Inflections of Ζεύς

Question

Διώνη is the name of a Titaness, a nymph, and Phoenician goddess. And according to the Wikipedia article on said Titaness, it's derived from the feminine form of the genitive of Ζεύς. And according to Wiktionary, all of the inflected forms of Ζεύς change Ζ ("z") to Δ ("d").

Two questions:

1. Why and how does the initial letter in the word transform into a completely different one in the process of inflection? It's a bizarre anomaly, as in my experience I've never seen any Greek word (modern on ancient) that changes anything other than the last few letters when it's inflected.
2. Is Διώνη truly the only feminine form of the name Ζεύς? Or is there an alternative one? Because unless I'm missing something, one would be hard-pressed to realize the connection between the two names if they aren't already familiar with the aforementioned quirk in the declension of Ζεύς.

Answer 1

It's more accurate to ask why the Δ changes to Ζ than vice versa. Historically Ζεύς comes from a pre-Greek form *Dyeus, and just like in English we sometimes hear a "dy" combination merge into a "j" sound (i.e., [dj]) -> [dʒ]), the same thing happened in Greek, so that the initial consonant was originally pronounced something like [dʒ] (with subsequent development on to [dz], and then, depending on dialect to [zd] or [z]).

From the original "Dyeus" (or "Dieus" or "Dyews"), it's easily to see where the genitive "Diwos" (later "Dios"/Διός) comes from, and then Διώνη (from Diōnā < *Diwōnā) makes sense. (Note that the pan-Greek form of this name would be Διώνᾱ, the final change to Διώνη represents a general replacement of ᾱ by η in Attic/Ionic.)

In most cases, this type of alternation would get levelled out in the passage of time, but in the case of Ζεύς, this didn't happen. (Or actually, in some dialects, it did happen, with the development of forms like Ζηνός, &c.,extending the Ζ throughout the inflexion.)

This is getting rather far from the original question, but the genitive form Ζηνός is an interesting example of analogic extensions. The expected form, based on purely phonological principles, of the accusative of original form (something like *Dyēm) would be Ζήν, but by analogy with other 3rd-declension nouns it got an additional accusative suffix to yield Ζῆνα (note that the normal Accusative of Ζεύς, Δία, is also formed by analogy). Then, from Ζῆνα, other inflexional forms were created based on the newly perceived stem Ζην‐.

Answer 2

Varro's answer is excellent (and entirely correct), but it's worth noting that we see the same sort of alternation in the same word in Latin: Juppiter and diēs both come from the same root! Just like in Greek, the initial *d mutated before *j/y.

Borrowing some explanation from this answer:

The Proto-Indo-European form behind Zeus is reconstructed as *dyēw-s, with the oblique stem *diw- used for all forms except nominative, accusative, and vocative. This sort of alternation between *yē and *i is common in PIE: look up "ablaut" for more information.

In Proto-Greek, the last common ancestor of the Greek dialects, this turned into something like *dzéus: dentals before /j/ palatalized into affricates. It's not entirely clear what the affricate sounded like, but Mycenaean (the oldest attested Greek dialect) uses the sign 𐀽 for earlier *dye and the sign 𐀆 for earlier *de, so there was definitely a difference there. The oblique stem was still *diw-.

Finally, in Ancient Greek, the *dz was written with the letter zeta; different dialects pronounced it differently, but the Romans just borrowed the letter zeta to represent it, and modern transcriptions do the same: "Z". This is how we get the forms Zeus, Zēn, Zeu alongside the oblique forms Diós, Día (previously Diwós, Díwa before most dialects lost /w/).

---

Similarly, Latin Juppiter comes from a vocative form combined with the word for "father"; compare the Christian formula "Heavenly Father". The PIE vocative form is reconstructed as *dyew, which became Proto-Italic *djou (aka *dyow, different ways of writing the same thing).

Unlike in Ancient Greek, the *dj didn't survive as a special palatalized consonant: it simplified into /j/ by Latin times. Then *Jou-patēr became Jū-piter by sound change, and then sometimes became Juppiter due to something called the "littera rule"; the two forms were in free variation.

The oblique forms on *diw- also survived into Latin, creating the noun *djow- > Jov-is (where Classical Latin v represents /w/). This had normal third-declension (consonant-stem) forms: Jov-is, Jov-ī, Jov-em, and so on. Through suppletion, the special form Juppiter with "father" glommed on took over the nominative and vocative, while the regular Jov- was used for all other cases.

The accusative form of the root, *dyēm (cf AGrk Zēn), also survived in Latin, in the form *diēm > diem. Uniquely, the dy here didn't become *dj > j; some scholars suggest that it evolved in a different Italic language/dialect, which didn't have the palatalization rules, then got borrowed back into Latin. When it did come back to Latin, though, it was with a new meaning of "day" (instead of "sky/sky god"), and it was extrapolated into a whole paradigm on the stem di-, and forced into the fifth declension where it didn't really fit (di-ēs, di-ēī, etc).