¹ Special abbreviations: ChS: Haas, V., Salvini, M., Wegner, I., Wilhelm, G. (ed.), Corpus der hurritischen Sprachdenkmäler (Roma, 1984 ff.)Google Scholar; GHL: Bush, F. W., A Grammar of the Hurrian Language (Brandeis University, 1964)Google Scholar; GLH: Laroche, E., Glossaire de la langue hourrite (Paris, 1980)Google Scholar; HuHurr. Haas, V. (ed.), Hurriter und Hurritisch, (Xenia 21; Konstanz, 1988)Google Scholar; IH: Speiser, E. A., Inlroduction lo Hurrian, (AASOR 20; New Haven, 1941)Google Scholar; Mit.: The Mittani letter (last complete transliteration: Friedrich, J., Kleinasiatische Sprachdenkmäler, Kleine Texte für Vorlesungen und Übungen 163 (Berlin, 1932), 8–32Google Scholar; last copy: Schroeder, O., Die Tontafeln von El-Amarna, Texte Nr. 190-202, nebst Zeichenliste (VAS 12; Leipzig, 1915), Nr. 200Google Scholar; last complete translation: Les lettres d'El Amarna, traduction Moran, de W. L. (Littératures anciennes du Proche-Orient 13; Paris, 1987), 139–51Google Scholar: EA 24. A new edition with transliteration, morphemic transcription, translation, photographs, full indices of names, words and morphemes, and a bibliography is being prepared by the author for ChS II/1); Neu, Das Hurritische: Neu, E., Das Hurritische: Eine altorientalische Sprache in neuem Licht, (Akademie der Wissenschaften und der Literatur, Abhandl. der geistes- und sozialwiss. Klasse Jg. 1988/1983; Mainz/Stuttgart, 1988)Google Scholar; Fs. Otten 1988: Neu, E./Rüster, C. (ed.), Documentum Asiae Minoris Antiquae. Festschrift für Heinrich Otten zum 75. Geburtstag (Wiesbaden, 1988)Google Scholar.

² Güterbock, H. G., Kumarbi (Istanbul, 1946), 95Google Scholar.

³ Salvini, M., Sui testi mitologici in lingua hurrica, SMEA 18 (1977), 84 f.Google Scholar; Laroche, E., GLH, 279Google Scholar; Salvini, M., Die hurritischen Überlieferungen des Gilgameš-Epos und der Kešši-Erzählung, in HuHurr, 158Google Scholar.

⁴ Astour, M., Semitic Elements in the Kumarbi Myth, JNES 27 (1968), 174–5Google Scholar.

⁵ Bush, , GHL, 360 n. 76Google Scholar.

⁶ Cf. Laroche, , GLH, 279–80Google Scholar and add ul-lu-t Mit. III 44, 75, ul-lu-ḫu-ni Al.T. 138:31 (a profession), and—may be even from another stem with a long vowel u-ul-lu-u-ši-ik-ku-un-ni ChS I/1,2 Vs. 7′ (term of profession following the pattern of ašḫ = až = i = kko = nni “sacrificer”).

⁷ Speiser, IH, § 125, first supposed a pronominal meaning of this stem, but he did not distinguish it from the verbal stem mann- (for which cf. Speiser, E. A., Studies in Hurrian Grammar, JAOS 59 (1939) 302–3Google Scholar; Bush, , GHL, 88, 202, 241, 260Google Scholar; Laroche, E., in: Ugaritica 5 (Paris, 1968), 452–3Google Scholar; Friedrich, J., Churritisch, in: HdO I/2 (Leiden/Köln, 1969), 27Google Scholar; Diakonoff, , HuU, 59Google Scholar; Wilhelm, G., Hurritische Lexikographie, Or 54 (1985), 493Google Scholar; Huehnergard, J., Ugaritic Vocabulary in Syllabic Transcription, (HSS 32; Cambridge, Mass., 1987), 86Google Scholar) and the particle of introduction of direct speech ma- (for which cf. Wilhelm, G., Die Inschrift auf der Statue der Tatu-ḫepa und die hurritischen deiktischen Pronomina, SMEA 24 (1984), 220 n. 10Google Scholar; idem, Or. 54 (1985), 493; with alternative explanation C. Girbal, Der Paragraph 24 des Mittani-Briefes, ZA 78 (1988), 136Google Scholar).

⁸ Cf. Wilhelm, , SMEA 24 (1984), 221–2Google Scholar; idem, Or 54 (1985), 490, 496 (both pronouns with final -i, not -u).

⁹ Cf. Wilhelm, , Or 54 (1985), 488Google Scholar.

