¹ Published with commendable promptitude by Sürenhagen, Dietrich, “Ein Königssiegel aus Kargamis” (MDOG 118 (1986), pp. 183–90)Google Scholar. An earlier and only partially preserved example of an impression of this seal was known to me from a photograph brought to the British Museum in 1981. This permitted the reading of the names of Kuzi-Tešub and Talmi-Tešub only, not the epithet. The photograph of the bulla was accompanied by another showing the reverse of a tablet with impressions of a stamp seal reading in Hieroglyphic Ku-ti-tonitrus rex.infans and a Cuneiform epigraph ^na4kišib ^Iku-un-ti-^D10-ub dumu ^Ital-mi-^Dim lugal kar-ga-miš. The tablet is now known to be in Japan: see Tsukimoto, A., Acta Sumerologica 6 (1984), pp. 68, 70Google Scholar. At the time it looked as if the element ku-zi- was merely a variant of kunti-, and that both names designated the same individual. The Lidar examples make this unlikely, providing the equivalent Hier. ku-zi- = Cun. ku-zi-, against the tablet's equivalent Hier. ku-ti- = Cun. ku-un-ti-. It is not known to me whether the bulla, like the tablet, has travelled to Japan. Nor of course is the provenance of the bulla or the tablet certainly known. Emar-Meskene seems likely for the tablet on internal evidence, as for a number of other tablets in the same collection. I do not know whether the bulla originally belonged with the tablet, but an Emar provenance for it too is not improbable.

² Sürenhagen read: “Kuzi-Tešub, King(!) of the land of Karkamiš, servant(!) of Kubaba, son of Talmi-Tešub King(!) of the land of Karkamiš, Hero(!)” (composite text from the two exemplars, translated into English). The main drawbacks to this are the forms of the signs marked (!) (Sürenhagen's notation). Albertine Hagenbuchner proposed an alternative reading based on the observation of the apparent sign kišib (taken by Sürenhagen as part of lugal) on the lower left of the circle. She would begin the reading here: “Seal ([^na4]kišib) of the King of Karkamiš, Hero, Kuzi-Tešub, of the King of Karkamiš, Hero (for “servant of Kubaba”), (!) of Talmi-Tešub”. While her reading ur.sag (not su!), in place of Sürenhagen's “servant of Kubaba”, appears to be correct, her other readings suffer from the disadvantage of the insertion of the titles both times before the royal names, also the double restoration of “son”.

³ Principally the unparalleled combination of a Hurrian element kuniti- with Luwian-muwa; and the supposed appearance of a nom. sing. MF ending -s, unparalleled in Empire Period titularies; also the impossibility of reading X + infans as rex.infans, “king's son, prince” (see following note).

⁴ deus for ku: both signs possible, but comparison with the example of either form in deus.tonitrus and ku-zi-tešub even on the photograph favours the identification as deus.

u-ni-mi- for muwa-: Sürenhagen is certainly right in observing that mu(wa) on Empire period seals may be written u + mi (but here u is always the full ox or ox-head Hieroglyph, never the cursive form). But he totally ignores the clear presence of the -ni-. For uni-, “know”, see following note.

x + infans for rex.infans: the element above infans, a small stroke or oval, certainly cannot be interpreted as rex. It must indeed be admitted that the identity, and hence the interpretation, of this element does remain uncertain. However, it can hardly be doubted, in spite of uncertainties of the reading there, that the Cuneiform legend does indeed record the affiliation of Kuzi-Tešub to Talmi-Tešub, and therefore the Hieroglyphic, whatever the status of the extra element, must do the same. Professor H. G. Güterbock tentatively suggests (by letter) the identification of the element as the numeral “one”, and an interpretation “only son” or “first son”.

