

Reread the grammarians, especially Velius Longus, with an open mind. Seriously, it's not that hard.

however i also need to remind you that wiktionary like wikipedia is NOT the platform for your own research and it is NOT the place to settle your grudges.

Good advice. Why not take it yourself?

that stop-sibilant metathesis is possible and does happen.

Did I or anyone else ever claim it was impossible?

why is it so hard to believe that ζ was [sd] or [ds] even if later it was [z:/z], knowing that attic here of which we are speaking is very early and conservative.

Maybe reread my post again? Also, "very early" and "conservative" don't square well with any of the grammarians' descriptions, given that they all lived postclassically.

also dionysius thrax an ancient grammarian says: "ζ ξ ψ. διπλᾶ δὲ εἴρηται, ὅτι ἓν ἕκαστον αὐτῶν ἐκ δύο συμφώνων σύγκειται, τὸ μὲν ζ ἐκ τοῦ σ καὶ δ". so there is an ancient source that says what i have claimed. i dont know how you can claim such sources do not exist - more precisely you said "not one grammarian describes ζ as equivalent to σδ"

Dionysius Thrax nowhere states equivalency. One can describe something as being composed of various elements without literally being those elements itself.

/z/ was not native to Classical Latin. It appeared in Greek loanwords starting around the first century BC, and was pronounced [z] initially and doubled [zz] between vowels - however by this time greek had lost gemination and vowel length.

Excuse me? Gemination still exists today in some Greek dialects! And while absent from the majority, vowel length was most likely still a characteristic of "learned" speech.

so how do we know that Velius Longus isnt applying the rules of his language to that of 5th BC greek?

And what of the other grammarians, given that none of them wrote in the "best" age of classical Attic. Or what if Velius Longus is simply describing... you know, just how ancient Greek works?

i do hope you not the author of this site - http://attic.kanlis.com/zeta.html or are copying from that site. because that site is not trust worthy because it is the work of one determined individual (greek nationalist overtones are detected on said site).

Nice ad hominem. That's actually a good page. I don't agree with all of his conclusions though. Ἀττικός Β' (talk) 01:57, 18 March 2023 (UTC)

-

let me start with a reminder - this site is not the place for you to settle personal grievances or establish YOUR NEW CONSENSUS against scholars that you don't like or whose conclusions you don't find "reliable".

-

attic. canals is not a good page because it is an unreliable source run by one man with a prominent bias. if you consider that site a good page - knowing that you've clipped that response out of my larger argument which is that your arguments are quite specific and can be found only in a couple of spots on the internet, which does not lend credence to your position - then I cannot help you and I do not wish to continue this conversation for much longer. I did not engage in an ad hominem attack, at best you could claim it resembled the genetic fallacy, but given that you are making similar arguments, that your arguments are not widespread on the internet, and given that that source is unreliable, I have to ask the question whether you are sourcing material from that website. that is precisely not the definition of an ad hominem. I did not attack your character - which is clear from the larger argument. I have seen numerous instances of Greeks, online, claiming that reconstructed pronunciation is wrong, is imperialist, is a-historic, anti greek etc. this has a lot to do with schooling in Greece, and how Ancient Greek is perceived, received, understood and taught in schools today. Therefore it is very much possible to see the signs of this nationalistic view of ancient greek in websites that attempt to correct the online community's understanding of ancient greek pronunciation - most evident of all in the derision they have for non-Greek sources (which disagree with them), and even Greek sources. therefore just as one would hesitate to use a pro-trump site as a source for Donald Trump's Wikipedia entry, so too must one hesitate to use information from a biased and fringe source that does not engage with or is engaged with by other scholars. and doing it all on a 1995 style HTML site, as opposed to a journal article or encyclopedia entry - indeed these sources form the basis for this template and others, NOT blog posts - which would lend more credence to his hypotheses; since he would have to face peer review.

-

in no particular order: quoting you: ""Dionysius Thrax nowhere states equivalency. One can describe something as being composed of various elements without literally being those elements itself."" you are playing word games and I don't know why, you dismiss evidence you don't like - you repeatedly ignore my larger argument and pick phrases at random seemingly to nitpick on, ... and when confronted with the inconsistencies in your own evidence you play a game of whataboutism. LET ME EXPLAIN what this entails:

-

quote: ""And what of the other grammarians, given that none of them wrote in the "best" age of classical Attic. Or what if Velius Longus is simply describing... you know, just how ancient Greek works?"" instead of ENGAGING with the evidence.

you of course again pick and choose clips out of my larger argument and find one thing you don't like and dismiss the rest without dealing with it.

