Heracles on the Pyre and His Insatiable Belly

One of the most memorable scenes in the Hymn to Artemis of Callimachus is where Heracles, having attained the status of a hero and superseding Apollo in this office, receives the goddess on Mount Olympus (142–169). Though recently a god, Heracles seems to be only interested in the quarry Artemis might bring him back from the hunt, so he encourages the goddess to look for big game, and begins to prepare his meal.^[1] An authorial note explains his eagerness: his gluttony did not abate, even after he had ceased his earthly career and entered the abode of the gods (159–161):

οὐ γὰρ ὅγε Φρυγίῃ περ ὑπὸ δρυῒ γυῖα ϑεωϑείς

παύσατ᾽ ἀδηφαγίης· ἔτι οἱ πάρα νηδύς ἐκείνη,

τῇ ποτ᾽ ἀροτριόωντι συνήντετο Θειοδάμαντι.

In this typically Callimachean expression the two opposites (‘deification’ vs. ‘gluttony’) are expressed by means of two mythographic vignettes (‘cremation on the pyre’ vs. ‘ravenous hunger at meeting Thiodamas’).^[2] This opposition amounts to a witticism: the flames of the pyre (ὑπὸ δρυΐ)^[3] were able to consume the body of the hero and to help him acquire a divine status (ϑεωϑείς), but they could not extirpate his ill-reputed belly (νηδύς ἐκείνη), which clings to him even on Olympus. The epithet Φρυγίῃ, however, constitutes a problem, since it is not the common ‘from Oeta’, but the mysterious Φρύγιος. Although the attribute seems to be wholly ornamental, it is worthwhile examining it in detail. It will be shown next that the word is far from redundant.

First of all, Φρύγιος could be simply ‘Phrygian’, if only some connection could be found between Phrygia and Thessaly (~ Oeta). But such a connection is very hard to establish. The scholia all do their best to relate the word to Φρυγία, a mountain in Trachis, which according to them is the location of Heracles’ death on the pyre.^[4] This would square well with the penchant of Hellenistic poetry for replacing the common with the uncommon (here the abstruse in geographical matters).^[5] But there is an alternative explanation supplied by some sources in late antiquity and medieval times. These connect the toponym Φρῠγία with the verbal stem φρῡγ- (‘fry’, ‘dry’) relating to the death of Heracles by incineration.^[6] However, Bornmann is right to warn us that this notion might be influenced directly through Callimachus, since φρύγειν as an expression for cremation is rather strange.^[7] It is hard to believe in the existence of a second source, independent of Callimachus, which would also have employed the notions Φρυγία and φρύγειν concerning Heracles’ death. Given the plausibility of the direct influence of h. 3.159 on the concept of ‘Φρῠγία – φρῡγ-’, the mere existence of Φρυγία as a Thessalian place name might be doubted as well. Therefore, it is tempting to see in the whole concept an ex tempore interpretation (αὐτοσχεδίασμα), probably originating in antiquity, which had been contrived in order to explain away the problem posed by the ‘Phrygian oak’ (on the Oeta in Thessaly).^[8] So we are back to the original – simple and natural – supposition ‘Φρύγιος = Phrygian’ and constrained to practise ars nesciendi regarding the connective element between the Thessalian oaks and the Trojan region. But surely Callimachus could have found one, if he wanted to, being as fond of obscure mythical combinations as he was of rare toponymns.^[9]

But we must also be wary of being hypercritical. Stephanus of Byzantium (or more probably his source) might be blamed for inventing a nonexistent place name Φρυγία by somehow misinterpreting h. 3.159;^[10] nevertheless he could have identified quite correctly the poetic etymology hidden in the words of Callimachus. This fictional etymology of Φρύγιος in the ‘deep structure’ will be not affected by the more plausible interpretation of the epithet as ‘Φρύγιος = Phrygian’ in the ‘surface structure’. Certainly it would be rash to write φρυγίῃ as a word derived directly from the verb φρύγειν,^[11] though this could be done easily without altering the paradosis. Callimachus the grammarian would never have dreamed of tampering with quantities (φρῠγ- in the attribute vs. φρῡγ- in the verb) and sacrificed the analogical correctness on behalf of a pun. With Φρύγιος the problem vanishes, as the short vowel is correct, and the association with φρῡγειν is secondary. This would suggest that the oak logs are ‘dry’, a quality which fits in well with δρῦς, although the epithet remains ornamental all the same. It is, however, a little odd that Callimachus would go to such lengths in contriving a pun, only to arrive at a conventional epithet. Parturient montes...?