⁵ The verb uni-, “know” (written in the Late Period (lituus)uni-) has been identified by myself: An. St. XXV (1975), p. 150 fGoogle Scholar. app. 2; Kadmos 19 (1980), p. 125 fGoogle Scholar. An excellent Cuneiform parallel to the present phrase appears in a titulary of Hattusilis III: šiunit k[anešša]ndaš numun-aš, “(of) the seed recognized by the god(s)” (KBo VI, 28 obv. 5)Google Scholar. A close Hieroglyphic parallel, besides that cited in Kadmos, loc. cit., n. 13, appears on maraş 1, §1h: deus-na-ti (lituus)á-za-mi-sà caput-ta-ti (lituus)u-ni-mi-sá fines-ha-ti audire-mi-sá rex-ti-sá, “the king loved by the gods, recognized by men, heard of abroad”; see my remarks An. St. XXX (1980), p. 142Google Scholar, and note that the reading (lituus)u-ni-, based on repeated collation, was not given there—I felt it to be too insecure, until the appearance of the present example provided confirmation. A useful comparable phrase has been noted on fraktin by Güterbock: … filia deus á-za-mi, “… the daughter, loved by the god(s)” (see FS Matouš (ed. B. Hruška and G. Komoróczy; Budapest, 1978 [1980]), pp. 127–136). These two Empire Period writings, the Lidar seal and fraktin, usefully demonstrate how the verbs aza-, “love”, and uni-, “know”, both written with the lituus in the Late Period, lack it earlier.

⁶ Noted by Sürenhagen in his publication of the bulla: MDOG 118 (1986), p. 189Google Scholar.

⁷ See Klengel, , GS I, p. 87 fGoogle Scholar.

⁸ A recent review of the evidence proposes a drastically reduced date for Suppiluliumas's conquest of Karkamiš, following the lowering of Egyptian chronology, to 1325 or 1322: see below, nn. 14–15. One of the fragmentary Boǧazköy references, KUB XXI, 7Google Scholar (cf. above, n. 5), points to a synchronism between Kuzi-Tešub and Suppiluliumas II, but there is nothing to establish that he was already king of Karkamiš at the time.

⁹ For recent considerations of the date of this event, see Otten, H., “Die letzte Phase des Hethitischen Grossreiches nach den Texten” (Sitzungsbericht Öst. Akk. Wiss. 418 (1983), pp. 13–21)Google Scholar; Singer, I., “Dating the End of the Hittite Empire” (Hethitica VIII (1987), pp. 413–421)Google Scholar. Note especially Singer's remarks on p. 418 with nn. 39 and 40.

¹⁰ The actual sequence of events is perhaps worth noting. At the XXXIV Rencontre Assyriologique Internationale held in Istanbul, 6–10 July 1987, I made the acquaintance of Dr. Hagenbuchner, and in conversation we arrived at the readings of the seal of Kuzi-Tešub as detailed above p. 99 f. and nn. 2–5. At the Rencontre, I read a paper, “Hittites and Assyrians at Malatya”, in which among other things I considered the Hieroglyphic evidence of the genealogies of the kings of Melid, including those from the inscriptions gürün + kötükale, ispekçür and darende. But it was not until later that summer on a visit to Sivas, on 7 August, to collate ispekçür, that I was able to establish the reading of the grandfather's name on that stele, and thus also on gürün and kötükale.

¹¹ Cf. my remarks in Iraq 36 (1974), pp. 76–9, esp. p. 77 n. 5Google Scholar.

¹² I had previously considered the first syllable to be la/i/u, but after the definite establishment of ku on ispekçür, one can see from photographs that ku is a perfectly possible reading on gürün too.

¹³ With the identification of kuzi- as a clear Hurrian onomastic element, the Hurrian (as opposed to the Luwian) reading of the logogram tonitrus follows.

¹⁴ Cf. a similar approach to the Kings of the Hittite Old Kingdom and “Middle Kingdom” by Gurney, , in Anatolian Studies … Güterbock (Istanbul, 1974), pp. 105–111Google Scholar. Cf. also the remarks of Henige, , BASOR 261 (1986), pp. 60–3Google Scholar.

