REVIEWS

Hehnut Humbach, Die Kani~ka-lnschrift yon Surkh-KotaL Wiesbaden, 1960. 64 pp. The sub-title of this booklet is Ein Zeugnis des jiingeren Mithraismus aus Iran, which indicates the different interpretation given to this important inscription by Humbach as opposed to Maricq and Helming. The inscription, of which two copies now exist, is written in modified Greek characters in an Iranian language which was current in what is today northern Afghanistan in Kushan times. The first copy was found by the French excavators at Surkla-Kotal in 1957, the second in 1960 predictably on the opposite side of a stairway leading up the hill to a temple complex. The details of the finding and bibliography are given by Humbach. With the keenness and highly developed intuition of the Avestan specialist Humbaeh has reconstructed the entire text of the inscription whereas Maricq and Henning had made only preliminary attempts at reading scattered words. Humbach claims that the inscription is written in an old Iranian dialect while the others interpret it as a Middle Iranian tongue. Humbach further criticizes the other attempts at reading as false since they were based on etymologies of the words and not an overall view of the nature of the inscription, which, according to Humbach, is of cultic significance. The inscription is in fact not a building memorial but is written in verse and is a paean to Mithra. Oddly the inscription resembles a Yasht of the Avesta, although Humbach does not go so far as to say that it is a lost fragment of the Avesta. Now a priori there is no reason why one should not expect an inscription of religious significance dating probably from the third century A.D. on a temple in ancient Bactria. It is also conceivable that the contents of the inscription in form might be very similar to the Yashts of the Avesta. But such an inscription would be unique and if one could show that the ingenious translation by Humbach was not as satisfactory as a more prosaic, direct interpretation, then his attempt would have fallen into the same methodological error as the notorious "Feuerlehre" of Hertel. The task of revealing this error is now being undertaken by Helming, Harmatta and others. In reading Humbach's neat and complete translation with explanations one is struck by the lack of hesitation with which he sets aside difficulties and makes a complete text. One can only admire this dogmatic certainty which is also a characteristic of the schools of Altheim, and Hertel, for example. Inasmuch as other and different translations of this inscription are in process of publication I only wish to make a few brief remarks to indicate that I am not convinced of the correctness of Humbach's translation of the inscription. Interpretations of inscriptions usually are started at the end and, of course the beginning, so the two monograms in the last line at once attract one's attention. Humbach has interpreted them as combinations of Greek letters without syntactical relation

REVIEWS

243

to the rest of the inscription, spelling A E I O 0 " G o d ! " and at the very end M I T P O " M i t h r a ! " The first might be read as Humbach declares, but the latter, as he admits, presents difficulties. One's interpretation of the inscription, naturally, will determine the understanding of the monograms, but Humbach has not considered possible analogies to the monograms; his interpretation of them is unique. The immediate parallels which came to the mind of Henning were the heraldic devices, personal or family insignia common in Sasanian Iran. Further study however, makes a more precise relationship possible. The devices of the inscription are most closely related to the monograms on Graeco-Bactrian and Seleucid coins, and even more specifically to the signs on tetradrachmae of Antiochus IV. This is the first point about the devices; the second is in regard to inscriptions in general. There are three sources which should be considered: Greek inscriptions, Achaemenid and Sasanian Iranian inscriptions, and Indian inscriptions. None of the Iranian inscriptions, as far as I know, have heraldic devices similar to S u r k h - K o t a l . Greek inscriptions have very few monograms or compendia, although in later Christian inscriptions abbreviations and devices are frequent. Among relevant Indian inscriptions I have only been able to find the Kharo.sthi inscription on the Taxila silver scroll with a symbol at the end which also appears on the coins of some of the Kushan rulers. 1 It is interesting that this one example is from Knshan times as is the inscription of Surkh-Kotal. Therefore, it seems to me more probable, though by no means certain, that the two devices on the Surkh-Kotal inscription are parallels to the symbols on the coins and on the Taxila silver plate rather than names of gods as Humbach reads them. To turn to the first line, both Maricq and Humbach consider ~ o tz~),t~o ~o xc~v~q]gxo o~v~v~oto be a complete sentence, except Humbach adds the following word [3ccyo)~yyo as an appellative of Kanigka. The first word all agree is "this", but Humbach interprets the second word as "king" from Aramaic malka. Why a zeta should represent a x or even X sound escapes me since both x and ~ are found in the inscription. Further, there is no parallel for the use of malka or a derivative of it in Iran in this period except as heterograms and we surely do not have a heterogram here. If one were to refer to Semitic languages the Arabic word maliz "sanctuary, refuge area" would be a better guess and more appropriate than malkd. Finally we have a word for "king" in px0, which makes Humbach's interpretation superfluous. The word ~ ) , ~ o which occurs eight times is a problem and I can propose no answer but may raise a query whether it might not be an indigenous non-Iranian word, comparing Burushaski muyege "dwelling place". The present day words for a large building mala in Igkagimi, mala in Rugani, maal(l)a in ~ugnani and mala " r o o f " in Waigali may be related to our word. In any case their origin cannot be sought in Arabic mahal as Skrld proposed. ~ Humbach's interpretation of ~c~yo),ccyyo as Avestan *baga-daghah "possessing divine qualities" is hardly acceptable. To reject Henning's convincing explanation of the word, also attested in the fragmentary Palamedes inscription (surely not a cultic inscription), as the name of the site "the sanctuary", hodie BaghlAn, for an artificial combination of daghah- is no less than folly, especially when daehah, q. v. Pokorny sub *dens, would hardly fit the meaning of the compound proposed by Humbach. He next proposes to interpret ~ o ~ ~CcTol3~o • as Avestan *cit~ from a root ci "to pile up", plus ~ - o " G o d " a compound which he interprets as "Opferfeuer". Yet, as realized by others who have studied the inscription, ~=Toloc~o •215 must go together as amply attested in Sogdian and elsewhere in Iranian. For [3ccpyo Humbach 1 Cf. Corpus Inseriptionum Indiearurn, II, part 1 (Calcutta, 1929), p. 77. Incidentally, similar devices or tamgas, in a stylized Brahmi script appear on the headgear of several Kushan notables from Mathura. Cf. the forthcoming volume by John Rosenfield on Kushan art to be published by the University of California Press. 2 H. Sk6ld, Materialien zu den iranischen Pamirsprachen (Lund, 1936), 267.