Before we suggest a more plausible interpretation of the pun, let us consider the evidence for the existence of such a play on words with Φρύγιος. Callimachus seems to use the same association in his Hymn to Apollo:

καὶ μὲν ὁ δακρυόεις ἀναβάλλεται ἄλγεα πέτρος,

ὅστις ἐνὶ Φρυγίῃδιερὸς λίϑος ἐστήρικται,

μάρμαρον ἀντὶ γυναικὸς ὀϊζυρόν τι χανούσης. (h. 2.22–24)

Here the pun on Φρυγίῃ (‘Phrygian’) is made explicit by means of the contiguous word διερός.^[12] The latter is a Homeric δὶς λεγόμενον,^[13] whose meaning has been heavily disputed since antiquity.^[14] It has been interpreted on the one hand as ὑγρός (‘wet’), and on the other as ζῶν (‘living’) – the latter view being most famously championed by Aristarchus.^[15] By choosing to introduce in his poems a philological ζήτημα,^[16] Callimachus must have taken sides in the dispute. Somewhat wittily, he seems to elude a clearcut decision and takes a conciliatory stance or capitalizes on the ambiguity by letting διερός have both meanings:^[17] Niobe who has been turned into stone is a ‘living stone’^[18] – a natural wonder very palatable to a poet interested in every kind of ϑαυμάσια^[19] – as well as a ‘wet stone’.^[20] This ambiguity (ἀμφιβολία) has its rhetorical expression in the pointed arrangement of the words: ‘stone’ is counterposed to διερός (~ ζῶν), διερός (~ ὑγρός) to the notion of ‘dryness’ (Φρυγίῃ). If one is willing to accept the equation ‘διερός = ὑγρός’, one will immediately spot the etymological play with Φρυγίῃ ~ φρῡγ- (‘dry’).

There is, however, another reason for the stone being moist besides the fact that it is being washed by the tears (δακρυόεις ~ διερός)^[21] of the grieving woman. Tears are actually an outward sign of the woman being alive. Thus, Callimachus does not only validate both contrasting interpretations of διερός, but he also goes out of his way to make them interrelate as two opposites of a single metaphysical concept: namely, moisture as a symbol of life and dryness as a symbol of death.^[22] This peculiar notion is expounded by Heraclitus,^[23] but it also occurs in the philological discussion about the meaning of διερός in Homer,^[24] a discourse Callimachus might have been fully conversant with.^[25] Interestingly enough, modern etymological studies have also arrived at the conclusion that διερός does not represent a case of homonymy, whose divergent meanings could be accounted for by assuming two originally separate roots, each one with a separate meaning; but here we are dealing with one single root with a broad semantic gamut arching (as it were) from the notion of humidity towards life and liveliness, and having its common denominator in some kind of mobility and agility.^[26] Thus, the witty quality of the Callimachean expression lies in the paradoxical notion that the dry stone (Φρυγίῃ ~ φρῡγ-), though being usually a symbol of death, in this exceptional case will be διερός, which is wet as well as living.^[27]

