Hostname: page-component-74656f9884-j7cmm Total loading time: 0 Render date: 2026-11-07T17:34:38.913Z Has data issue: false hasContentIssue false
  • English
  • Français

GYGES AND DELPHI: HERODOTUS 1.14

Published online by Cambridge University Press:  28 December 2020

Alexander Dale
Alexander Dale*
Affiliation:
Concordia University, Montreal
*
alexander.dale@keble.oxon.org
Rights & Permissions [Opens in a new window]

Extract

Herodotus’ Histories begin in earnest with Lydia and the infamous tale of the fall of Candaules and the rise of the Mermnad dynasty under Gyges. Yet, for all that Gyges was evidently a transformational figure in Lydian history and, through the story of his usurpation of the throne from Candaules, came to occupy a prominent place in the received memory of the Lydian world, Herodotus tells us very little about Gyges himself or his reign. Chapters 1.13–14 tell us about the role of the Delphic oracle in legitimizing the rule of Gyges in Lydia, as well as his lavish dedications at Delphi, in which he set a precedent followed by Alyattes and Croesus, the more notable among his successors. Yet, the remainder of his reign is summarized in less than one sentence: ‘He sent an army to Miletus and Smyrna as soon as he took power, and seized the citadel of Colophon, but no other great deed was done by him during his thirty-eight year reign … ’.

Information

Type
Research Article
Information
The Classical Quarterly , Volume 70 , Issue 2 , December 2020 , pp. 518 - 523
Check for updates
Copyright
Copyright © The Author(s), 2021. Published by Cambridge University Press on behalf of The Classical Association

Herodotus’ Histories begin in earnest with Lydia and the infamous tale of the fall of Candaules and the rise of the Mermnad dynasty under Gyges. Yet, for all that Gyges was evidently a transformational figure in Lydian history and, through the story of his usurpation of the throne from Candaules, came to occupy a prominent place in the received memory of the Lydian world, Herodotus tells us very little about Gyges himself or his reign.Footnote 1 Chapters 1.13–14 tell us about the role of the Delphic oracle in legitimizing the rule of Gyges in Lydia, as well as his lavish dedications at Delphi, in which he set a precedent followed by Alyattes and Croesus, the more notable among his successors.Footnote 2 Yet, the remainder of his reign is summarized in less than one sentence: ‘He sent an army to Miletus and Smyrna as soon as he took power, and seized the citadel of Colophon, but no other great deed was done by him during his thirty-eight year reign … ’.

Gyges’ dedications were impressive, an ‘inexplicable splendour of Ionian white and gold’. Among the multitude of silver and gold, most noteworthy were six golden bowls weighing a combined thirty talents—some 775 kg by the Attic-Euboean standard of weights. These magnificent offerings were to be seen in the treasury of the Corinthians. Herodotus continues:

ὁ δὲ χρυσὸς οὗτος καὶ ὁ ἄργυρος τὸν ὁ Γύγης ἀνέθηκε, ὑπὸ Δελφῶν καλέεται Γυγάδας ἐπὶ τοῦ ἀναθέντος ἐπωνυμίην

This gold and silver which Gyges dedicated is called Γυγάδας by the Delphians, in accordance with the eponym of the dedicator.

There is an issue in this last sentence that, as far as I am aware, has not been fully appreciated. The dedications are called Γυγάδας by the Delphians. It is evident from translators and commentators that this is taken to mean ‘Gygean’, that is, the dedication of Gyges.Footnote 3 Support for this interpretation, though seldom actively sought, is seemingly found in Pollux's Onomasticon, where, in a discussion of coinage, we find the phrase Γυγάδας χρυσός at 3.87: εὐδόκιμος δὲ καὶ ὁ Γυγάδας χρυσὸς καὶ οἱ Κροίσειοι στατῆρες, ‘highly esteemed are the Γυγάδας gold and the Croesean staters’. In other words, it is seemingly taken for granted that Γυγάδας is functioning as a relational/possessive adjective, equivalent in sense to a form *Γύγειος or *Γύγαιος, the latter perhaps presupposed in the collocation Γυγαίη λίμνη, the ‘Gygaean lake’, otherwise lake Coloë, attested from Homer onwards.Footnote 4 The problem is that one can search the annals of Greek high and low and not find a nominal derivative in -(ι)δᾱς functioning as a relational adjective. By any pattern of nominal derivation in Greek or Lydian, Γυγάδας cannot mean ‘Gygean’.

