« ZurückWeiter »
A NEW DARIUS INSCRIPTION
CARL D. BUCK
UNIVERSITY OF CHICAGO
Any addition to our scanty store of Old Persian inscriptions is certain to yield something of interest, and this is true of the new Darius inscription published by Sidney Smith in JRAS, 1926, 433 ff. It is in the usual trilingual versions, and occurs in duplicate on a gold and a silver tablet. These presumably belonged to a series of three or more (gold, silver, and baser materials), such as have been unearthed in foundation deposits. The editor gives a copy made from a photograph of the gold tablet, and restorations of the text (of the Elamite version) from a photograph of the silver tablet. A photograph of the gold tablet was received from a dealer by the Oriental Institute of the University of Chicago, and I have used this.
Just recently, after my comment was written, I have seen Herzfeld's communication in the Deutsche Literaturzeitung, 1926, 2105 ff., where is given the first account of the provenience of the gold tablet. It was found between two blocks of an ancient foundation which came to light in rebuilding a small house in Hamadan and which belongs to an extensive series of Achaemenian ruins of ancient Egbatana. Herzfeld makes no mention of the silver tablet, though he must have used it in constituting the Elamite text, which is much mutilated in the gold tablet. In the Old Persian text too, his line division after the first two lines is not that of the gold tablet. It looks as if his text, though nominally that of the gold tablet, were in reality that of the silver tablet, which is in much better condition. Mr. John P. Kellogg, who was in Persia last summer, and to whom I am indebted for first calling my attention to the find and the publications, informs me that the silver tablet was also found at Hamadan and presumably in the same building, and that the gold tablet is now in New York.
Herzfeld deals mainly with the historical and archaeological importance of the find. His brief grammatical notes have not made my own comments on the Old Persian text superfluous. He gives a letter for letter transcription (except in rendering the ideogram for "king"). The editor in JRAS has followed the more usual method of fuller transcription, but with a considerable number of errors-the misreading
of some signs (or misprints), several misinterpretations of sequences, and the omission of one whole word.
The correct reading, in the system of transcription commonly employed' in Iranian studies, but with ř (cf. Boh. ř) instead of the usual or Meillet's ç, or Weisbach's r", is as follows:
1 Dārayavauš XŠ vazarka XŠ XŠyanām XŠ dah
2 yuvnām Vištāspahyā puřa Haxāmanišiya.
3 @atiy Darayavauš XS ima xšařam tya ada
m dārayāmiy hača Sakaibiš tyaiy pa
ra Sugdam amata yātā ā Kušā hačā Hi(n)da6 uv amata yātā ā Sparda tyamaiy Aurama7 zdā frābara hya mališta bagānām. m
8 am Auramazdā pātuv utāmaiy viłam.
"Darius the great king, king of kings, king of the lands, son of Hystaspes, the Achaemenian. Says Darius the king: This (is) the kingdom that I hold, from the Scythians beyond Sogdiana to Kush, from India to Sardis, (the kingdom) which Auramazda, who is the greatest of the gods, gave me. May Auramazda protect me and my family."
Instead of an enumeration of the subject lands, such as we have elsewhere, the extent of the kingdom is here defined by the extremities, from northeast to southwest and from southeast to northwest. This orientation is of peculiar interest in connection with the conclusions of J. L. Myres, Geographical Journal 1896, 605 ff., as to the distorted axis of ancient geography.
From the fact that Sardis (i.e. Lydia) is given as the (north) western frontier, Smith thinks it probable that the inscription was written before the expedition to Samos in 516 B. C., and for the same reason Herzfeld states that it must antedate the expedition against the European Scythians in about 515 B. C. These seem somewhat doubtful inferences, since Lydia might well figure as the (north) western frontier land without excluding islands off the coast or lands further north but not so westerly. More significant perhaps are the (south) eastern and south (west)ern borders, India and Kush (Ethiopia), which are not included in the earliest of the lists of subject lands, India occurring in the second and third, Kush only in the third, the Naxš-i-Rustam list of 486 B. C. But I leave this question to the historians, and pass to some grammatical comments.
In the form of the word-divider, and in the spelling of the first syllable of Vištāspa (vii, not simply vi), our inscription differs from the Behistun inscription,
Throughout this inscription, as in many of the other short inscriptions, xšāyabiya is indicated by the ideogram, which with Tolman we represent by XŠ. For the genitive plural we expect XŠyānām as in Dar. Suez c 5, or XŠānām as in Xerx. Pers. ca 6, cb 10. Cf. gen. sg. XŠyahyā and XŠhyā = xšāyabiyahya. Smith reads -yānām and that is doubtless the form intended. But the original text has XŠyanam", the vowel sign after ya having been omitted, so that the only faithful transcription is -yanām.
