Laudator Temporis Acti

Homer, Iliad 3.276-277 (tr. Peter Green):

Zeus, Father, ruling from Ida, most glorious, greatest, and you, Sun, the all-seeing, all-hearing...

Ζεῦ πάτερ Ἴδηθεν μεδέων κύδιστε μέγιστε,
Ἠέλιός θ᾽, ὃς πάντ᾽ ἐφορᾷς καὶ πάντ᾽ ἐπακούεις ...

Jacob Wackernagel (1853-1938), Lectures on Syntax with Special Reference to Greek, Latin, and Germanic, tr. David Langslow (Oxford: Oxford University Press, 2009), p. 14 (from Lecture I, 2):

In this course on syntax, we shall have constantly to deal with inherited material, which is to be found in even the smallest details and in oddities concerning which one would hardly think in terms of inheritance. For example, at Iliad 3.276–7, in a prayer to Zeus and the Sun-god, we read: Ζεῦ πάτερ ... Ἠέλιός τε ('Zeus father (voc.) . . . and Helios (the Sun) (nom.)'), i.e. in the invocation one god is named in the vocative, the other in the nominative. It would be superficial simply to refer this to the requirements of the metre, since the poet would have had other means at his disposal for turning out a correct hexameter. From the point of view of Greek this is an oddity, particularly as vocatives elsewhere occur in coordination. This puzzle was solved by an outstanding philologist, Theodor BENFEY (1809–81). He showed (1872: 30–4) that in the Rigveda, the oldest written remains of an Indo-European language, when two forms of address are joined together with the little word ca 'and' (corresponding to Greek τε 'and'), the second is in the nominative rather than the vocative. So Homer's use of the nominative instead of the vocative is conditioned by the little word τε. This tiny detail reveals the power of linguistic habit and the influence of inheritance.

The reference is to Benfey, "Über die Entstehung des indogermanischen Vokativs," Abhandlungen der hist.- philologischen Classe der Königlichen Gesellschaft der Wissenschaften zu Göttingen 17 (1872) 1–92.

Wackernagel's rejection of a metrical explanation looks like it could be a criticism of Basil L. Gildersleeve, who said in his review of Delbrück's Die Grundlagen der griechischen Syntax in American Journal of Philology 2.5 (1881) 83-100 (at 88):

Benfey's parallel between the copulation of voc. and nom. by τε in Γ 277 Ζεῦ πάτερ ... Ἠέλιός θ' ὃς κτέ, and a similar union by means of ca in Sanskrit does not seem to me to rest on a sufficiently wide basis of induction. Ἠέλιε cannot be got into the combination on account of the verse, and while it is easy enough to abuse the old device of metri causa, we cannot at all events deny the influence when there is a balance.

M.L. West (1937-2015), Indo-European Poetry and Myth (Oxford: Oxford University Press, 2007), pp. 306-307:

Vedic and early Greek agree further in a syntactic peculiarity that sometimes appears when two deities are invoked together. One is addressed in the vocative, the other appended in the nominative with the copula ca or τε (both from *k^we): RV 1.2.5, 4.47.3 Vā́yav Índraś ca; Il. 3.276 f. Ζεῦ πάτερ Ἴδηθεν μεδέων κύδιστε μέγιστε, | Ἠέλιός θ᾽, ὃς πάντ᾽ ἐφορᾷς καὶ πάντ᾽ ἐπακούεις; Aesch. Sept. 140, Supp. 26; Ar. Nub. 264 f. ὦ δέσποτ᾽ ἄναξ, ἀμέτρητ᾽ Ἀήρ | ... λαμπρός τ᾽ Αἰθὴρ, σεμναί τε θεαὶ Νεφέλαι ... | ἄρθητε φάνητε. It is as if the one god is treated as being in the second person, and the other as in the third: 'O Vayu, hear, and (let) Indra (hear)'; a dual or plural verb normally follows.<sup>4

4</sup> Cf. Pind. Ol. 10.3 ὦ Μοῖσ᾽, ἀλλὰ σὺ καὶ θυγάτηρ | Ἀλάθεια Διός, ὀρθᾷ χερί | ἐρύκετον ψευδέων | ἐνιπὰν ἀλιτόξενον; PMG 884.4 Παλλὰς Τριτογένει', ἄνασσ' Ἀθάνα, | ὄρθου τήνδε πόλιν ..., σύ τε καὶ πατήρ; Berthold Delbrück, Syntaktische Forschungen (Halle 1871–88), iv.28; Renée Zwolanek, Váyav índraśca. Studien zu Anrufungsformen im Vedischen, Avestischen und Griechischen (Munich 1970).