¹⁰ The independent pronoun of the 1st ps. sg. also should be mentioned here, though it displays some peculiarities: Absolutive: ište(n), oblique cases šo/u- except ergative ižaž.

¹¹ The same applies to more supposed attestations of nešši in this paragraph: e-eḫ-ḫi-ne-eš-ša (15, 16).

¹² For the remote possibility of such a root-complement cf. Speiser, IH, § 191, who cites the form a-ri-ir-e KUB XXVII 38 IV 25. The recently established participle ending -iri, however, has to be taken into account there; cf. G. Wilhelm, Gedanken zur Frühgeschichte der Hurriter und zum hurritisch-urartäischen Sprachvergleich, in HuHurr, 56.

¹³ Wegner, I., Phonotaktischer n-Verlust in Jussivformen des Boǧazköy-Hurritischen, Or 59 (1990), 298–305Google Scholar.

¹⁴ Neu, , Das Hurritische, 10–12Google Scholar.

¹⁵ Laroche, , GLH, 105 fGoogle Scholar.

¹⁶ Cf. šū = da = man pašš = ož = i “to me he sent” Mit. I 50; šēn(a) = iffə pašš = ož = i “my brother has sent” I 65; ḫill = oz = ī = tt(a) = ān “and I have told” II 26; pašš = ed = ī = d = an šēn (a) = i[ffu = d]a “and I will send to my brother” III 116; pic = and = ož = i = ll(a) = ān “I have rejoiced” IV 9; ḫill = ož = i “he has spoken” IV 11. For a comprehensive definition of the term “anti-passive” in the framework of Hurrian grammar cf. Ch. Girbal/I. Wegner, Zu einer neuen hurritisch-urartäischen Grammatik, ZA 77 (1987), 149 fGoogle Scholar.

¹⁷ Laroche, , GLH, 285Google Scholar.

¹⁸ Cf. Speiser, , IH, §§ 165–7Google Scholar; Bush, , GHL, 166–70Google Scholar; E. Neu, Zum hurritischen “Essiv” in der hurritisch-hethitischen Bilingue aus Hattuša, Hethitica 9 (1988), 164Google Scholar; idem, Hurritische Verbalformen auf -ai aus der hurritisch-hethitischen Bilingue, in Studia Indogermanica et Slavica, [Fs.] Werner Thomas, Specimina philologiae Slavicae, Suppl. 26 (München, 1988), 512–13Google Scholar. [The form ending in -kai (pu-ú-ta-an-ka-a-i) described by Neu as “infinite hurr. Verbalform” (p. 511) could be explained as an infinite verbal form in -ae following the verb pūd = ang-; for a root-complement -ang- cf. pir = ang = umma “to flee” (inf.) AASOR 16, no. 52:18].

¹⁹ urġ = ae = lla with the enclitic personal pronoun 3rd ps. pl, referring to the object in an ergative construction, i.e. in the given case to “sheep”.

²⁰ na-WA-[š]a is to be connected with the noun naṷni “pasture” attested in the Hurrian Hittite bilingual from Boǧazköy KBo XXXII 14 Vs. I 5; cf. Neu, E., in: HuHurr, 101Google Scholar. According to Neu's transliteration, the text says: a-a-i na-ú-ni-i-e pa-a-pa-an-ni a-me-la-a-an-ni la-a-ar-re-eš, which, with regard to the Hittite version, has to be translated “May the fire burn the mountain (serving as) pasture!” In a personal communication from Dec. 1985, E. Neu kindly informed me of the attestation of naṷni in the Boǧazköy bilingual and indicated the possibility of a relationship between naṷ- and Akkadian nawû “steppe pasture”. Neu, l.c. 102, explains the difficult ending -i-e as the enclitic possessive pronoun 3rd ps. sg. = i = in combination with the genitive -ve, which, according to Speiser, IH § 69, and Bush, , GHL 118Google Scholar, becomes -i-i-e-e in a-a-i-e-e Mit. III 28, 29 and e-ti-i-e-e Mit. I 91, 98, II 64, IV 19, 25, 28. This, however, is questionable, because forms like aš-ti-i-i-we (ašt(i) = î = ve) “his wife's” Mit. III 76 show that the rule established by Speiser is not valid. In my view (cf. already Wilhelm, G., Or n.s. 54 (1985) 494Google Scholar sub šeni) the ending -e-e has to be defined as another case- or adverbialending, hitherto overlooked, as becomes evident in Mit. III 28, 29, IV 49 f.Google Scholar: še-e-ni-íw-wu-〈ú-〉e-né-e a-a-i-(i-)e-e “in the presence of my brother” . Here the ending -e-e is repeated at the preceding genitive according to the pattern of Suffixaufnahme. The article -ne- in relational function cannot be followed by a possessive suffix, and consequently the assumed change cannot be adduced as an explanation here. Furthermore, the forms in -e-e are nowhere followed by nouns to which they could refer, as it would be expected, if they were genitives. The passage from the Hurrian Hittite bilingual containing naṷ- and naṷni has also been treated by H. Otten, Die Bronzetafel aus Boǧazköy. Ein Staatsvertrag Tuthalijas IV. (StBoT Beiheft 1), Wiesbaden 1988, 40Google Scholar. The lexeme in a-ga-bé-e-na-a-ša (agabe = n(a)= až = a) is also found as a-qa-a-we in the following passage from a Nuzi letter: ù ⌈lu⌉ a-na e-be-er-ta ša a-qa-a-we lu-⌈ú⌉ e-be-er-ta ⌈ša e⌉-ša-a-we HSS IX 5 (= AdŠ 592), 12-15; cf. CAD E 9b. There might be a connection with the pair of alternative pronouns akki—agi “the one (of two)—the other”. For the desiderative form ending in = i = l = anni attested in the bilingual passage quoted above, an exact counterpart is preserved in Urartean: ard = i = l = an(n)i “he may give!”, ḫa = i = l = an(n)i “he may take away!”; cf. Melikišvili, G. A., Die urartäische Sprache (StPohl 7), Roma 1971, 64 sqGoogle Scholar.

