Talk:Ζεύς

The accusative of Ζεῦς should be expanded to include dialectal or poetic forms Ζῆν and Ζῆνα. 99.231.240.64 07:31, 7 February 2016 (UTC)

dz or sd?

Etymology states: "From Proto-Hellenic *dzeus". Pronunciation gives: "/sdeú̯s/".
How come? Is etymology or pronunciation wrong? Was there metathesis? Is it uncertain if Ζ stands for sd or dz, so that the pronunciation would correctly be "either /sdeú̯s/ or /dzeú̯s/"? -80.133.100.58 14:44, 10 April 2018 (UTC)

Yes, the reconstructed Attic transcription reflects the theory that there was metathesis in Attic (/dz/ → /sd/, pronounced [zd]). In other dialects, the pronunciation /dz/ may have remained, and it is a more likely as the ancestor of the Koine, Byzantine, and Modern Greek pronunciation [z(z)]. I'm not sure, some scholars might disagree with this theory about Attic. A sort-of explanation for this metathesis is that Attic had a phonotactic constraint that prohibited any alveolar stop–fricative clusters; they changed either to fricatives (*métsos → μέσος) or to fricative–stop clusters (*dzeus → Ζεύς /sdeú̯s/). — Eru·tuon 19:09, 10 April 2018 (UTC)

@Erutuon: One thing I've wondered is why do we represent this sound as /sd/ rather than /zd/? --WikiTiki89 19:28, 10 April 2018 (UTC)

@Wikitiki89: Because Attic didn't have a /z/ phoneme; [z] was an allophone of /s/ before voiced consonants. —Mahāgaja (formerly Angr) · talk 19:32, 10 April 2018 (UTC)

Might it make more sense to treat /zd/ as a single phoneme? --WikiTiki89 19:43, 10 April 2018 (UTC)

@Wikitiki89: That would be a pretty weird phoneme. I'm not aware of any case in which a fricative–stop cluster is considered a phoneme. — Eru·tuon 19:46, 10 April 2018 (UTC)

Considering it a single phoneme would make it hard to justify not treating the rest of the fricative–stop clusters as phonemes, of which there are many: σπ, σβ, σφ, στ, σθ, σκ, σγ, σχ (sp, sb, sph, st, sth, sk, sg, skh). Aside from place of articulation, /sɡ, sb/ in particular are pretty similar to /sd/. — Eru·tuon 20:03, 10 April 2018 (UTC)

If ζ (z) and σδ (sd) were completely equivalent, you would have expected that σδ (sd) would develop into /z/. Furthermore, I just noticed that we do represent σ (s) as /z/ before voiced stops in our module. --WikiTiki89 21:00, 10 April 2018 (UTC)

@Wikitiki89: Yeah, there are a number of other non-phonemic things transcribed by the module. It needs cleaning up. See above the origin of later Greek /z/: if I recall right one hypothesis is that the /z(z)/ pronunciation originated from non-Attic dialects, in which ζ (z) was still pronounced /dz/ and was distinguished from σδ (sd) /sd/. — Eru·tuon 23:57, 10 April 2018 (UTC)

Just out of curiosity, how exactly do we know that /s/ was voiced to [z] before voiced consonants in Attic? --WikiTiki89 00:09, 11 April 2018 (UTC)

The evidence that I remember is Koine-era spellings with ζμ (zm) instead of σμ (sm), indicating that in Koine the two sounded the same, as well as the assumption that processes of assimilation didn't change between Attic and Koine, in the same way that they didn't between Koine and Modern Greek. In other cases assimilation was regressive (λέγω, ἐλέχθην (légō, elékhthēn), not *ἐλέγδην (*elégdēn)), so I find it convincing. — Eru·tuon 03:02, 11 April 2018 (UTC)

Assimilation rules can definitely change over time and between dialects, and there is no reason to assume that because one type of consonant cluster assimilates one way that another type will assimilate a different way or not at all. For example, in Russian, /s/ voice-assimilates to following /b/, but not to following /v/ (meanwhile, Ukrainian has no voicing assimilation at all). --WikiTiki89 14:35, 11 April 2018 (UTC)