Where we do find forms in -(ι)δᾱς is, of course, in patronymics and gentilics.Footnote 5 Patronymics in -(ι)δᾱς were a productive feature of poetic diction in the Archaic period, from Πηληϊάδης to Ἀτρεΐδης to Πριαμίδης. Such patronymics do not exclusively reflect paternity, but can also reflect descent across generations: in the Iliad, Αἰακίδης can be used of both Peleus (18.433) and Achilles (2.860+). Furthermore, patronymics in -(ι)δᾱς were, by the historical period, primarily if not exclusively a feature of poetic diction. We have personal names which could reflect frozen patronymics, such as Thucydides, and clan names, such as the Ἀλκμεωνίδαι or the Πενθιλίδαι, but by the time of our written sources paternity was almost exclusively expressed with the genitive or, in the Aeolic dialects, with patronymic adjectives in -(ι)ος/-ειος. Thus Γυγάδας in Herodotus should—indeed could—only mean ‘son of’ or ‘descendant of’ Gyges, and furthermore should reflect a poetic source. We thus must ask if there is any way to reconcile a patronymic Γυγάδας with a Lydian dedication at Delphi.

In a recent article, I advanced a new interpretation of the Lydian coin legends walwet and kukalim.Footnote 6 The latter form, once thought to refer to Gyges himself and thus provide proof that coins were already being minted during Gyges’ reign in the first half of the seventh century, is best analysed as consisting of the personal name kuka-, /gūgas/, that is, Gyges, + -l(i)-, an appurtenance suffix functioning as a patronymic/gentilic, and -im, the first-person reflexive pronoun marking a first-person nominal sentence. Thus kukalim, ‘I (am) of Gugas’, would identify Alyattes, who is expressly named in the walwet legends.Footnote 7 For Lydian patronymics formed with the adjectival suffix -l(i)-, cf. manelid kumlilid silukalid, which can be interpreted as ‘of Manes, son of Kumlis, grandson of Silukas’.Footnote 8 Thus the nominative-singular common-gender Lydian patronymic of a son or, like Alyattes, descendant of Gyges would be *kukalis. Γυγάδας would be the obvious translation of such a form in, for example, a Greek verse inscription.

In light of all this, I posit that the form Γυγάδας transmitted by Herodotus reflects a Greek translation of an original Lydian dedicatory inscription in which the dedicator was identified as kukalis, and thus almost certainly refers to Alyattes. Why Alyattes and not, say, Ardys or Croesus? To begin with, we have the concrete epigraphic evidence that Alyattes did identify himself via the patronymic kukalim. We never find this used on any of Croesus’ coinage.Footnote 9 Second, the details of dynastic succession following the death of Gyges, which have always been murky, are rendered more unclear still now that the accession of Alyattes to the throne is likely to be pushed back to c.630 or even earlier.Footnote 10 Rather than the sixty-one years that Herodotus gives for Ardys and Sadyattes, we have to contend with no more than fifteen or so years between Gyges and Alyattes. Given that the Cimmerian sack of Sardis in 644/3 resulted in the death of Gyges, with a further sack in 637 no doubt doing little to ameliorate turmoil in Lydia, it is hard to see a Lydian king in this brief period having the wherewithal or inclination to give the better part of a tonne of gold to a Greek god half a world away. The archaeological and numismatic evidence points to coinage first developing under Alyattes’ reign, and it is during the reign of Alyattes that Lydian power and wealth began to truly extend beyond the Hermus valley and impinge upon the Greek world.Footnote 11

