The Origins of the Goddess Ariadne

III. Delos, Cyprus, and Argos

After consideration of the major variants of the Ariadne-Theseus myth, the next step must be a division on the basis of content. The last three sets discussed, those of Cyprus, Delos, and Argos, all contain one particular element absent from the Cretan and Naxian variants: a reference, however small, to the goddess Aphrodite and the consequent implication of her eventual predominance over an earlier figure. It is fitting, therefore, to deal with the variant versions of these three localities together, since the evidence of each area might well be complementary to that of the others.

At first glance, the Delian variants appear to be the most valuable in determining the precise nature of the relationship between Ariadne and Aphrodite, simply because of the greater degree to which the pair of deities enters into the basic pattern of the tale. This extensive involvement with the workings of the tale, in fact, suggests that the stories may even be inventions “attached to the myth of Theseus for the purpose of exalting the Delian image and cult.” [1] {34|35}

A more conclusive argument, however, for the ascendancy of Aphrodite over Ariadne on Delos comes from the epigraphic evidence of the island. For in inscriptions dealing with Aphrodite, the goddess is known as:

These inscriptions have been explained as recalling the name Ἀρι-άδνη by way of a variant Ἀρι-άγνη. The implication, of course, is that Aphrodite, either directly or indirectly, had become the successor of Ariadne. [
3] Yet with these opinions Nilsson takes exception, proposing rather that the inscriptions refer to the Dea Syria and can be carried no further back in time. A major failing of this argument is that it takes no account of the form of the epithet. For, even if Nilsson is correct in assigning these references to a Dea Syria who stands as the successor of Aphrodite, this information is of only minimal assistance in determining the reasons behind the choice of the epithet ἁγνή for the goddess. Why would this particular form be chosen out of a myriad of other appropriate epithets? [4] The form Ἀριάδνη/ Ἀριάγνη {35|36} may well have been indirectly involved here, passing successively to the appellations of Aphrodite and even of the Dea Syria. Those who postulated Ariadne’s early influence were probably correct even if the Dea Syria is taken into account: she involves only a further extension of their theory; Nilsson, on the other hand, may have been misled by the context of the epithet. But further consideration of this epithet ἁγνή and its relation to Ἀρι-άδνη must be preceded by mention of the Cypriote evidence.

In the Homeric evidence, Ἀριάδνη occurs only in the dative and accusative singular forms and only in line-final position:

Σ 592 Δαίδαλος ἤσκησεν καλλιπλοκάμῳ Ἀριάδνῃ.
λ 321 Φαίδρην τε Πρόκριν τε ἴδον καλήν τ᾽ Ἀριάδνην.

As for the Homeric feminine adjectival ἁγνή, it appears only in the nominative and accusative singular cases; furthermore, only in either line-final or line-initial position:

ε 123 ἕως μιν ἐν Ὀρτυγίῃ χρυσόθρονος Ἄρτεμις ἁγνή.
λ 386 ἁγνὴ Περσεφόνεια γυναικῶν θηλυτεράων.
σ 202 αἴθε μοι ὧς μαλακὸν θάνατου πόροι Ἄρτεμις ἁγνή.
υ 71 εἶδος καὶ πινυτήν, μῆκος δ᾽ ἔπορ᾽ Ἄρτεμις ἁγνή.
φ 259 ἁγνή· τίς δέ κε τόξα τιταίνοιτ᾽; ἀλλὰ ἕκηλοι.
H. to Dem. 203 πολλὰ παρασκώπτουσ᾽ ἐτρέψατο πότνιαν ἁγνήν.
H. to Dem. 337 ἁγνὴν Περσεφόνειαv ὑπὸ ζόφου ἠερόεντος. {38|39}
H. to Dem. 439 πολλὰ δ᾽ ἄρ᾽ ἀμφαγάπησε κόρην Δημήτερος ἁγνήν.

It is worth pointing out that there are no occurrences of ἁγνή the Iliad.

As for the masculine and neuter adjectival forms which appear in the Homeric corpus, these deviate only slightly from the feminine, occurring in the genitive and accusative cases and occupying only the line-initial position:

H. to Hermes 410 ἄγνου· ταὶ δ᾽ ὑπὸ ποσσὶ κατὰ χθονὸς αἶψα φύοντο.
H. to Hermes 187 ἁγνὸν ἐρισφαράγου Γαιηόχου· ἔνθα γέροντα.

Within the Hesiodic corpus, the name Ἀριάδνη occurs only once, in the accusative case and again occupying the line-final position:

Theogony 947 χρυσοκόμης δὲ Διώνυσος ξανθὴν Ἀριάδνην.