Velius Longus presented contrary information to what you initially provided - which you clearly omitted, and was later still than the Grammarians whom you claim were somehow nigh unto equally distant or removed or unfamiliar with classic attic pronunciation. with the result that you trust a Latin speaker 600+ years after the attic period; towards the end of the koine period, over a Greek Grammarian writing in the Hellenistic period??? this is a false equivalency. 200BC is better than 200AD, even though neither is 500BC. both are not of the same kind.

""Dionysius Thrax nowhere states equivalency."" you are playing word games and I don't know why? σ δ represent the sounds [s] [d], with σ becoming voiced before βγδμ, thus Dionysius is saying quite clearly that ζ = σδ because, σδ = [zd], so what is this demand for a claim of "equivalency"? knowing that ξ, ψ are described in the same way "τὸ δὲ ξ ἐκ τοῦ κ καὶ σ, τὸ δὲ ψ ἐκ τοῦ π καὶ σ." can also be written as separate consonants ""[ks] could be ΚΣ, ΧΣ, Χ, or Ξ [ps] could be ΠΣ, ΦΣ, or Ψ""

-

and while we are at it you again deliberately chopped up my words to make me say something I didn't say. I did not claim or say that you said metathesis of an affricate is impossible, you said: Did I or anyone else ever claim it was impossible?, but I never said you said it was impossible my quote in full says ... New Comparative Grammar of Greek and Latin by Sihler, and does NOT contain a reference to "medža" or "mežda". and if we look for research done on metathesis - what do we find? that stop-sibilant metathesis is possible and does happen. the most obvious example is Hebrew in the hithpael conjugation. tš or ts metathesize to št and st. i don't know why you would claim that because the metathesis of stop-sibilant is rare that this would not in fact be the case in Greek, it is less likely but we do have evidence from sources which indicate the pronunciation [sd] in some parts, and to ignore the fact that ζ comes from gy and dy knowing that yod regularly triggers palatalization and knowing that ντς resolves as ς not σς, and that σσ and ττ are both reflexes of Proto Hellenic t' [ts] with ττ being the attic pronunciation in the 5th c, why is it so hard to believe that ζ was [sd] or [ds] even if later it was [z:/z], knowing that attic here of which we are speaking is very early and conservative see metathesis of stop-sibilant clusters in modern Hebrew for more cross linguistic examples of stop-sibilant metathesis. the evidence presented here comes from ... METATHESIS OF STOP-SIBILANT CLUSTERS IN MODERN HEBREW: A PERCEPTUAL INVESTIGATION - Kyle S. Jones - university of Arizona. your conclusion that because said linguistic phenomenon is rare THEREFORE it is unlikely to happen in Greek is only the beginning of an argument, one which you have not supported with relevant data, and we do know that stop sibilant clusters are not viable in ancient Greek - and yet despite the large body of evidence i presented to you from this very site - you ignore it and instead play a game of semantics with me, and a game of spot the misquote. did you or were you unwilling to acknowledge that i framed my response around the possibility of stop sibilant metathesis as it relates to ancient Greek - which you have argued is likely to not be the case because of its rarity - even though the paper i alluded to and which i have now referenced in full mentions several examples from languages around the Mediterranean and Europe generally. OVS OSV languages are very very rare - but those languages which do have them are mostly clustered around the amazon jungle. thus while OVS and OSV are rare cross-linguistically, they are certainly not that rare or unusual in the amazon - and may be an areal feature. therefore you need to come up with a better argument than simple rarity to explain why ancient Greek did not have stop sibilant metathesis - when such examples can and do occur in languages around the Mediterranean.