Then there is the question of the location of the dedications. Many Lydian dedications at Delphi were housed in the treasury of the Corinthians (the treasury of Cypselus).Footnote 12 Now, it is possible that most of the Lydian dedications (or, to be more precise, the dedications which by Herodotus’ time were identified as Lydian) were moved into the treasury after the great fire which destroyed the temple of Apollo in the mid sixth century. Indeed, Herodotus (1.50) tells us that the gold sculpture of a lion originally weighing ten talents that formed part of Croesus’ dedication was moved into the treasury of the Corinthians after the fire. Other Lydian dedications were moved elsewhere: two large bowls, one gold and one silver, which initially stood by the entrance to the temple, were moved after the fire, one to the treasury of the Clazomenians (a city under Lydian sway in the sixth century), the other to the forecourt of the precinct. Perhaps everything Lydian was placed in the Corinthian treasury after this time. However, given the links between Corinth under Periander and Lydia under Alyattes,Footnote 13 as well as the evidence for dedications at Delphi from the East beginning in the late seventh century,Footnote 14 it might seem overly coincidental that almost everything Lydian just happened to end up in the Corinthian treasury. As the Cypselids and the Mermnads ascended to power at roughly the same time, and likely fostered trade contacts, the association of the Lydian dedications with the treasury of Cypselus might be more than mere coincidence, and here again the evidence would point to the roughly coeval rules of Periander and Alyattes as a suitable temporal locus for Lydian dedications to be housed in the Corinthian treasury.Footnote 15

To return to the dedications in Herodotus, I suggested above that an inscription was the ultimate source for the form Γυγάδας, as well as the tradition that Gyges was the dedicator. Commentators have suggested that the long alpha in Γυγάδᾱς reflects the dialect of his Delphic guides.Footnote 16 This is highly unlikely. In the few passages where Herodotus directly quotes an inscription, manuscripts are fairly consistent in giving appropriate dialect forms, for example Doric τέτορες in the inscription for those who fought at Thermopyle (7.228). But Herodotus does not simply slip into other dialects when recalling things he has been told. The phonology of Γυγάδας tips the balance in favour of autopsy, whether of an original dedicatory inscription or of a copy made after the fire in the mid sixth century. I suspect autopsy on the part of Herodotus himself, though this was likely in tandem with information provided by temple officials at Delphi. Given that patronymics in -(ι)δᾱς were not a productive feature of everyday speech, an original verse inscription presents itself as the most plausible source of the form. That -ᾱς reflects a Phokaian dialect inscription is furthermore unlikely. The interpretation of Γυγάδας presented here is predicated on a translation/calque of an original Lydian kukalis, and the likelihood of any Phokaian being fluent in Lydian is slim to say the least. We can then entertain two possibilities: 1) that Γυγάδας reflects a stonecutter's error for the antigraphē's Γυγάδης; such slips are readily attested in interstate inscriptions;Footnote 17 2) that Γυγάδας reflects the actual phonology of the inscription, in which case I would suggest that Aeolic, rather than Doric, might be the dialect reflected in the text.Footnote 18 We have evidence for Greek poets at the Lydian court already in the time of Gyges,Footnote 19 while Sappho and Alcaeus—both presumably active during the reign of Alyattes—suggest close links between Mytilene and Sardis.Footnote 20 It would require no great feat of imagination to see how a simple inscription along the lines of, purely exempli gratia, Γυγάδας βασιλεὺς Λυδῶν με θεῶι ἀνέθηκεν or Γυγάδας μ’ ἀνέθηκεν ἑκαβόλωι Ἀπόλλωνι would have provided the requisite amount of information which, in the 150–200 years separating the dedication and Herodotus’ visit to Delphi, came to be misconstrued as that of Γύγεω (τοῦ πολυχρύσου) and not of Γυγάδᾱ(ο)/Γυγάδεω.