¹⁵ The approximate dates of the reign and the life of Suppiluliumas I were until recently generally agreed, at least approximately: accession (aetat. 20), c. 1380 B.C., reign of c. 40 years. This has now been sharply reduced by a recent review to: accession (aetat. ?), 1343 B.C., death 1322 or 1318 B.C.; see Wilhelm, G. and Boese, J., “Absolute Chronologie und die heth. Geschichte des 15 und 14. Jahrhunderts v. Chr.”, in Åström, P. (ed.), High, Middle or Low? (Acts of an International Colloquium on Absolute Chronology; Paul Åströms Forlag, Gothenburg, 1987), Part I, pp. 74–118Google Scholar. This actually has the effect of shortening the average length of the five generations in Karkamiš, Suppiluliumas–Talmi-Tešub, towards a reasonable 30 years each.

¹⁶ The most obvious difficulty of such an early date for gürün and kötükale is the occurrence of the signs za and ia, i.e. the forms differentiated from zi and i by the addition in ligature of à. While we have no specific evidence for the date of this innovation, the fact that it is not found on the karahöyük stele must be borne in mind. Cf. below, p. 105 f., for the obstacle to a high dating of gürün and kötükale presented by karahöyük.

¹⁷ As explained in my contribution to the papers of the XXXIV R.A.I. (see above, n. 10), forthcoming, ispekçur is the work of Arnuwantis “the grandson”, shown pouring a libation to the deified Arnuwantis (his grandfather) and the latter's deified wife (his grandmother). darende is the work of a man with name written logographically avis, grandson of avis-wa/i-tá-sa ₅ (genitive), shown pouring a libation to Hebat and Sarruma. The reading of avis as Arnuwantis has been demonstrated, and the identity of grandfather and grandson with those on ispekçür is considered entirely probable.

¹⁸ See Orthmann, , Untersuchungen zur spätheth. Kunst (Bonn, 1971), p. 116 f.Google Scholar; Genge, , Nordsyrisch-sudanatolische Reliefs (Copenhagen, 1979), p. 177 fGoogle Scholar.

¹⁹ It cannot be emphasized too often that the identification of this king with Sulumal, named by Tiglath-pileser III in the period 743–732 B.C. is quite out of the question; and that the dependent reading of his name as Sulumili, now general in the literature, is thus without foundation. Evidence for the reading of pugnus is still awaited. The name is probably connected with a verb pugnus-mi- occurring in the emirgazi inscriptions, for which cf. An. St. XXV (1975), p. 129 fGoogle Scholar. citation 10 (iv, vi).

²⁰ Classified by Orthmann, as his style Malatya IGoogle Scholar, and placed in his period Spätheth. II(?), linked to Karkemis II/III (Suhis-Katuwas), etc.: Untersuchungen, pp. 91 ff.; cf. id., Propyläenkunstgeschichte 14: Der Alte Orient (Berlin, 1975), p. 430, no. 353Google Scholar (“… kaum vor dem 10. Jahrhundert v. Chr. entstanden …”); also Genge, , Reliefs, pp. 172–9Google Scholar. A radio-carbon date for wood from a burnt layer underlying the Lion Gate comes out at 885–845 B.C. (Alessio, M. et al. , Radio-carbon 8 (1966), p. 405 f.Google Scholar). If this is a genuine terminus post quem for the building of the Gate, then it must be supposed that the sculptures, which can hardly be dated so late, must have been reused from an earlier structure.

²¹ The dating of further Malatya sculpture and inscriptions depends on the examples already examined: principally that of the Lion Hunt (malatya 1), author Halpasulupis, linked to the Lion Gate by the epigraph of his name (malatya 4) on the portal lion; the Stag Hunt (malatya 3), stylistically close to the Lion Hunt; and the stele izgin, author probably identical with the grandfather of Halpasulupis.

²² Text reconstructed from a number of fragmentary exemplars: Weidner, , AfO 18 (1957–1958), p. 350Google Scholar, = Grayson, , AR1 II, §96Google Scholar. The king's name occurs only on exemplars N (KAH II, no. 69, 1Google Scholar. 8 al-lu-x […]) and L (Weidner, loc. cit., Taf. XXVI, 1. 31: […m]a?-ri).