Through the common element Φρυγίῃ in the same sedes in both hymns and the pun on it relating to the symbolism of dryness and moisture in the Hymn to Apollo, we are now in a better position to understand the role of the same word play in the Hymn to Artemis. Within the structure of oppositions the following analogies between Niobe and Heracles can be discerned: the stone Niobe corresponds to the cremated (Φρυγίῃ ὑπὸ δρυΐ), but deified (γυῖα ϑεωϑείς) Heracles, the living quality of the woman to the gluttony of the hero. In the same manner as death through petrifaction the cremation entails an extreme dryness as a symbol of death: one should only think of the same lack of substance in the case of the εἴδωλα in the netherworld.^[28] The dryness is expressed in both cases by the means of the etymological pun Φρυγίῃ ~ φρῡγ-. At the level of this etymology the epithet in Φρυγίῃ περ ὑπὸ δρυΐ meaning ‘dry’ explains the full force of the preposition ὑπό: expressing a causal or instrumental dative^[29] it drives home the effect of the fire (reinforced by the particle περ after the significant word) which extracts the last drop of moisture from the body, making the hero – surprisingly enough – no εἴδωλον, but a god (ϑεωϑείς). There is, however, an even bigger surprise in store for us. Heracles, though from now on a god, cannot abandon his ravenous appetite: οὐ ... / παύσατ᾽ ἀδηφαγίης (159 f.). Here one can once again perceive a connection with Niobe’s fate: while the woman does not stop weeping,^[30] the hero does not abstain from endless gobbling. In the same way as Niobe’s living quality manifests itself in her tears, Heracles’ bodily humours do not desiccate. The manner is typically Callimachean, where a pathetic scene about a tragic heroine is transformed into a comical sketch about a gluttonous hero.^[31] By doing this, the poet goes so far as to poke fun at the traditional concept of symbolical dryness and moisture.

This opposition of ‘dry death’ vs. ‘moist life’ directs our attention to the belly of Heracles (160: νηδύς) which – being the most important part of the hero – dominates the description of his Olympian existence. Even with the belly the poet does not stop short of etymologizing. But there is a difference in scope. For whereas in the case of Φρυγίῃ he did not intend to give an etymology for the word, but only used its range of association for poetical effect, for νηδύς he seems to suggest a real etymology. Due to the context of its occurence (the picture of the gormandizing hero) we are invited to think of νηδύς as an unfathomable abyss (νη-δύς) in which everything disappears. That Callimachus conceived of νηδύς as a compound from νη- (‘nether’) and -δύς (‘sink’),^[32] is shown by a fragment of the Aitia:

καὶ γὰρ ἐγὼ τὰ μὲν ὅσσα καρήατι τῆμος ἔδωκα

ξανϑὰ σὺν εὐόδμοις ἁβρὰ λίπη στεφάνοις,

ἄπνοα πάντ᾽ ἐγένοντο παρὰ χρέος, ὅσσα τ᾽ ὀδόντων

ἔνδοϑι νείαιράν τ᾽ εἰς ἀχάριστον ἔδυ,

καὶ τῶν οὐδὲν ἔμεινεν ἐς αὔριον· ὅσσα δ᾽ ἀκουαῖς

εἰσεϑέμην, ἔτι μοι μοῦνα πάρεστι τάδε. (Ait. fr. 43.12–17 Harder)

The context is obviously the description of a symposium where the poet expresses his view that he prefers the delights of the intellect to those of the body. The abundance of food is circumscribed as the stuff which sinks through the teeth to his ungrateful stomach (15: νείαιραν ... εἰς ἀχάριστον ἔδυ). Here, νείαιρα is a synonym for νηδύς. The attributive occurrences of the word in Homer clearly indicate that it was understood as ‘nether’ and then transformed to a substantive meaning ‘belly’ as the lower part of the body.^[33] If one combines the notion of ‘nether’ with the verbal stem δυ- from ἔδυ meaning ‘sink’, then the more common epic word for ‘belly’ (νηδύς) leaps to mind, and Callimachus seems to have given for this an explaining periphrasis by νείαιραν ... ἔδυ. This is quite plausible in itself, but it surely makes more sense if we recall that the diphthong ει in νείαιρα must have been a secondary development due to the metagrammatism at the end of the fifth century b.c., the original form being *νήαιρα (as evinced by the related νέατος ‘outermost’ / ‘lowest’).^[34] Callimachus may have used the Homeric word in this original form or there might have been a ζήτημα which he wanted to settle by writing νήαιραν. This form could then be corrupted to νείαιραν due to the conventional orthography of the Homeric manuscripts. But it is also conceivable that the diphthong ει used to be pronounced at that time as a simple vowel similar to eta.^[35] At any rate a strong case can be made for the first syllable of νείαιραν being tantamount to the νη- element in νηδύς. Irrespective of phonetical quibbles, the notion of an insatiable belly informs the passage in the Aitia and it should also be present in one’s mind when reading h. 3.159–161. The two passages might have been carefully constructed so as to invite a comparison with each other – in a similar way to the hymns to Apollo and Artemis relating to Φρυγίῃ. The evidence, however, is too subtle and tenuous to support chronological arguments with respect to priority: νηδύς could have been within the ken of the poet of Aitia without h. 3.160 and conversely the notion of ‘sinking’ concerning the belly might have been suggested to the poet of the Hymn to Artemis without Ait. fr. 43.15.