Herodotus explicitly uses and even quotes, supposedly verbatim, epigraphic material throughout the Histories. Furthermore, as Stephanie West has suggested in her detailed study of inscriptions in Herodotus, Herodotus likely derives information from epigraphic sources in many instances where he makes no avowed claim to do so.Footnote 21 Recently, evidence has come to light that would appear to provide a shining example of just this sort of tacit use—and misuse—of epigraphic sources, with specific reference to Lydian dedications at Greek sanctuaries. Peter Thonemann has revisited the story of Croesus’ testing of the oracles in Hdt. 1.46–54 with reference to the votive of a golden shield and spear Croesus supposedly dedicated to Amphiaraus.Footnote 22 Herodotus makes no reference to a dedicatory inscription, but simply states that in his day these were still to be seen in the temple of Apollo Ismenios at Thebes (1.52). In 2014 a new votive inscription from the Theban sanctuary of Apollo Ismenios was published which appeared to provide corroboration for Herodotus’ story.Footnote 23 The text records the ‘rediscovery’ and rededication of a votive shield which ‘Croesus’ had dedicated to Amphiaraus, with the key text in lines 4–5: ἀσπ]ίδα τὰγ Ϙροῖσος … Ἀμ]φιαρέο̄ι μνᾶμ’ ἀρετ[ᾶς … ἀνεθε̄́κεν. Not only is the inscription the likely source for Herodotus’ story of Croesus’ dedication to Amphiaraus, but also Thonemann has suggested that the Croesus in question is not the great Lydian king but a scion of the Alcmaeonidae who died in the later sixth century; this same Croesus is commemorated in the famous ‘Croesus epigram’ (CEG 27). Thonemann's arguments are compelling (if not conclusive)Footnote 24 and, if true, would provide a particularly apposite comparandum for what I have argued here, namely that, through the mists of time and with no malice aforethought, inscriptions can come to be misattributed to more famous namesakes. Just as the possession of dedicatory offerings from a fabulously wealthy and exotic Lydian king would hold greater cachet for the sanctuary in Thebes than those of a minor Athenian aristocrat, so too would the name Gyges—a figure already mythologized by the time of Herodotus and Plato—have excited the imagination more than the historically more tangible and thus more mundane figure of Alyattes, particularly given the ease of misinterpretation of the patronymic Γυγάδας.

References

1 In addition to Hdt. 1.8–14, our main source for Gyges is Nicolaus of Damascus (FGrHist 90 F 47). On the fraught question of Nicolaus’ debt to Xanthus of Lydia, see Fritz, K. von, Die griechische Geschichtsschreibung 1 (Berlin, 1967), 348–77Google Scholar. For the dates of Gyges (before 664–no later than 644/3 b.c.), see Cogan, M. and Tadmor, H., ‘Gyges and Assurbanipal: a study in literary transmission’, Orientalia 46 (1977), 65–85Google Scholar; Spalinger, A.J., ‘The date of the death of Gyges and its historical implications’, JAOS 98 (1978), 400–9Google Scholar. I follow Spalinger's dating, though Parker, V., Untersuchungen zum Lelantinischen Krieg und verwandten Problemen der frühgriechischen Geschichte (Stuttgart, 1997), 76 nGoogle Scholar. 320 warns that we can only date the death of Gyges to sometime shortly after 650 b.c.; see now Payne, A. and Wintjes, J., Lords of Asia Minor: An Introduction to the Lydians (Wiesbaden, 2016), 31–2CrossRefGoogle Scholar.

2 On the Lydian dedications, see Kerschner, M., ‘Lydische Weihungen in griechischen Heiligtümern’, in Naso, A. (ed.), Stranieri e non cittadini nei santuari greci (Florence, 2006), 253–91Google Scholar and Kaplan, P., ‘Dedications to Greek sanctuaries by foreign kings in the eighth through sixth centuries b.c.e.’, Historia 55 (2006), 129–52Google Scholar.

3 Thus the translations of e.g. Powell and Godley. Neither How and Wells nor Asheri is explicit in how they take the form. Creuzer, G.F. and Bähr, J.C.F., Herodoti Halicarnassensis Musae (Leipzig, 1856), 35Google Scholar translate ‘a Delphis vocatur Gygaeum’.

4 Whether Γυγαίη is in origin an adjective (‘the Gygaean lake’) or a noun (‘the lake Gygaiē’) is unclear (though Strabo 13.4.5 uses Γυγαίη unambiguously as a noun), but it was evidently understood as the former by the Romans (e.g. Prop. 3.11 Lydia Gygaeo tincta puella lacu, and cf. Plin. HN 5.111), whence our familiar ‘Gygaean lake’.