²¹ For this case cf. Wilhelm, G., ZA 73 (1983) 96–113CrossRefGoogle Scholar.

²² Cf. Laroche, , GLH, 52 fGoogle Scholar.

²³ Cf. E. Neu, Varia Hurritica. Sprachliche Beobachtungen an der hurritisch-hethitischen Bilingue aus Boǧazköy, in Fs. Otten 1988, 241-2.

²⁴ In the Hurrian Hittite bilingual from Boǧazköy, a variant form of this word, i-šu-uḫ-ni (ižuġni), is attested; cf. Neu, , Das Hurritische, 16 n. 39Google Scholar; idem, Zur Grammatik des Hurritischen auf der Grundlage der hurritisch-hethitischen Bilingue aus der Boǧazköy-Grabungskampagne 1983, in HuHurr, 99.

²⁵ Cf. irg = am = ōž = a = mān … itt — ummi “Keliya announced the departure of the gift” Mit. I 92; ḫa = i = en = i = l — an šēn(a) = iffu = ž fir = ādē = na šūa = lla = man 'may my brother take all the nobles” III 30 f.; cf. also III 47, 61, 85, 87, etc.

²⁶ āsk-: ChS I/5 41 Rs. 11; 75 IV 24′. ChS I/1 41 III 43, 44, 45, 76, IV 2, 5; I/5 40 Rs. 36′. For the latest functional analysis of -ōl- cf. Neu, E., in: Fs. Otten 1988, 238–46Google Scholar.

²⁷ I owe the general information about this previously unattested infinitive to an oral communication of Dr. Bahija Ismail Khalil, Baghdad. In a letter from 20 Feb., 1991, Dr. M. Müller, Leipzig, kindly communicated the text number and line (6′) to me and added: “Nach dem Kontext scheint aškumma e. zumindest allgemein “zur Rede stellen”, “zur Rechenschaft ziehen”, wahrscheinlich in dem vorliegenden Zusammenhang spezieller “Ersatz/Regress fordern” zu bedeuten.”

²⁸ ChS 1/540 Rs. 36′Google Scholar; cf. also Haas, V./Thiel, H. J., Die Beschwörungsrituale der Allaituraḫ(ḫ)i und verwandte Texte (AOAT 31; Kevelaer/Neukirchen-Vluyn, 1978), 302Google Scholar.

²⁹ Cf. H. Otten, Blick in die altorientalische Geisteswelt. Neufund einer hethitischen Tempelbibliothek (Jahrbuch der Akademie der Wissenschaften in Göttingen 1984; Göttingen, 1985), 58–9Google Scholar; Wilhelm, , in: HuHurr, 55–7Google Scholar; Neu, E., Zum hurritischen Verbum, Or 59 (1990), 230 n. 17Google Scholar.

³⁰ CAD A/2, 444b.

³¹ Cf. Adler, H.-P., Das Akkadische des Königs Tušratta von Mitanni (AOAT 201; Kevelaer/Neukirchen-Vluyn, 1976), 263Google Scholar.