Well, unlike Russian /v/, Ancient Greek /m/ did trigger assimilation of stops: /k, kʰ/ + /m/ → /ɡm/, /p, pʰ, b/ + /m/ → /mm/. There was assimilation in phonation or nasality. But I suppose this assimilation could have affected only stops as easily as it could all obstruents. — Eru·tuon 22:37, 11 April 2018 (UTC)

Did /t, tʰ/ + /m/ become /dm/? --WikiTiki89 13:59, 12 April 2018 (UTC)

Not in τμῆσις (tmêsis) and ῥυθμός (rhuthmós); at least not early enough to affect the spelling. —Mahāgaja (formerly Angr) · talk 14:09, 12 April 2018 (UTC)

Then that's a really big hole in the argument. Since /s/ is coronal just like /t/ and /tʰ/, perhaps this voicing assimilation didn't affect coronal consonants in Attic. Or even if you can't generalize to coronal consonants, it's proof that not all consonants were treated equally, so there is no reason to assume voicing assimilation of /s/ without evidence. --WikiTiki89 14:22, 12 April 2018 (UTC)

From Attic inscriptions. Spellings like τοῖσζ’ (490 BC) for τοῖσδε are very early; ζ or σζ for σ before μ appear starting from the 4th century BC: these may reflect the [z] pronunciation, but a simple epenthesis is also possible. Guldrelokk (talk) 22:30, 15 April 2018 (UTC)

And of course ζ for σδ occurs from the earliest time. Guldrelokk (talk) 22:37, 15 April 2018 (UTC)

Oops, I have to correct what I said. Actually, the assimilation /k, kʰ/ + /m/ → /ɡm/ is a morphophonological rule that applies when suffixes beginning in /m/ are added, for example -μενος, -μα, -μαι. /km, kʰm/ do exist: ἀκμή (akmḗ), δραχμή (drakhmḗ). (But clusters of a labial stop and /m/ don't, based on Perseus Project searches: 1, 2, 3.) — Eru·tuon 21:35, 13 April 2018 (UTC)

Before those m-initial suffixes, /t tʰ d/ all become /s/, don't they? —Mahāgaja (formerly Angr) · talk 07:21, 14 April 2018 (UTC)

Yes. I didn't mention it because it seems like something other than assimilation but I don't know what. The *tt, *dt → *tst → ss change in Latin is somewhat reminiscent, but it's not in the same environment (before *t rather than before *m). — Eru·tuon 01:29, 16 April 2018 (UTC)

Pronunciation of /eu/

Is it really /zefs/ in modern Greek? The IPA is auto-generated so there'd be no way to know for sure if it was a mistake. Are there other words in modern Greek that end in the cluster /fs/, aside from reborrowings of ancient Greek names? Soap (talk) 20:39, 13 November 2018 (UTC)

@Soap:: Yes, it's zefs. See Greek Wiktionary, where the pronunciation has been entered manually. I don't speak Modern Greek so I don't know how common the cluster is at the end of a word. Perhaps @sarri.greek can answer. — Eru·tuon 23:52, 13 November 2018 (UTC)

Yes, @Soap, as Eru writes, we pronounce [efs] Ζευς (ipa rules here). All the inherited -εύς masculines have never stopped being used. Pronounced since late antiquity and mediaeval times /efs/ (plural -εῖς [is]). The greeks pronounce all ancient greek texts in Reuchlinian fashion: with modern pronunciation. Extra modern ending -έας [eas] developed as well from the ancient accusative ‑έα. 90 modern words found at “*έας”, in Λεξικό της κοινής νεοελληνικής [Dictionary of Standard Modern Greek], Triantafyllidis Foundation, 1998

We use the old style in formal conservative speech, and the modern ending in both formal and informal speech.

Examples in old style (and modern style) as used today: Προμηθεύς (Προμηθέας), Ἀχιλλεύς-Αχιλλεύς (Ἀχιλλέας-Αχιλλέας), βασιλεύς (βασιλέας), γραμματεύς (γραμματέας), ιερεύς (ιερέας), etc, and the irregular Ζεύς-Ζευς (Δίας). sarri.greek (talk) 07:31, 14 November 2018 (UTC)

Okay thank you both. So, if I understand correctly, the affix -eus is generally a reborrowing, having been lexically replaced by -eas in Modern Greek, but everywhere that it does occur, it is always pronounced /efs/. That answers my question, thank you. Soap (talk) 17:55, 14 November 2018 (UTC)