5 On the suffix, see Keurentjes, M., ‘The Greek patronymics in -(ί)δας/(ί)δης’, Mnemosyne 50 (1997), 385–400CrossRefGoogle Scholar, who argues that its use to form patronymics is an extension of the role of the element -ιδ- in forming gentilics and demotics. Unrelated to the present case is the use, largely restricted to iambus and comedy, of patronymics in -(ί)δης to identify people as belonging to categories or character types, for which see G. Meyer, Die stilistische Verwendung der Nominalkomposition im griechischen: Ein Beitrag zur Geschichte der διπλᾶ ὀνόματα (Leipzig, 1923), 140–6. I find implausible the arguments advanced in Guasti, D., ‘Pittaco ζοφοδορπίδαις e la storia dei sostantivi in -ίδης’, Graeco-Latina Brunensia 22 (2017), 57–67Google Scholar.

6 A. Dale, ‘walwet and kukalim: Lydian coin legends, dynastic succession, and the chronology of Mermnad kings’, Kadmos 54 (2015 [2016]), 151–66.

7 As Dale (n. 6) notes, it is now generally agreed that walwet identifies Alyattes and, given that certain kukalim coins were evidently minted after some walwet coins, the kukalim series cannot predate the reign of Alyattes. For the punch-links between kukalim and walwet coins, see Dale (n. 6), citing R.W. Wallace, ‘kukaliṃ, walwet, and the Artemision deposit: problems in early Anatolian electrum coinage’, in P. van Alfen (ed.), Agoranomia: Studies in Money and Exchange Presented to John H. Kroll (New York, 2006), 37–48. Incidentally, Poll. Onom. 3.87 (quoted above), unknown to me at the time of writing Dale (n. 6), would seem to confirm the interpretation of kukalim set forth there, with Γυγάδας χρυσός transparently reflecting an original ‘gold [i.e. electrum] (coin) of the descendant of Gyges’.

8 Text 1 line 4 in Gusmani, R., Lydisches Wörterbuch (Heidelberg, 1964), 250Google Scholar. Following Yakubovich, I., ‘An agreement between the Sardians and the Mermnads in the Lydian language?’, IF 122 (2017), 265–93, at 279–80Google Scholar, the ‘endingless’ forms in -l (e.g. manelim, artymalim and kukalim) might instead be genitive singular, though this does not affect the interpretation offered here.

9 None of the coinage issued by Croesus (the so-called ‘Croeseids’, something of a misnomer, as the universal form attested in antiquity was simply Κροίσειοι, sc. στατῆρες; cf. Poll. Onom. 3.87, quoted above) bears any legend whatsoever.

10 The backdating of Alyattes’ accession is based on reanalysis of the stratigraphy of the Artemision deposit (the source of many walwet and kukalim coins) to the third quarter of the seventh century, for which see the references in Dale (n. 6), 158–60 and M. Kerschner, ‘Das Artemision von Ephesos in geometrischer und archaischer Zeit. Die Anfänge des Heiligtums und sein Aufstieg zu einem Kultzentrum von überregionaler Bedeutung’, in T. Georges (ed.), Ephesos: Die antike Metropole in Spannungsfeld von Religion und Bildung (Tübingen, 2017), 3–67, at 45–51, 56–60. The earlier date for Alyattes integrates well with Wallace's recent reassessment of the accession of Croesus to c.587–582: Wallace, R.W., ‘Redating Croesus: Herodotean chronologies, and the dates of the earliest coinages’, JHS 136 (2016), 168–81CrossRefGoogle Scholar.

11 For the rise of Lydia under Alyattes, cf. M. Kerschner, ‘The Lydians and their Ionian and Aiolian neighbours’, in N. Cahill (ed.), The Lydians and their World (Istanbul, 2010), 247–65.

12 On the location of the Lydian dedications, see Parke, H.W., ‘Croesus and Delphi’, GRBS 25 (1984), 209–32Google Scholar; Kerschner (n. 2); Kaplan (n. 2); Scott, M., Delphi and Olympia: The Spatial Politics of Panhellenism in the Archaic and Classical Periods (Cambridge, 2010), 45–6Google Scholar.