In 11. 1–2 Smith reads dahyunām, which is the proper transcription for every other occurrence of the form. But here the text has dahyuvnām (dah yuvanam"), paralleled by paruvnām in NR a 6 ff. and the four copies of Art. Pers. The spelling uv is taken over from the final position, just as in the compound paruvzanānām beside paruzanānām. Cf. Skt manobhis with sandhi treatment of -as before the case-ending.
1. 2. Vištāspahya. Smith gives the second sibilant as 8, but it is s as always. As noted above, the first syllable is written here v'i, not simply v.
1. 4. Smith reads Sakibiš, remarking "elsewhere Sakā.” The correct reading of skibiša is of course Sakaibus, like bagaibiš. The plural Sakā is usual in the enumeration of the provinces, as in NR a 25 ff, Dar. Pers. e 18, though the singular occurs once, Sakam Bh. 5. 21. Noteworthy is the instrumental form after hača "from," with which the ablative is the natural construction and the usual one in both Avestan and Old Persian. At the same time one must bear in mind that in Old Persian the syncretism of ablative and instrumental is only somewhat less complete than in Latin. The forms are to a considerable extent identical, and the uses are indicated by the preposition, e.g. hačā Parsā "from Persia," but hadā karā, "with the army." Without preposition the ablative is unknown and the instrumental restricted. In the singular of a-stems, which are the most numerous, the two cases are merged in -a, through the loss of final d of the ablative. In the pronouns the ending -nā, corresponding to the Sanskrit instrumental ending -na, serves also for the ablative, and we have hačā aniyanā Dar. Pers. d 11, e 20. In the plural no form in -biya, answering to Sanskrit -bhyas is quotable, but hitherto there has been no occurrence of a plural with hačā. The hačā Sakaibiš is the first example, and in connection with the situation in the singular of a-stems it justifies the
but agrees with the Darius Persepolis inscriptions, as well as with the Naxš-iRustam inscription, so that there is no material help from this.
provisional assumption that here too the two cases were merged, the instrumental ending serving for both.
But the syncretism of ablative and instrumental has no bearing on the hačā Hi(n)dauv of 11. 5-6, which stands in contrast to the previously known hačā Bābirauš with the regular genitive-ablative form. For Hi(n)dauv (wrongly transcribed in JRAS, as if the text had du instead of the actual dou) is a locative form like Bābirauv. One may recall the fact that the formally identical Vedic sacă "with" is used with the locative, or again that in Avestan the locative forms in -ō = OPers. -auv, and in -vō, and also other locative forms sometimes serve as ablatives, as hačā gātvō, hačā bar'smən (cf. Bartholomae, Grd. d. iran. Phil. I 222, 229). But the contrast between hačā Bābirauš and hačā Hi(n)dauv remains surprising.
11. 4-5. tyaiy para Sugdam. Smith remarks that we should expect tara, and Herzfeld that para is new. It has been customary in the corresponding phrase NR a 28-9 to read Sakā tyaily ta]radraya, with tara Av. tarō, Skt. tiras "across, beyond." But the t is a restoration, for which we should now substitute p". For para = Skt. paras "beyond" (cf. Grk. Téрα, Téрav) gives the same sense.
As between Sugdam here and the previously known Suguda (suguda and saug"d") it is the latter that is abnormal with its anaptyctic vowel between stops.
11. 5, 6. amata yātā ā Kušā (Spardā) = hinc ad. In all preusque viously known passages yātā is a conjunction meaning "while" or "until." The a was known only as a postpositive with the locative, but here is used like Skt. a with the ablative. The new word amata, for which neither editor offers any satisfactory explanation,3 is, I think,
* I believe in this formal identity, despite the skepticism of some (cf. Bartholomae, Altiran. Wtb. 1753) and the fact that Delbrück's explanation (Vergl. Syntax I 752) of the shift of meaning in Iranian is insufficient. The shift is no more violent than that which once took place in Eng. with, only that here we know better the particular context in which it started; nor is it more startling than the opposite poles reached by the group represented by Latin sub and super.
In JRAS; "Is it another derivative from the same root as āmātā (if the reading is correct) variously rendered in the Bisitūn inscription Col. I, 3? In that case should amata yātā a be translated 'extending up to'?" Herzfeld: "a-mata kann wohl nur von der Wurzel mad- 'zumessen,' im jungawestischen belegt, abgeleitet werden." Such a participle would have the form masta. His further remark "Im Elamischen ist das Altpers.āmata nur an erster Stelle durch das ebenfalls neue mittuma ausgedrückt" is due to a misunderstanding of the Elamite structure. For mittuma is a postpositive with Šukta, answering to the OP para before Sug