244

REVIEWS

proposes the etymology vihara (~y'r) > *~ry (~pyo) > Sogdian ~ry'r (by "Volksetymologie"). I do not understand this since there are other examples of a parasitic -r- intruding Sogdian words borrowed from Sanscrit. Sogdian ~ry'r is thus a direct borrowing from Sanscrit and ~ p y o can have nothing to do with vihara. Obviously further critique of Humbaeh's etymologies of various words would only reveal complete disagreement with his cultic interpretation of the inscription? His equation of ~zo (para. 102, p. 36) with Semitic 'amrn "uncle" or "A~z~tcovis unacceptable while Maricq's comparison with Av. hama is much more likely. More convincing is Humbach's suggestion (p. 43) that ~ o y~vcre~y~ (line 22) is to be compared with Av. asti- "guest" and da- "giving" plus *ganzika, but not as he takes it "dem Gaste spendender Wirt". Rather it is a personal name plus title in the series with the other names. Compare the Aehaemenids' names, 'Aa~S&x~g and "Acr~qg, while *gandika would be a possible origin of y~vor in sense probably parallel to the *ganzabara of OP. On the other hand ~O~oy~vcrr may be just a Kushan personal name as those of the other kanarangs of the inscription. The last line presents a problem since after saying that it was written by both Mihr~m~m and Burzmihrpuhr the crest or sign of the latter is given, after which the name MihramAn is repeated followed by his sign. 3 The translation of Humbach, if I understand him, goes as follows: "This is the king Kani~ka, the victorious, who possesses divine attributes. By the lord of the sacrificial fire, by the ruler, was the (temple bearing) the name of Kani~ka built. By him, seed of heaven ruling far and wide, were made a press for the flowing liquid and a vat for an intoxicating drink. By the king of the sacrificial fire was set up a mixing bowl, and there was set up a goblet for the swift and strong steer of the cows by the prince of the watercourse, by the bearer of the quickly killing sword, who lavishes the flowing liquids for pleasure, who causes the drops to glitter, by the arranger of he liquids, by the king of the intoxicating drink. By him in the goblet was released the conqueror of all, the speed of the wind, for the elephant riding army leader, for the friend of the cowborn commander, the friend of the stars and water, for the ruler, the son of the gods with the two mighty (friends), the source of the streams, he (the violent stream) who destroys the canals with shining sun glory. In the temples the lord of the henbane was inflamed in the thirty-first year of the era at the beginning of the month of Nisan, the Malo (month). Thou, lord of the water course, possessing divine attributes. Thou, king of the water course who controls much profit, thou chief of the intoxicating drink, possessing one hundred cities, thou powerfully shining father of the intoxicating drink, and furthermore king of the water won from sundered cliffs. Thou, (prince) of the milk mixed (liquid) under the sunshine creating liquids and of the (watercourse) becoming in the jugs the swift and strong steer of the cows, under the prince of the watercourses, thou bearer of the swiftly killing sword, thou king of the sought-for (intoxicating drink) by the seekers of pleasure among the drinkers, master of the drink which holds the released (steer), hundred fold possessor and overseer. In the temple built (by him) was then also otherwise present he, the priest and the noble, seer, shepherd and king, the excellent chief of the intoxicating drink, controlling much profit, the hundred fold possessor and the long-beamed, spiritually born ruler, the mighty Mithra. Thou ruler, son of Kozga~ka, thou ruler, host who looks after his guest, thou ruler, thou elephant riding army leader, thou lord of subjects, thou care-taking commander, thou chief of the intoxicating drink, thou designated by the spirit, thou ruler, thou with the mind of Mithra, thou ruler, thou noble son of MithraI God, oh thou with the mind of Mithra! Mithral