13 Though the extent of contacts between Corinth and Sardis has been questioned: cf. Salmon, J.B., Wealthy Corinth (Oxford, 1984), 225Google Scholar; Kaplan (n. 2), 148. For the chronology of this period, and the synchronism of Periander and Alyattes, see Parker, V., ‘Zur griechischen und vorderasiatischen Chronologie des sechsten Jahrhunderts’, Historia 42 (1993), 385–417, at 389–97Google Scholar.

14 Cf. Scott (n. 12), 45–6.

15 The treasury was probably built in the late seventh/early sixth century (thus postdating Cypsilus himself), which again accords with the reigns of Alyattes and Periander (the latter traditionally dated to c.627–585). For the date of the treasury. see E. Østby, ‘Delphi and archaic Doric architecture in the Peloponnese’, in A. Jacquemin (ed.), Delphes cent ans après la grande fouille (Athens, 2000), 239–62, at 241–2.

16 Thus Asheri comments ad loc.: ‘information typically provided by Delphic guides (Herodotus even keeps the Doric form)’.

17 Buck, C.D., ‘The interstate use of the Greek dialects’, CPh 8 (1913), 133–59Google Scholar.

18 We might expect Γυγάδαις if an Aeolic form, but there have always been problems with the masculine a-stem nominative singulars in -αις in the Lesbian poets. As has often been noted, we never find this in Lesbian inscriptions, which regularly show the expected -ᾱς (cf. E.-M. Hamm, Grammatik zu Sappho und Alkaios [Berlin, 1958], 24–5, §49; W. Blümel, Die aiolischen Dialekte [Göttingen, 1982], 74–5, §86). Motivating an analogically marked masculine a-stem nominative is not transparent (we cannot posit a general shift -ᾱς > -αις on analogy with the regular -ᾰνς > -αις in the face of e.g. the feminine genitive singular in -ᾱς). The source of an analogically marked masculine nominative could be identified as the masculine nominative of athematic participles such as e.g. κέρναις, where -αις is the expected outcome of *κέρνα-ντς (contrast thematic κέρνᾱν <*κέρνα-ων in IG XII 2.1). But the lack of attestation of such an analogical reflex in the inscriptions means that the old suggestion of -αις as a hyper-Aeolicism introduced in the (post-?)Alexandrian transmission cannot be ruled out (for which, cf. Blümel [this note] and G. Liberman, Alcée: fragments (Paris, 20022), 1.xliii n. 137 with further references).

19 Namely Magnes of Smyrna, on whom see West, M.L., ‘Magnes of Smyrna: a Greek poet at the court of Gyges’, in Hellenica: Selected Papers on Greek Literature and Thought (Oxford, 2011), 1.344–52Google Scholar.

20 Cf. the prestige value attached to all things Lydian in the Lesbian poets. Here we should recall that ‘the Lydians’ gave Alcaeus and his faction 2000 staters (fr. 69 V.). Assuming that the historical context for the relevant poem comes from a period predating Pittacus’ election as aisymnētēs, traditionally dated to 597/6 (Diog. Laert. 1.75; Plut. Sol. 14.7), then Alyattes would have occupied the throne in Lydia at the relevant period.

21 West, S., ‘Herodotus’ epigraphical interests’, CQ 35 (1985), 278–305, at 303–4Google Scholar.

22 Thonemann, P., ‘Croesus and the oracles’, JHS 136 (2016), 152–67CrossRefGoogle Scholar.

23 N. Papazarkadas, ‘Two new epigrams from Thebes’, in id. (ed.), The Epigraphy and History of Boeotia: New Finds, New Prospects (Leiden and Boston, 2014), 223–51.

24 Thonemann's argument for a post-mortem dedication commemorating a fallen warrior is, in part, predicated on his supplement ἀρετ[ᾶς τε πάθας ἀνεθε̄́́κεν in line 5 of the epigram, which in turn is suggested by Herodotus’ ἀρετὴν καὶ τὴν πάθην at 1.52. Though ingenious and highly plausible, there is a certain amount of circular reasoning here.