-

again why cut out part of my argument and not quote it in full? I said "why only post part of Longus' and not the full account? is it because Longus' own opinion is favourable to you but Verrius Flaccus is to be ignored? /z/ ζ was not native to Classical Latin. It appeared in Greek loanwords starting around the first century BC, and was pronounced [z] initially and doubled [zz] between vowels - however, by this time greek had lost gemination and vowel length. so how do we know that Velius Longus isn't applying the rules of his language to that of 5th BC Greek? but then you say "Excuse me? Gemination still exists today in some Greek dialects! And while absent from the majority, vowel length was most likely still a characteristic of "learned" speech." so now you are extrapolating from the exceptions to the general? and thence to 5th-century attic? on what basis??? ""Horrocks agrees with Gignac on finding evidence that geminate consonants tended to simplify beginning from the 3rd century BC, as seen in their arbitrary use in less literate writing. However, degemination was not carried out universally, as seen where the South Italian, south-eastern and some Asia Minor dialects preserve double consonants.""'from Wikipedia's article on koine Greek phonology.

""... neutralization of phonemic gemination of consonants seems to have already begun in Egyptian Greek during the Hellenistic period."" and ""Egypt was likely one of the first regional varieties of Koine Greek to neutralize phonemic gemination43 —it probably did so earlier than Judea-Palestine—this same phenomenon might have occurred there already at an earlier stage."" and ""C.8. Geminated consonants were generally simplified. While this likely began in some regions during the Hellenistic period, other regional varieties of Koine Greek apparently retained geminated consonants into the Roman period. This change is indicated in part by spelling interchanges like αλα (for ἄλλα) and γραματα (for γράμματα)."" and again ""C.1. The original combination ζ = /zd/ shifted to /zz/ by the early Hellenistic period. Later, probably around the same time that gemination was generally simplified, /zz/ was further simplified to /z/. This likely took place by the Roman period."" The LXX and Historical Greek Phonology: Orthography, Phonology, and Transcriptions. Thus, my comment refers to the general trend in koine Greek - and considering that while the Athenian dialect remained prestigious, did Velius Longus study Greek in Athens? Did he learn it from Greek-speaking Romans [in italy]? Velius Longus in writing about Ancient Greek pronunciation most likely based his work on the traditional pronunciation preserved in the dialects of Italy which preserved gemination - in part because of the influence of Latin - which did have phonetic gemination, rather than the dialects of mainland Greece. therefore in sum - what evidence do we have for 5th century BC pronunciation of ζ as [z:], when the data you provided is limited to an even later period and a group of people which had contact with Latin for some time before Longus wrote, whilst ignoring the evidence of largely monolingual Greek speakers who preceded Latin cultural expansion in Greece, who by and large had lost gemination.

-

I said: ξ, ψ, and ζ are not "geminated" consonants. on the basis of the writings by Grammarians like Dionysiοs Thrax. he calls them double consonants

For the first two, you're correct. again... read Thrax, let me quote in English this time: ""Again, of the consonants, three are double, ζ, ξ, ψ. They are called double because each one of them is composed of two consonants"" nowhere does he refer to κκ λλ μμ νν ττ σσ in the same way, in fact, he does not even discuss the notion of gemination, so now you want to take 2/3 of his explanation as being correct, whilst claiming the other is not as he describes but the word he uses to describe zeta somehow also means gemination even though he does not refer to the phenomenon otherwise??? and why should I read a very late Latin grammarian with an open mind but ignore the plain witness of Thrax which is 4 centuries earlier? now you might opine that I am saying that you said to ignore Thrax. no of course you didn't. but your reply was that I read Velius Longus with an open mind - despite your lack of other sources except for Aristotle, whom you are very creative in your exegesis thereof. if as we should - read it in context - we see an entirely different claim is being made. ""On the other hand, assuming that this knowledge should turn out to be innate,1 it is astonishing that we should possess unawares the most important of the sciences. Further, how is one to know of what elements things consist? how is it to be established? Even this presents a difficulty, because the facts might be disputed, as happens in the case of certain syllables—for some say that ZA is composed of S, D and A, while others say that it is a distinct sound and not any one of those which are familiar to us. notice how the issue is framed thus "how" does one know?, and how does one establish facts? clearly the issue is not about the pronunciation of zeta. since he does not pick it up again. and we know there were dialectical variations of Greek, if this is the case then Aristotle is not even talking about 5th-century Attic Greek specifically, - but He being part of the post-attic elite is engaged in or knows of the contemporaneous issue, namely - what is the correct pronunciation of Greek? As part of the larger context being, do we speak like Athenians? as an example. furthermore, his second option is that it is some kind of sound that is unknown to us, who is this "us"? speakers near to Aristotle presumably (Aristotle and friends?). furthermore, this does not prove that ζ being composed of σ καὶ δ is but a "theory" is to misuse the word theory. a theory is an explanation of one or more facts. the word you should be using here is hypothesis. but Aristotle is not describing a hypothesis, but a dispute. what do you even mean by fact? we know that greek had many dialects, and since the template for Greek includes 5th century attic, i do not understand why you are looking for a "fact" that would not be disputed. clearly, zeta was pronounced differently in different places and at different times. the facts you are searching for might not be found. people cannot even agree about which is the correct pronunciation of scone. Greek speakers pronounced the name of Zeus differently: Δάν (Lesbian) Δεύς (Laconic) Ζάν (Doric) Ζάς (Doric) Ζήν (Poetic) Θιός (Boeotian) Σιός (Boeotian) Τάν (Cretan), and we know they all come from PIE dyḗws via Proto Hellenic *dzeus. therefore we know that the pronunciation of zeta or more specifically Proto Hellenic dz, dy, gy, etc was highly variable in the Greek dialects HOWEVER this template is not for all Greek dialects but includes only 5 main eras, of which 1 is Attic. and in Attic Greek, it was [sd]. I don't know why you would expect there to be unanimity then in how ancient writers understood the sound of zeta, especially following the expansion of Koine and the consequent dialect mixing.