Moreover, the Hesiodic attestations of the adjectival form ἁγνό-/ ἁγνή- follow the already-described Homeric pattern, occupying the line-final position:

W. & D. 465 εὔχεσθαι δὲ Διὶ χθονίῳ Δημήτερί θ᾽ ἁγνῇ. {39|40}

and also once appearing in the slot immediately preceding the trochaic caesura:

W. & D. 122 τοὶ μὲν δαίμονες ἁγνοὶ ἐπιχθόνιοι καλέονται.

The restrictions within this large body of material are most remarkable. For one thing, all of the attestations of the appellation Ἀριάδνη and the feminine adjectival form ἁγνή are in no more than three declensional slots. Indeed, even the few masculine and neuter adjectival forms of ἁγνό- stray from these slots only sporadically (genitive singular once and nominative plural once). Secondly, the name of Ἀριάδνη always occurs in line-final position, corresponding to the line-final position of ἁγνό-/ ἁγνή-. Those instances of ἁγνό-/ ἁγνή- not in line-final position occur almost without fail in the line-initial position. The already-mentioned solitary exception, occurring before the trochaic caesura, can be derived from an original line-final position.

As for the one exception, the position immediately before the trochaic caesura, it too can be explained in terms of the line-final position. For an example, let us use one of the noun-epithet formulae cited above:

χ 251 τῷ νῦν μὴ ἅμα πάντες ἀφίετε δούρατα μακρά.

Besides this line-final position of δούρατα μακρά, the formula also appears in the slot immediately preceding the trochaic caesura:

ε 251 ἀλλ᾽ ἄγε δούρατα μακρὰ ταμὼν ἁρμόζεο χαλκῷ.

Thus, the line-final position of ἁγνή is clearly primary, and it is with this that the position of Ἀριάδνη must be compared. It too is line-final.

To sum up: every attestation of the adjectival ἁγνό-/ ἁγνή- falls in a position corresponding directly or indirectly to the final portion of the name Ἀριάδνη. This is indeed significant for corroborating the {41|42} frequently-proposed etymology of the goddess’ appellation, “very holy.” Moreover, this etymology lends strong support to the mythology, in that the Delian and Cyprian variants reveal the name of Ἀριάδνη in an epithetical phase of development.

If, then, the material from Delos is reconsidered from the comparative standpoint, the supersession of Ariadne by Aphrodite, proposed on the basis of the information gleaned from various Delian inscriptions, becomes all the more likely. However, the information derived from a close examination of the textual patterning leads to the revelation of an additional stage in the development of the goddess. Ariadne’s supposed donation of the cult image of Aphrodite and her attendance at the dances performed as part of the ceremonies which accompanied the installation of this image on Delos suggest a period of syncretism preceding the succession of Ariadne by Aphrodite. Little can be said about the length of this period of coexistence or about its precise nature. Yet the evidence points to the presence of such a transitional period between the reigns of the two figures.

The question arises whether syncretism might be the explanation for the evidence from Cyprus as well. {42|43} Indeed, the very presence of a compound formation “Ariadne-Aphrodite” suggests that at the time of this formation the name Ariadne still had an epithetical quality vis-à-vis the new goddess Aphrodite, just as it once had vis-à-vis the old Mother Goddess. But then the substantive becomes independent of its epithet, and syncretism gives way to supersession. The attested coexistence in the compound formation “Ariadne-Aphrodite” is as strong an endorsement of syncretism as the retention of the more basic epithet ἁγνή for Aphrodite in the Delian inscriptions is an endorsement of supersession. Thus, the Cyprian and Delian variants reveal different phases in an evolution from syncretism to supersession.

In Argos, the presence of the temple of Aphrodite Urania close beside that of the Cretan Dionysus, in which Ariadne was entombed, suggests the coexistence of the two goddesses. The relationship cannot be forced beyond this stage; however, the argument that coexistence is one step removed from syncretism and, accordingly, two steps from supersession, should not be overlooked. For in this light, the situation on Argos is anticipatory of a resolution similar to those which occur on Cyprus and Delos. {43|44}


[ back ] 1. Nilsson, p. 524, n. 61.

[ back ] 2. P. Roussel and M. Launey, ed. Inscriptions de Delos (Paris, 1937), Nos. 2220–2300; 2275; 2299, 2300.

[ back ] 3. Wagner, Pauly-Wissowa, “Realencyclopädie der classischen Altertumswissenschaft,” II, p. 808; Pallat, de fabula Ariadnea, Dissertation, Berlin, 1891, p. 1; Neustadt, De Jove Cretico, Dissertation, Berlin, 1906, p. 31.

[ back ] 4. Homer alone employs seven different single adjective epithets for Aphrodite, as well as three additional epithets composed of more than one word. Cf. Dunbar and Pendergast.

[ back ] 5. Nilsson, pp. 526–527.

[ back ] 6. H. W. Stoll, “Ariadne,” in W.H. Roscher, Ausführliches Lexikon der griechischen und römischen Mythologie (Leipzig, 1884–1890 , I1, p. 540.

[ back ] 7. Hjalmar Frisk, Griechisches Etymologisches Wörterbuch (Heidelberg, 1960), I, p. 138.

[ back ] 8. Puhvel, p. 166.

[ back ] 9. For a list of such positional variants, see J. B. Hainsworth, The Flexibility of the Homeric Formula (Oxford, 1968), pp. 105–109.