³² The form tupp = ag = ō = šḫe = na Mit. II 21 shows that the suffix may be analysed as the derivational vowel -o/ō- and a suffix -šḫi/e, which supposedly is formed of the two well known suffixes -šše/-ži and -ḫḫe/-ġe.

³³ Cf. Draffkorn, Anne E., Hurrians and Hurrian at Alalaḫ: An Ethno-Linguislic Analysis (University of Pennsylvania, 1959), 217Google Scholar; Bush, , GHL, 112Google Scholar; Diakonoff, , HuU, 70 (5), 72ad (5)Google Scholar; Khačikyan, M. L., Khurritskij i urartskij jazyki (Jerevan, 1985), 65Google Scholar.

³⁴ Denominal verbs are not rare in Hurrian; cf. e.g. *ašt = ugar- “to establish a relationship by marriage” (only attested in ašt = ugar = i “relationship by marriage”) from ašti “woman, wife”; ewurumma “to inherit” from *ewuri “heir” (attested as Akkadianized ewuru); ilimd- “to swear in” from ilmi, elami “oath”; maġilumma “to buy” from *maġili < maġiri < Akk. maḫīru “purchase price”; nikkassamumma “to account” from *nikkassi < Akk. nikkassu “account”; šalaššumma “to adopt as daughter”(?) from šala “daughter”; urġ = upt- “to keep faith” from urġe “true”.

³⁵ The two root-complements -am- and -an-, which Speiser, IH, § 176 (1) and (10), had still kept apart, were taken as variants by Laroche, E., Études hourrites, RA 54 (1960), 201Google Scholar. Laroche's view has been adopted by later scholars such as Bush, , GHL, 179Google Scholar (“tentatively grouped together”), Friedrich, , Churritisch, HdO I/2 (Köln/Leiden, 1969), 17Google Scholar, Diakonoff, , HuU 114Google Scholar, and Khačikyan, M. L., Khurritskij, 60Google Scholar. The reason for the assumed allomorphy has been seen in a dissimilation of -an- to -am- in roots ending in -n like eman- and šin- (Laroche, l.c., Friedich, l.c.), and, as Bush, , GHL, 180Google Scholar, added, also in roots containing an -r- such as šur- and irg-. The appearance of -am- in the verb nikkass = am- or nissakk = am- attested at Nuzi (“to account”, derived from Akk. nikkassu), however, is not explained by these dissimilation rules. It is more likely that (following an oral suggestion of M. Krebernik) -am-and -an(n)- have to be kept apart, the first being a factitive suffix (eman = am- “to make tenfold”, šin = am- “to double”, nikkass = am- “to make an account”) and the second a causative-suffix (keb = ān- “to make (s.o.) bring”, corresponding to Akk. šūbulu; cf. Müller, G., Zu einigen hurritischen Verba mittendi, Mesopotamia 21 (1986), 230–2Google Scholar; ar = ann- “to make (s.o.) give”; cf. Laroche, l.c.) -an- is consistently written plene in connection with the root keb- (16 ×), but always non-plene after tīġ- (6×).

³⁶ Cf. Bush, , GHL, 184–6Google Scholar; Diakonoff, , HuU, 117Google Scholar, separates it from the group of root-complements, but leaves its function open.

³⁷ Cf. Bush, , GHL, 185, 351 n. 20Google Scholar.

³⁸ Also in Mit., -nna follows -en without the connecting vowel -i- which is present between -en and the pronoun 3rd ps. pl. -lla; cf. a-ra-an-ni-en-na-ma-an (ar = ann = i = e(n) = nna = man) “he may order it to be given (to him)” Mit. III 41, but a-ra-an-ni-e-ni-la-an (ar = ann = i = en = i = l = an) “and he may order them to be given (to him)” III 39.

³⁹ Speiser, , IH, § 212aGoogle Scholar; Bush, , GHL, 264–9Google Scholar; Diakonoff, , HuU, 150 fGoogle Scholar. Lastly, Girbal, C., ZA 78 (1989), 135Google Scholar, argued that the sentence particles (“associatives”) -an and -man are allomorphs of one and the same morpheme. Although it is true, that the use of -an and -man seems to be conditioned by preceding elements (e.g. -an never appears after another vowel than -a, -man never after the short forms of the enclitic pronouns), there is, however, no strict complementarity; cf. anammi = ll(a) = ān and anammi = lla = man etc.

⁴⁰ Cf. Farber, W., Zu einigen Enklitica im Hurrischen, Or 40 (1971), 39Google Scholar; Girbal, C., ZA 78 (1988), 135Google Scholar.

⁴¹ For the latest treatment of -ol- cf. Neu, , Fs. Otten 1988, 238–43Google Scholar.