R~VmWS

245

Maricq's suggestion that the last word o q z m p ~ v o is a haplograph for ~z (o~) ~up~z~o is attractive. The variant orthography of the inscription, e.g. two forms of ~ and of ~(, as well as variant spellings, e.g. o'ro and oBo for 'and' would indicate the lack of a long and strong tradition of writing of the Iranian tongue in Greek letters. Surely the language is native Bactrian with Kushan elements as the Parthian language was the tongue of the province of Parthia with Parni elements added to it. Now that a new middle Iranian language has been found we may expect new linguistic and historical vistas in a little known time and place. 4 Cambridge, Mass.

Richard N. Frye

T. Ja. Elizarenkova, Aorist v "Rigvede", Izdatel'stvo vosto~noj literatury. Moskva, 1960. 149 p p ? The Rigvedic verbal system shows a remarkable lack of regularity and symmetry. To some extent at least, the distinction between imperfect and aorist forms is determined, not by clear formal criteria but by the existence or absence of a correlation to present forms: e.g. dtirat functions as an imperfect because of the existence of a present tirdti, but druhat is an aorist, because a corresponding present form *ruh6ti is nonexistent. On the other hand, there is a category of forms which does not fit in with the general system based on the distinction of tenses and modes: in both respects the injunctive shows no clear opposition. As a result, the Veda and the Avestan Gathas show a vagueness of expression which embarrasses the modern translator. Since the Pluperfect and the Future are no doubt comparatively recent new formations, the earlier system must have consisted of Pres./Imperf., Aorist, and Perfect. Now it has long been recognized that the athematic primary endings -mi, -si, -ti, -toi, -ntoi, etc. must be due to a secondary differentiation from the secondary endings -m, -s, -t, -to, -nto, etc. This suggests the idea, that in the original verbal system the athematic present forms were not formally distinguished from the corresponding secondary forms. Starting from these observations Miss Elizarenkova assumes an original system on the sole opposition of an injunctive (expressing action and state) and a perfect (which expressed a passive state)) From this injunctive arose in later times both the present and the preterite by the addition of resp. -i and the augment, and, on the other hand, the modal distinction of subjunctive and indicative. Traces of this older state of things are supposed to survive in the discrepancies of the Rigvedie verbal system. Miss Elizarenkova, who is perfectly well acquainted with recent research in Western Europe, has exposed this theory with great competence. Since the historical background of the Vedic verbal system is far from dear, her explanation deserves serious

In an excursus of three pages called Divus Vima Kadphises, R. G6bl proposes to interpret a Kushan coin legend OI-CHO as Vima (Kadphises), a later minting in memory of Vima, on the analogy of R o m a n coins. This reading is unconvincing and also it would be unique in Kushan coinage. One would rather expect the name of a goddess. 1 An English extract, which briefly states the general theory without the ample analysis of the Rigvedic aorist-system, is contained in the Papers presented by the USSR Delegation, XXVInternational Congress of Orientalists and is entitled "On the Problem of the Development of Tenses in Old Indo-Aryan" (Moscow, 1960), 21 pp. 2

Cf. p. 122.

246

REVIEWS

consideration. However, although she does not try to support her theory by a reference to the comparatively simple Hittite verbal system (p. 133), she nevertheless thinks that the Rigveda may have preserved the traces of an evolution that was peculiar to IndoIranian, Greek and Armenian only, whereas the original proto-IE system is supposed to have been more faithfully preserved in Old Germanic and Balto-Slavonic. This explains, why the reconstruction of the earlier system has been based only on the Rigvedic evidence, with the exclusion of all comparison with the other IE. languages. Some details, however, suggest the conclusion that such a simpler verbal system could at best be attributed to the pre-IE., rather than to the pre-Vedic, period. The existence of an s-aorist as early as proto-IE, seems warranted by the correspondence between the OChSlav. aorist and the Latin s-perfect. On the other hand, the present-class with an infixed nasal must date back to a very archaic stage of proto-IE., and thus proves the existence of the present/imperfect-system at that early stage. Still, though we may have to date these reconstructions further back, to a more remote past than the author supposed, this is no reason why we should dismiss them entirely: The irregularities of the Vedic verbal system cannot be explained without some theory about its earlier prehistoric stages. Leiden

F, B. J. Kuiper