Abbildungen der Seite




With regard to the vowel of the final syllable, Hesychius offers a gloss Κανδαύλας · Ἑρμῆς ἢ Ηρακλής, and on this basis some modern scholars distinguish a god Kandaulas and a king Kandaules. I think that the validity of this distinction may be doubted. Hipponax' vocative Κανδαύλα is the proper form to Κανδαύλης, cf. Λυκάμβα, Λυκάμβης; κυνάγχα, Kuvάyxns. So Hesychius, if his copyists are not at fault, has either κυνάγχης. formed the nominative wrongly from Kavdauλa2 or drawn on a nonIonic source. For Ionic we need reckon with Kavdavλns and with nothing else.

The identification of this god with Hermes seems to rest solely on the familiar line 4.2 (Diehl) of Hipponax: 'Epμñ κuváyxa, Mnovioti Kavdavλa and that seems to be an improvisation.

I can find no other evidence for a 'Еpuns κuváyxns, and Hesychius seems to have known nothing to the point, for in spite of the clear etymology3 he glosses it as vaguely as кλéлта. The invocation is put in the mouth of Boupalos and is meant to be derisive, suggested by the epithet ȧpyeïpovrns and the fact that "Apyos is a dog's name. Then it is probable that a similar twist has been given to the Maionian term Kavdaúλns. Kretschmer rightly warns against attaching any historical importance to the use of Mnovori instead of Avdiori; but stylistically the former word is high-flown and may perhaps give a hint in the same direction. At all events the great god of the Lydians must have done some mightier exploit than the choking of a dog, and I think it is so far safer to interpret Kandaules merely as "Throttler-of-the-Beast'.

It was 'a dog-like monster', says Kretschmer-trusting to the literalness of the translation Kuvάyxns-and it was this exploit that lead to the

1 Prehn in PW 10.1860, Solmsen, to some extent Kretschmer.

2 The meter does not reveal the quantity of the final vowel.

3 Even Tzetzes: σκυλλοπνίκτης.

The implicit interpretation of 'Apyéïçóvrns at so early a date is interesting and in agreement with Kretschmer's conclusions about the etymology, cf. Glotta 10.45-9 (1919).

identification with Herakles because he too throttled a monster.

The last seems to me right, and may take us farther. The monster that Heracles throttled was a lion, and the lion is so prominent in Lydia that it has been regarded as the arms of the royal family. We may then assume that it was also the throttling of a lion that gave to the Lydian god his title Kandaules, and agree (except for the unnecessary qualification) with Sayce, Sardis, 6.2.86 (note) that this god 'was doubtless represented in art like the Babylonian Gilgames, holding a strangled lion or similar animal in either hand'. The type may be seen on a Minoan gem, so that it is also attested for the Pre-Hellenic peoples of the Aegean.

This god became the founder of the Lydian dynasty whose members the Greeks called Herakleidai-having identified Kandaules with Herakles. It is probable that in it Kandaules became a titular name comparable with Caesar, Pharaoh, Syennesis. It is true that in Herodotus the name Kandaules seems personal. But Herodotus himself (1.7) tells us that the Greeks called this king Myrsilos and speaks of him as the son of Myrsos; while Nikolaos of Damascus' knows him (from Xanthos) as Adyattes or Sadyattes. From these facts Gelzer, drew the essentially correct conclusion that Kandaules was a 'sacral' name. Essentially correct may be said, for 'sacral' and 'royal' are responses to stimuli undifferentiated in such a context.


The decipherment of the Lydian inscriptions has confirmed Herodotus' use of Myrsos Myrsilos by showing that in Lydian patronymics were actually formed by such a suffix, cf. Bakivas, Bakivalis. For Kandaules however, they seem to have given no information.

But most recently Buckler' has published a series of electrum coins that antedate the gold coinage attributed to Croesus and certainly earlier than 546 B.C., the fall of Sardis. The device of the die (it appears completely on no specimen through the inexperience of its cutters) was two lion-heads confronted, with a vertical inscription between them. Thanks to a new specimen Buckler can now show that the inscription is Lydian and that it reads valves [walwe S]. He then equates valves with "Aλns and explains the legend as Gyges' cry of triumph over the conquest of Colophon: "The port on the Ales river

'Cf. B.V. Head, Brit. Mus. Excav. at Ephesos 91.

6 JHS 21.163 (1901).

FGrHist 90 F 47 (Jacoby).

8 RhM 35.5172 (1880).

JHS 46.36-41 (1926)

now belongs to the Lydian kingdom.' Against the equation valves "Aλns no objection can be brought; but the remainder of the interpretation appears most unlikely.

When the inscription was imperfectly known, the practice was to connect it with ̓Αλυάττης. That is now impossible, but it shows a fact which Buckler himself sees,10 that the natural thing to be looked for in such a legend is a 'personal, i.e. royal, name.' I would then submit the following considerations:

1. The most probable name is Kandaules; a) as being the name of Lydia's strong-arm god, and of her ever-existing priest-king, and b) because of his close connection with the Lydian lion as Throttler-of-theBeast.

2. The obvious affinity of the device to the art type posited for Kandaules by Sayce on other grounds; the vertical inscription taking the place of the male figure, the lions being represented by their heads. 3. The frequent use of ideograms and of ideograms combined with alphabetic writing in the systems of writing in vogue in Asia Minor.