²³ Note in particular the difficulty of bringing the archaic features of the karahöyük stele into too close a juxtaposition to the style of gürün and kötükale, below pp. 105 f., 107.

²⁴ See e.g. Bittel, , Sitzungsberichte Öst. Ak. Wiss. 418 (Vienna, 1983), pp. 39 f., 46Google Scholar; Akurgal, ibid., p. 76.

²⁵ For the identification of Hitt. Malitiya, see my contribution “Hittites and Assyrians at Melid(Malatya)”, papers of the XXXIV R.A.I., forthcoming; also my contribution to R1A VII, “Melid”, §1.2, forthcoming.

²⁶ See Gonnet, , Hethitica III, pp. 18 f., 25Google Scholar.

²⁷ The Ugarit text RS 18.06 + 17.635 (Nougayrol, , PRU IV, p. 137 f.Google Scholar, 1. 20 f.); cf. the remarks of Klengel, , GS I, pp. 83, 86Google Scholar.

²⁸ E.g., in the Ugarit edicts RS 17.352, 17.68, 17.108 (PRU IV, pp. 121, 164 f.Google Scholar).

²⁹ Neve, , AA 1987/1983, pp. 401–3Google Scholar.

³⁰ Otten, , AA 1987/1983, pp. 410–2Google Scholar: note in particular §18.

³¹ Tawagalawas letter, KUB XIV, 3 i 73 f.Google Scholar: see Singer, , An. St. XXXIII (1983), p. 212Google Scholar.

³² The presence of examples of impressions of his seal at Lidar Höyük, and possibly at Emar (above, note 2) may indicate his control of these places.

³³ The name is written logographically magnus.tonitrus, for which a Hurrian reading (Talmi-Tešub) would theoretically be as possible as the Luwian Ura-Tarhunzas. However, this ruler is to be identified in the later inscriptions of Katuwas, where reference is made to the “grandsons of magnus.tonitrus” (Karkamiš A 11b + c, §§4, 30). Here the name is written ^Imagnus + ra/i-tonitrus-tá/ta-, where the phonetic complement indicates the reading -Tarhunta- rather than -Tešub(a)-. Cf. also the father's Luwian name, and my remarks Iraq 36 (1974), p. 70 fGoogle Scholar.

³⁴ The reading mons.t[u] magnus.rex ka[r-ka-mi]-sà [… regio.rex] confirmed by my collation of the surviving fragment, A 16c1, in Ankara Museum.

³⁵ Note the contributions made by the recent treatments of Masson, E., in Florilegium Anatolicum (Paris, 1979), pp. 225–41Google Scholar; and Nowicki, H., KZ 95 (1981), pp. 251–73Google Scholar.

³⁶ Luwian reading, Arma-nani, perhaps to be preferred to Hurrian *Kušuh-šenni, a formation unparalleled in both elements.

³⁷ The proper understanding of this clause depends on the recognition of the verb, pes₂ + ra/I, previously misidentified as a superfluous repetition of the sign HH. no. 67.

³⁸ tonitrus can of course be read either Hurrian -Tešub or Luwian -Tarhunzas as required. The first element i(a) + ra/i- is open to a number of phonetic interpretations ranging from yara- to ir-. Ir-Tešub appears to be an attested name-form, although there are doubts as to whether ir- is phonetic or logographic and what it means—it is not recognized in Laroche's Glossaire Hourrite. The Boǧazköy occurrences are confusing: KBo III 3Google Scholar gives ^IIr-Tešub (ii 6, 9)Google Scholar = du-Tešub (ii 10, 11, 15, 17, etc.), son of Abiratta; and KUB III 14, 15, etc.Google Scholar gives DU-Tešub = KBo I 8 obv. 8Google Scholar, SUM-Tešub, son of Aziru. Klengel follows Hrozný and Weidner in taking ir as the phonetic rendering of DU without considering the further equation sum “give” = Hurr. ar (Or. 32 (1963), 41 n. 3)Google Scholar; cf. Kammenhuber, , Die Arier, 55, 64 ffGoogle Scholar. The Alalakh attestations of ^Iir-^dIM/^dTeš(š)uba appear to confirm that ir is indeed a phonetic writing, since logographic writings in the onomasticon here are not usual, apart from a few divine names. Thus if we can confidently accept Ir-Tešub as a commonly occurring name (cf. Masson, op. cit. p. 234 n. 3), the Hurrian reading of the KARAHÖYÜK Great King's name may be preferred to the Luwian.