-

and again I am not sure why you are unable to read charitably. but let me quote in full and then remind you of the context. "": why is it so hard to believe that ζ was [sd] or [ds] even if later it was [z:/z], knowing that attic here of which we are speaking is very early and conservative."" : and your reply "" Maybe reread my post again? Also, "very early" and "conservative" don't square well with any of the grammarians' descriptions, given that they all lived postclassically."" notice the attic which I refer to is the aforementioned 5th century BC attic greek. NOT early and conservative "grammarians". thus what you are trying to do is cast doubt upon the pronunciation of zeta as [sd] in the 5th century BC, given that the grammarians who describe it are much later, but since we cannot avoid the issue of reconstruction, it does not follow that we cannot reconstruct the pronunciation based on later works by working backwards and following the sound changes. and again why: "is it so hard to believe that ζ was [sd] or [ds] in the 5th century BC! in ATTICA even if later it was [z:/z] in Ionic or the koine period before gemination was lost in most Greek dialects. you insist on imposing the pronunciation of the 3rd 2nd century BC KOINE onto 5th century BC attic which as I mentioned already has other spelling/phonological differences between it and Ionic and other dialects, and even Epic Greek is different from all of those. will you also insist that ττ in Attic was pronounced the same as σσ in ionic and later koine greek? because later dialects did not have ττ???

and again you ignore most of my argument and nitpick on one thing and then ignore the rest. as I said with regards to latin Z as it comes from Greek: ""ζ in the earliest borrowings was pronounced as [s] not [z], for more evidence see Ancient Greek Dialects and Early Authors by Gary miller."", meaning when it was borrowed around the 1st century BC ζ was pronounced as roman S not [z]. and only later did Romans adopt the sound of [z] since the phoneme /z/ was not part of classical latin, unlike archaic latin. therefore if /z/ is a foreign sound to Latin, and affricates, and voiced stop sibilants are not part of the language either, then it is at least possible that Velius Longus' conclusions were incorrect since these sounds previously described were part of educated speech, not normal speech, and if Japanese speakers who know that English words have both L and R, but cannot perceive a difference between them, it is equally possible that as a Roman Velius could not distinguish ZZ from Z, from DZ, from ZD or TS. therefore despite your assertions, i see no reason to take Velius over other Grammarians. and this is not as you say: ""Reread the grammarians, especially Velius Longus, with an open mind. Seriously, it's not that hard."" not being open-minded, but one of being able to weigh up the evidence, and make conclusions about, considering the context, and the history of the text - and of course, that involves considering the author's own context.

-

let me be clear I am not here to settle a grudge with you, BUT if you are here to conform Wiktionary's ancient Greek pronunciation template to YOUR PREFERRED model, to your preferred idea, to your preferred conclusions, then I will not let that slide, this is the problem on this platform where one individual such as yourself takes it upon himself to rewrite whole articles and templates just because he or she thinks he knows better than the body of scholars, Wikipedia's goal is not to represent the best argument but the general consensus of a topic or issue as is reported by reliable sources.

