II. THE RELATIONSHIP OF LYDIAN TO HITTITE AND TO ETRUSCAN Sayce thinks that Lydian is very close to the Hittite of Boghaz Kevi. He says (p. 49): "The two languages agree so remarkably in structure, grammar, and vocabulary as to show that they must either stand in the relationship of two sister languages or that Lydian is a dialect or a daughter of Hittite." This statement seems to me to go beyond the evidence, but some relationship between the two languages is fairly clear. Among the most remarkable resemblances are these. Lydian amu, amuk means 'of me,' while Hittite ammu, ammug means 'I.' The enclitic possessive meaning 'my' is mēs, mit in Lydian, mis, mit in Hittite. The Lydian relative, pis, pid (sometimes kud) reminds one of Hittite kuis, kuid. The demonstrative ēnas, ēẽnat of Lydian is similar to Hittite enis 'that.' Lydian has an enclitic demonstrative ms, ml, which seems to have the same stem as Hittite mas, man. The Lydian system of gender was apparently identical with the Hittite. Both languages have s in the nominative singular, masculine-feminine, and n in the accusative. The d of the Lydian neuters recurs in the Hittite pronominal declension, but not in the Hittite nouns. Just so Hittite shows in the genitive of certain pronouns, but in no nominal caseending. In the verb, the Hittite third person singular ending zi may well be cognate with the Lydian ending d; in both languages the third person plural differs from the singular in having a nasal before the stop. Both Lydian and Hittite form preterites with a suffixed k. 'Cf. Joh. Friedrich, Die bisherigen Ergebnisse der hethitischen Sprachforschung, Festschr. f. Wilhelm Streitberg, 304-18, (Heidelberg, 1924.) G.M.B. This is a long list in view of the fact that we can so far read only a few scattered phrases of the Lydian documents. Other points of contact will probably appear as our knowledge advances. For Sayce neither Lydian nor Hittite is Indo-European. I have recently published a reports on the case for the connection of Hittite. with Indo-European, and I need not go into the matter here, except to answer the two arguments which Sayce mentions in connection with his treatment of Lydian. On page 38 he discusses "the remarkable conglomeration of particles and pronominal forms at the commencement of the sentence" in Lydian and Hittite. He has elsewhere cited this feature of Hittite as evidence against its Indo-European character, and he undoubtedly wants us to interpret it in that sense here. He seems not to realize that Homeric Greek shows the same characteristic more constantly and in higher degree than either Lydian or Hittite. For example, г 442: où Yάρ TÚ TOTÉ μ' ωδέ γ' έρως φρένας ἀμφεκάλυψεν, is headed by a group of a pronoun and six adverbs. If it is the use of several pronouns in such a group that seems to Sayce to be significant, that also can be abundantly illustrated from Homer. There are three pronouns and two adverbs at the beginning of Δ 54: τάων οὔ τοι ἐγὼ προσθ ̓ ἵσταμαι οὐδέ μεγαίρω. More serious is Sayce's contention (p. 31) that "suffixes are numerous and can be attached one to the other, retaining a semi-independent character." He comes back to this point frequently and assumes (often without sufficient reason) all sorts of agglutinative processes in declension and conjugation. For example, he analyzes est "this" (neuter) into es 'this' and t, the personal ending of the third singular aorist. He observes in passing that ist 'it is' has the same second element. It must be admitted that Lydian shows some strange phenomena in the way of enclisis. For example, compare these two parallel clauses: No. 23. 11-12: elverskin sarokak. No. 24. 14-15: Artimulkin etversn sarokak. Sayce (pp. 39, 42) translates the first, 'and the anger <of the gods> let him incur(?),' and the second, 'and of Artemis the wrath let him incur(?).' The context makes it fairly clear that k in the first word of each clause is the enclitic meaning 'and,' and it is obvious that the in of etverskin is equivalent to the case-ending (accusative singular or oblique plural) of etversn. Furthermore Artimul-k-in seems to be a Classical Weekly 18. 171-175 (1925). There is a discussion of the matter by Heinrich Zimmern in the Streitberg Festgabe 439-441 (Leipzig, 1924). case form of the possessive adjective from Artimus. We have, then, the enclitic k interposed between stem and case-ending. Such things as this Sayce considers (p. 44) "sufficient to exclude Lydian from the Indo-European family of speech." Before we go as far as that we had better try what can be done with Forrer's hypothesis that Hittite <and Lydian>are elder sisters of the previously known Indo-European languages. Quite possibly they represent a branch which broke away from the parent stock before the inflectional system had been fixed in the form in which we know it. At any rate, if Lydian is related to Hittite, it must be related also to Indo-European; for the Hittite question has been settled, and not in the way which Sayce has favored. Nevertheless a word of caution is necessary. On account of its later date. Lydian is even further removed from the well-known Indo-European languages than is the Hittite of the thirteenth century B.C. We must not expect to find much resemblance in it to Greek, Latin, and Sanskrit. Any attempt to interpret the Lydian inscriptions with the help of these languages is foredoomed to failure. After discussing the connection between Lydian and Hittite, Sayce continues (p. 50): "On the other hand there is no genetic relationship between Lydian and Etruscan, which to me is a very unexpected fact." I do not quite understand the phrase "genetic relationship"; but I fully agree that there is no evidence of any such close relationship between Lydian and Etruscan as is implied by Herodotus' tale that the Etruscans migrated from Lydia. To me this does not come as a surprise. 