³⁹ Otten, loc. cit. above n. 30, n. 38.

⁴⁰ For the publication of the last inscription, together with a general survey of Hartapus's inscriptions, see Alp, S., “Eine neue hieroglyphenhethitische Inschrift der Gruppe Kızıldaǧ-Karadaǧ” (in Anatolian Studies … Güterbock (Istanbul, 1974), pp. 17–27Google Scholar.

⁴¹ Wasusarmas alone also bears these titles on the inscriptions of his servants at suvasa, closely linked palaeographically to topada.

⁴² See most recently Gonnet, H., “Nouvelles données archéologiques rélatives aux inscriptions hiéroglyphiques de Hartapus à Kızıldaǧ” (Archéologie et Religions de l'Anatolie Ancienne: Mélanges … Naster (ed. Donceel, R. and Lebrun, R.; Louvain-la-Neuve, 1983), pp. 119–25)Google Scholar; also ead., Hethitica V (1983), pp. 21–8Google Scholar; and Bittel, K., “Hartapus and Kızıldaǧ” (in Ancient Anatolia: Essays … Mellink (ed. Canby, J. V. et al. ; Wisconsin, 1986), pp. 103–111)Google Scholar. The reading of the name of the king as Hartapusa by Dr. Gonnet in her last cited article (i.e. taking the -sa- as part of the stem and not the nom. sing, mf ending -s) was attributed to a remark of mine, but was unfortunately a misunderstanding. There are very few “cartouches” with royal names for the Late Period, but those of topada (also suvasa) do clearly write the names of Wasusarmas and Tuwatis with the ending -s. Since in this article the close connections between topada and the Hartapus inscriptions are emphasized, it would seem that the -sa in the Hartapu- cartouches may also be safely identified as the ending -s.

⁴³ Bittel previously inclined to the earlier date in agreement with Alp: see loc. cit. above, n. 24, pp. 40–2, 46. In his latest pronouncement he appears to concede the late dating, but announces his intention to re-examine the question (loc. cit. above, n. 42, especially p. 105 and n. 15).

⁴⁴ To be argued in detail in my treatment of topada in my forthcoming corpus of Hier. Luwian inscriptions. The grounds for the late dating of topada are: (1) historical—the identification of the names of Warpalawas and Kiyakiyas in the inscription; (2) epigraphic—the identification of Wasusarmas and Tuwatis with the rulers named in other Tabalian inscriptions (kululu 1, çiftlik, sultanhan, kayseri) which form with further inscriptions a clearly late block; and (3) palaeographic—topada shows enough usages characteristically late to permit the recognition of the archaic features as deliberate, if sometimes unsuccessful, archaisms.

⁴⁵ On the whole these form a homogeneous group characterized by incised linear execution and by very full, regular phonetic writings. They include inscriptions of rulers and their servants (kululu 1, kululu 4, çiftlik, sultanhan, bohça, karaburun, niǧde 1, andaval, bor, bulgarmaden, niǧde 2, porsuk) and of “private” individuals (hisarcik 1, kululu 5, erkilet 1–2, kululu 2, kululu 3, eǧrek), also the economic kululu lead strips. The few inscriptions executed in monumental relief style (ivriz, kayseri, aksaray) have sufficiently clear links with the incised group. Of uncertain date are only the fragment eǧrikoy, and hisarcik 2 and the çalapverdi blocks.

⁴⁶ Loc. cit. above, n. 24, p. 46.