-

I am presenting to you the research, that is known, that is available - it is not my own research, but it seems to be the evidence you won't look at or carefully consider. It is you, you are the one questioning the sources and attempting to change the template to suit your preferences - NOT ME. even if you were a professor at Athens university it would still be highly inappropriate for you to go around modifying the entire template to suit your conclusions. if you think you have a better theory - go publish a journal article and one of us will incorporate that information into this discussion once it has been peer-reviewed. but it isn't your job to fix the template. i am not changing the template with my own research, my own work - articles/entries/theses etc - that is what you are doing. as editors we are supposed to refrain from this practice. this is the issue i have with your edits.

-

L0ngh3nry89 (talk) 22:51, 18 March 2023 (UTC)

https://antigonejournal.com/2022/02/latin-spelling/

Romans regularly spelt ζ as S in their inscriptions for some time before Z became part of the alphabet again.

there are two more entries for words suffixed with ἔν but both are followed not by σ alone but a stop ἔνσπονδος, ἔνστασις, so again νσ νζ is unstable and not viable in attic greek. there are of course about 7-8 polytonic entries on lsj and logeion amongst others which have ενζ however the majority are either misspellings, spellings for the sake of metre, rare forms, or as is the case for all the monotonic entries post classical formations. L0ngh3nry89 (talk) 21:42, 14 March 2023 (UTC)

Add "annotation" parameter for multiple pronunciation

I wonder if someone could create the "ann" parameter for different vowel lengths, as in Template:la-IPA? JoeyChen (talk) 05:47, 30 June 2023 (UTC)

I don't know how to do it myself, but I definitely support this suggestion. —Mahāgaja · talk 06:49, 30 June 2023 (UTC)

Fix display of hyphenation

Because this template encloses list items in a <div> tag, list items placed after it don't display correctly, as in ἕως (héōs), where we have

{{grc-IPA}}
*{{hyphenation|grc|ἕ|ως}}

and the bullet next to the hyphenation displays to the left of the content area because the wikitext parser puts the hyphenation in a <li> tag without an enclosing <ul> tag. That may be a parser bug, but even with it fixed, we would want the hyphenation and the pronunciation to be in the same list in the HTML (in <li> tags that are inside the same <ul> tag). Achieving that requires changing the output of this module. Possibly we could get rid of the <div> around the list items and output literal <ul> and <li> tags with collapsibility classes; in that case we would have to add a hyphenation parameter to this template because a * on a line below the template would generate a nested list, which would be just as bad as the current situation. — Eru·tuon 22:35, 30 January 2024 (UTC)

Missing Attic diphthongs

In the Attic dialect, ει and υι were diphthongs when they occurred before vowels, thus pronounced pronounced */ei̯/ and */yi̯/. This is according to Wikipedia (Ancient Greek phonology) as well as Polymathy (Greek Pronunciation: ει (epsilon iota), the full history) Yet, curretly, Ἀλεξάνδρεια is transcribed as /a.lek.sán.dreː.a/ Torr3 (talk) 18:58, 8 February 2024 (UTC)

Wiktionary uses phonemic spellings. I don't think there's any evidence of a phonemic contrast between /eː/ and /ei̯/ in 5th century Attic, so it would seem that even if a phonetic diphthong [ei̯] occurred in this context, it could be interpreted as an allophone of /eː/. W. Sidney Allen (Vox Graeca, 1968, 79-80) interprets variation between forms like ιερέα, δωρεά and ιέρεια, δωρειά as suggestive of pronunciations with [ej] vs. [ejj] respectively, but I don't really see why alternation between short [e] and long [eː] isn't an equally tenable interpretation. Are there other scholarly sources that discuss this?--Urszag (talk) 21:22, 8 February 2024 (UTC)

However, in the example ἀκούω given in the documentation, the 5th century Attic pronunciation is /a.kǔː.ɔː/ instead of /a.kǒː.ɔː/. I suppose there are similar issues with <ει> being represented by iː instead of eː. Besides that, the pseudo-gemination of intervocalic semivowels (as in <γαῖα> /ɡâi̯.ja/) isn't represented. Golmore (talk) 07:02, 6 September 2024 (UTC)