⁴² ChS 1/5 87 Rs. IV 8-29 passim, 88 Rs. III 4′; cf. Thiel, H.-J., Überlegungen zu einigen Demonstrativa und Partikeln des Hurritischen, in Studies on the Civilization and Culture of Nuzi and the Hurrians in Honor of Ernest R. Lacheman (Winona Lake, 1981), 335Google Scholar; the relationship between au and the pair of alternative pronouns *akku—*agu postulated by Thiel can no longer be sustained after these pronouns have been identified as i-stems; cf. supra n. 8.

⁴³ Cf. Wilhelm, G., Zur hurritischen Gebetsliteratur, Daniels, D. R.el al., Ernten, was man sät, Festschrift Klaus Koch (Neukirchen-Vluyn, 1991), 37–47Google Scholar.

⁴⁴ Salvini, M., RA 82 (1988) 64Google Scholar; for older views cf. Laroche, E., GLH, 187 fGoogle Scholar. s.v.; Haas, V./Wilhelm, G., Hurritische und luwische Riten aus Kizzuwatna, (AOATS 3; Kevelaer/Neukirchen-Vluyn, 1974)Google Scholar, 93 with reference to luw. nu(t)-which, however, has to be kept apart according to CHD L-N 471, 476 f.

⁴⁵ The restoration was rightly suggested by the editors V. Haas and I. Wegner.

⁴⁶ Cf. Bush, , GHL, 181 fGoogle Scholar.

⁴⁷ Cf. Müller, , Mesopotamia 21 (1986), 233–6Google Scholar; Wilhelm, , ZA 79 (1989), 131Google Scholar. The Hurrian Hittite bilingual from Boǧazköy provides further attestations of nakk-, this time in the context of manumission; cf. Neu, E., Or 59 (1990), 226 fGoogle Scholar. It seems difficult, however, to arrive at the basic meaning of nakk-through the correspondence of Hurr. nakk- und Hitt. , offered by KBo XXXII 15 I/II 19′ f. The phrase nakk- kirenzi “to promulgate a manumission” (Hitt.: parā tarnumar) could easily derive from a basic meaning “to dispatch, to release, to let go” etc., comparable to the Old Assyrian phrase ṭuppam waššurum “to send a letter”.

⁴⁸ Wilhelm, G.apud Moran, W. L., Les lettres d'El-Amarna (Paris, 1987), 147Google Scholar (unpublished German and English versions: “eindringen”; “invade”).

⁴⁹ KUB XXXII 13 I 3 wa _a-šu-ma-i //II 4–5 ma-aḫ-ḫa-an … an-da-an ú-it “when he came in”; cf. Neu, , Hurritische Verbalformen auf-ai, 513 n. 40Google Scholar; idem, in Fs. Otten 1988, 245 n. 39.

⁵⁰ Mit. IV 69 should be divided into šukko “one” and ōli “another” (contra Friedrich, J., Kleinasiatische Sprachdenkmäler, 29Google Scholar).

⁵¹ Either the ergative suffix is not followed by an associative at all (by far the most cases), or by -an (Mit. IV 14, 57). Only when the ergative suffix is followed by an enclitic pronoun ending in a vowel, the associatives -an and -man appear fairly frequently; cf. with an attempt of explanation Girbal, C., ZA 78 (1988), 135CrossRefGoogle Scholar.

⁵² Spellings: Še-en-nu-un-ni and Še-en-nu-ni; cf. NPN, 131b, AAN, 126a.

⁵³ Cf. also NPN, 240b with more personal names of this type.

⁵⁴ Cf. Groneberg, B., Die Orts- und Gewässernamen der altbabylonischen Zeit (RGTC 3; Wiesbaden, 1980)Google Scholar, passim.

⁵⁵ During the Middle Babylonian period, double determination with URU and KI is fairly common; cf. Nashef, K., Die Orts- und Gewässernamen der mittelbabylonischen und mittelassyrischen Zeit (RGTC 5; Wiesbaden, 1982)Google Scholar, passim. This pattern is also found in the letter of a king of Mittani (the so-called “Sauštatar-letter”) from Nuzi: URU pa-aḫ-ḫa-ar-ra-še. KI HSS IX 1 obv. 3; URU a-ti-lu.KI ibid. 8.

⁵⁶ URU na-war TB 8001:7, 12; URU na-wa-ar TB 8002:2; URU ta-i-de ₄, ibid. 3 (cf. Illingworth, N. J. J., Inscriptions from Tell Brak 1985, Iraq 50 (1988), 101, 105Google Scholar). URU i-ḫi-be-ni Mit. I 86; URU ni-i-nu-a-a-we III 98.