But we can go a step further. The voracious stomach in Ait. fr. 43 spontaneously conjures up the figure of Charybdis on account of her devouring habits. This association is favoured by the fact that in the language of the iambus, comedy and probably the everyday κοινή as well the sea monster had evolved into the common metaphor of a gorging and guzzling Gargantua,^[36] an image which accords well with the symposial context of the Callimachus passage. In addition, if this fragment is linked with Ait. fr. 178 (‘The banquet of Pollis’), as is plausibly suggested by most critics (fr. 178+43),^[37] then the Charybdis association contributes to the epic allusions of the aition about the cult of Peleus on Icus.^[38] The point of this double allusion involving the rival genres iambus and epic might be expressed as follows: one encounters Charybdises not only on the sea, but at the drinking party as well.^[39] If this imagery holds good, one can discern the name of the monster in the expression ἀχάριστον ἔδυ (15). There are two semantic components, namely ‘ingratitude’(ἀχάριστον) and the element of ‘sinking’ (ἔδυ). Phonetically the first part of the Charybdis (χαρ-) is clearly voiced. As to the iota in ἀχάριστον and ypsilon in ἔδυ, one may suspect a witty jest with the itacist pronunciation similar to that in Epigram 28.^[40] But it is more likely that this jest has been construed so as to make an appeal on the reading eye which is supposed to transpose both letters. This spell will unleash a ravenous Charybdis on the metaphorical ‘sea of the symposium’, the monster being only short of a beta in the middle and a sigma at the end.

A very similar image occurs in the Hymn to Demeter (89 f.): τὰ δ᾽ ἐς βυϑὸν οἷα ϑαλάσσας / ἀλεμάτως ἀχάριστα κατέρρεεν εἴδατα πάντα (the feasting Erysichthon). The attribute ἀχάριστα – which has only these two occurrences in our Callimachus, in this case with the objective meaning (‘which is not welcomed’ ~ ‘futile’) instead of the subjective (‘ungrateful’) – is a clear connective between Ait. fr. 43.1 and h. 6.90 and it has the same semantic implication, namely excessive eating is futile, since it does not satiate and its pleasure is transient. The context in the Hymn to Demeter is also a symposium (cf. 54 f. and 63 f.), but this time it is a perverted one.^[41] In addition, the image of the sea and the things devoured (food) reminds one of the description of Charybdis in Homer.^[42] Yet here no etymologic pun is attempted (ἀχάριστα is insufficient to establish it), although the Homeric passage presents such a pun (Hom. Od. 12.104: Χάρυβδις ἀναρρυβδεῖ μέλαν ὕδωρ).^[43] Callimachus, however, makes amends in Ait. fr. 43.15. Yet here – as in the case of Φρυγίῃ in h. 2.25 and 3.159 – no real etymology for Charybdis is suggested, but the poet seems to playfully capitalize on the associative force of ἀχάριστα and ἔδυ. In other words, Callimachus does not want to rival the Homeric etymology for the sea monster.

The results of our endeavours on the matter can be briefly stated. The classical reputation of Callimachus as poeta doctus has been reinforced once again through the interpretation of h. 3.159–161 and two related passages (h. 2.25 and Ait. fr. 43.15). By means of a skilful combination of various elements (poetic etymology, Homeric philology and religious-philosophical concepts) he is able to achieve a subtle and complex effect that might be astonishing or provocative, but it is always witty. At least to the cognoscenti.