I should therefore suggest that the coin may offer a combination of picture and phonetic writing and be read as KANDvalves [Kandwalwe.S]. From this form the Greek names derive readily. Epenthesis in Lydian will account for the diphthong in Kavdaúλns; while a form without epenthesis may be preserved in Kávdaλos which long ago Wilamowitz11 recognized as the non-Hellenic name of the founder of Kos. If Kavdovλos the name of one of the Κέρκωπες (but given also as 'Ανδοῦλος) belongs here, it has perhaps been assimilated to douλos by popular etymology. The same may be true of Κανδαλος glossed by Hesychius as κακοῦργος, Aporns, if it arose in a Doric-speaking community. Or it may be that behind these varying vowels we have dialectic differences of the Asia Minor languages.12

If we attempt the etymology, we must remember that the earlier efforts to explain the name started not only from a different form, but from a false hypothesis about the relationship of Lydian. That language was then put in the Thracian-Phrygian branch, whereas we now know that it meets the languages of our family only in a Pre-IndoEuropean period.

10 Note 9.

11 Hermes 18.430.

12 Cf. Lambertz, Glotta 6.17 (1914); Sturtevant, LANGUAGE 1.77 (1925) for the principle involved.

13 Kretschmer, Einl. 388-9 (1896); Hirt, Idg. 134-5, 599; Solmsen, KZ 34.77-80 (1897); 45.97–8 (1912); Herm. 46.286-91 (1911); Boisacq 541 (s.v. Kúwv).

The connection of the second half of the word, with OCS daviti 'choke', Lith. dovyti 'torment', Phryg. daos 'wolf', will have to be abandoned because the division of the compound must be Kand-valves.

It is clearly a compound in which the first member is governed by the second, of the type represented by Sanskrit dhanam-jaya-. In that type,14 the use of a case form for the first member was original. The use of a stem-form is ascribed to the analogy of the determinative compounds1 and is dated back to the time of the parent language itself. It is not surprising to find the earlier type persisting in Lydian, and I should regard kand as an accusative neuter, with the well recognized -d ending of Lydian.

The word kan I would still equate with IE kwon-, and the fact that the Lydian word does not begin with a sibilant is no longer a difficulty. Indo-European shows no trace of a form without the w-sound,16 and Solmsen must therefore assume a special law to account for its disappearance in Lydian. Starting from Kand-valves, it is easily explained in this compound as due to dissimilation. It has been suggested by others that the word is connected with Kvéw and meant originally 'young animal', 'cub', 'whelp'; if so, that it should be neuter is not surprising. The change of gender will have come in Indo-European with the specialized meaning 'dog'. What the word meant in Lydian must be uncertain: 'whelp', 'beast', 'monster', or 'lion', are all possible. If it became definitely 'lion', the semantic development would parallel that of catulus and caniculus discussed elsewhere in this issue.

Finally valves would contain an element corresponding in form to IE welu- seen in volvo, elλów, etc. The meanings seem to diverge greatly, but the idea of a rolling or twisting motion may underlie both the Lydian and the Indo-European senses of the words. Or the word may mean nothing more definite than 'killer', in which case Goth. walwjan should be compared.

To sum up, if we had the plainly written name of a Lydian king Kandvalves, we would unhesitatingly identify it with Kandaules, and we could etymologize it on somewhat the same lines as I have proposed. Kandaules is the name that we have most reason to expect to find on such coins as those discussed. The connecting link however, the assumption of rebus writing, seems to me weak. Parallels for it may be found, or a better interpretation of valves. Can the latter be a short name equivalent to Alyattes?

14 Cf. Brugmann 2.1.94.

15 Ib. 101-2.

16 On canis, cf. Kent, LANGUAGE 2.186-7 (1926).


Homenaje ofrecido a Menéndez Pidal; Miscelánea de estudios lingüísticos, literarios e históricos. Three volumes. Pp. 848, 718, and 696. Madrid, 1925.

On the completion of twenty-five years of service as professor at the Universidad Central, Madrid, the colleagues, friends and pupils of Don Ramón Menéndez Pidal have presented him this monument of linguistic, literary and historical studies. It is not possible to give here even a cursory account of each one of the valuable materials contained in the three volumes. All the great masters of Romance philology from Schuchardt, Morel-Fatio and Meyer-Lübke to the most humble pupil of Menéndez Pidal have contributed to this document of labor and love. In the first volume there is first of all an excellent photoprint of Menéndez Pidal, and then the homenaje begins with a German poem by the venerable scholar Hugo Schuchardt, where he states with deep feeling that although his physical faculties do not permit his contributing a philological article, at the same time his heart and mind still allow him to admire Spain, the Cid and Don Ramón and to send his most cordial greetings in German verse. One hundred and thirty-five scholars, representing twenty different nations, have contributed to the three volumes. A few of the outstanding contributors are: Schuchardt, Morel-Fatio, Meyer-Lübke, Wechssler, Pietsch, Schevill, Bonilla, Ribera, Rennert, Hills, Salverda de Grave, García Solalinde, Krüger, Pio Rajna, Tallgren, Navarro Tomás, Américo Castro, Rubió Lluch, Staff, Saroïhandy, Morley, Sarrailh, Millardet, Gauchat, Alonso, and Roques. The two outstanding contributions to the three volumes are, in the opinion of the reviewer, the work of Tallgren on the Arabic names for the stars, with their transcription in the time of Alfonso the learned, and the article by Navarro Tomás on Basque phonetics.

No greater honor could be conferred on don Ramón, now recognized as the greatest living Romance philologist, than this human document, a veritable encyclopedia of modern research into linguistic, literary and historical problems.


1 A brief account of the contents of each of the three volumes is given in HISPANIA 9, 191-3 (1926).

« ZurückWeiter »