10 Nevertheless it is true that both Lydian and Etruscan seem to have but one oblique case, which commonly ends in l. In Lydian as in Etruscan, this oblique case in I seems sometimes to be treated as a derivative adjective, which may take an additional case ending. The most important lexical agreement between them yet noted is the use of an encitic k‘and.' If, as seems very likely, Lydian proves to be akin with Carian, Lycian, etc., a more distant relationship with Etruscan will necessarily follow; for there is no doubt of the relationship of many proper names of Asia Minor with those of Etruria." III. CERTAIN LYDIAN GODS Among the most important results yet obtained from the study of the Lydian inscriptions is the identification of the names of several gods. • Mitteilungen der Deutschen Orient-Gesellschaft 61. 26-27 (1921). 11 See G. Herbig, Kleinasiatisch-Etruskische Namengleichungen. The Lydian pantheon included deities that we had previously known from Oriental sources and others whose names are familiar in Greek literature. I shall confine myself to the second group. The most frequent divine name in the inscriptions is Artimus, which is obviously to be identified with "Apreus. The name is several times accompanied by the epithet Ibsimsis 'Ephesian.' Artimus is often invoked to punish the violators of tombs, and in several inscriptions she is associated in this office with Pldāns, who can be no other than Apollo. Scholars have previously suspected that the worship of Apollo was indigenous in Asia Minor.12 His worship was more frequent and firmly established there than would be intelligible if it were merely an importation of the Greek colonists. Hitherto, however, it has been possible to reply that the barbarian god originally worshipped in the Apollo temples of Asia Minor had some different name which was Hellenized by the Ionians when they identified him with their ancestral deity.13 I recently predicted" that "the source of the cult of Apollo.... is likely to be definitely proven by our increasing knowledge of the preGreek languages of the Aegean basin." The name Pldāns in Lydian carries us a considerable distance toward this proof. It is unlikely that a god who was commonly invoked by the Lydians to protect the tombs of their dead was a Greek god or called by a Greek name. Furthermore the name Pļdāns is more primitive than its Greek counterpart, since the change of ld to ll is far more probable than the change of Il to ld.15 The inference that the name 'Aπóλλwv was originally not a Greek word is strengthened by its varying form in Greek ('Απέλλων, ̓Απείλων, "ATλoυ) and by the fact that it has no satisfactory Greek etymology. We have not as yet evidence enough to determine whether Lydian has lost an initial vowel or whether the A of the Greek forms is due to popular etymology (cf. ἀπόλλυμι, ἀπειλῶ, ἀπολύω). We have not yet any right to say, as Wilamowitz and Nilsson did in the passages above referred to, that the worship of Apollo was brought from Asia Minor by the Greeks. It is more likely that he was worshipped under a name related to Lydian Pļdāns throughout the Aegean basin 12 Wilamowitz, Hermes 38. 575-586 (1903); Martin Nilsson, Griechische Festen 102. 13 So Bethe, Festschrift Jakob Wackernagel 15. 14 CW 18. 118 (1925). Hence Kretschmer, Einleitung in die Geschichte der Griechischen Sprache, 327, assumes that Carian "Toowλdos is more primitive than "Toowλλos. Sayce (p. 39) assumes the contrary development in both these cases, but he is surely wrong; his description of ld for ll as a dentalized I is phonetically absurd, since I is a dental as well as d. before the coming of the Hellenes. The several Greek communities very likely took over both cult and name from the primitive inhabitants of their cities. Quite possibly the differences in the form of the name in Greek are partly due to dialectic differences in the primitive Aegean speech. There is nothing in the form of the name Artimus to disprove derivation from Greek "Apreus; but the latter name has no satisfactory Greek etymology, and it has all along seemed evident that Artemis of Ephesus was of barbarian origin. Furthermore Artimus was even more firmly established than Pļdāns as a protector of Lydian tombs. That she was associated with Pļdāns in Caria, as well as in Lydia, seems to follow from her epithet, Пeλdekeɩɩs, on an inscription from Caria (BCH. 12. 269). We may be confident, therefore, that Artemis as well as Apollo, came to the Greeks from the primitive Aegean languages. Another god who appears in several inscriptions is Bakis, that is Báκxos. His name does not occur in the nominative case; but this is clearly implied by the derivatives. There is, first of all, the possessive adjective Bakillis, BakillĮ, which we have already discussed. There is also a derivative or compound, with an appended element -vas, which functions as a personal name. It appears in the nominative masculinefeminine of the possessive adjective in No. 20: Nannas Bakivalis Artimul. Νάννας Διονυσικλέος 'Αρτέμιδι. This same possessive adjective appears with the nominative neuter suffix d in No. 51. The oblique case of the possessive is Bakivall (No. 10. 22). The presence of the name Báκxos in Lydian is unexpected; but we really ought to have seen that the word came into Greek from the primitive Aegean languages. The name Alóvvoos is Thracian, as Kretschmer long ago proved16 and the god has so many other names in that language that there is scarcely room there for the name Bacchus. From the fact that the Romans frequently called the god Bacchus, and scarcely ever Dionysus, we must infer that they learned the name from other than a Greek source; else why did they adopt the rare instead of the common name? Aside from Greece, the only known connecting link between the Lydians and the Romans is furnished by Etruria; and it was from Etruria" that Bacchic orgies spread over Italy about 200 B.C. To be sure, the god is called Fuflans on the bronze liver from Piacenza; but he was everywhere known by many names, and the absence of 16 Einleitung 241 f. and references. 17 See Livy 39. 8-9. |