Urartian was most likely the dominant vernacular around Lake Van and the upper Zab valley. It was written from the late ninth to seventh century BCE in the empire of Urartu.


IRAN vii, continued

(4) Urartian, with a Linguistic Sketch


Urartian was most likely the dominant vernacular around Lake Van and the upper Zab valley. It was written from the late ninth to seventh century B.C.E. in the empire of Urartu in an area that included modern-day Armenia (cf. Mt. Ararat), northern Iraq up to the Euphrates, Azerbaijan, and the modern Republic of Azerbaijan (Urartu: Melikishvili, 1964/1971; Kashkaĭ, 1977; Salvini, 1995; Zimansky, 1998; Hurrians: Wilhelm, 1989; Hurrian language: Neu, 1988; Wilhelm 1995, 2004). The Assyrian name for the country was Uraraṭu

On Iranian soil, the citadels of Bestam in West Azerbaijan and of Hasanlu southwest of Lake Urmia (qq.v.) are prominent traces of Urartian presence. Urartian, or one of its varieties, was most likely the language of the kingdom of Manna located along the eastern and southern shores of Lake Urmia and in Azerbaijan (Boehmer, 1964, pp. 15 ff.; Kashkaĭ, 1976, chap. 2; Zadok, 2002, p. 92). Mannean names of tribes and locations are mentioned since the ninth century, occasionally together with the Parsua and Medes. Some of their rulers show Iranian names: Bagdatta < Ir. Baga-dāta, and probably also Daiiaku/Dayuka. The onomastic evidence suggests that they were increasingly subjected to Iranization during the period between 834-600 B.C.E. Its only cognate language, but not direct dialectal predecessor, is Hurrian, which flourished from the end of the third to the end of the second millennium B.C.E. in the northern Fertile Crescent. In terms of linguistic affiliation, Proto-Hurro-Urartian represents an isolate, but may be remotely related to Proto-Northeast Caucasian, as discussed above. The 1st sing. clitic pronoun -me could reflect Indo-Iranian input (see below).

Urartian is written in the Akkadian cuneiform script, with an extremely reduced syllabary. The texts, including a single bilingual one (with Akkadian rendering), are almost exclusively commemorative inscriptions (more than 500), and thus refer mainly to building activities and military campaigns. Accordingly, the texts are largely repetitive and lack much grammatical information for a fuller understanding of the language.

The following sketch follows Wilhelm, 2004. While there is general agreement on the identification of morphemes and paradigmatic sets, many aspects in phonology, morphology, and syntax remain uncertain. These include semantic functions and lexical meanings. Accordingly, translations may vary widely. The meanings of less than a hundred nouns are well defined. These circumstantial problems, however, should not obscure “the actual close proximity of the two languages” (Wilhelm, 2004, p. 135).

Phonology. The cuneiform script makes establishing the phonemic system problematic. Overtly it distinguishes twenty consonants, some with restricted distribution, while some others may not be overtly distinguished. Typologically the system may have been similar to Caucasian.

There is a threefold phonemic distinction between voiceless, voiced, and a third feature which may be either aspirated voiceless, or glottalized (as found in Caucasian languages), and represented by the signs ṭ ṣ q (the so-called emphatic subset, which may have a labial member). The signs for sibilants may represent both fricatives and affricates: Chart 1.

There is a system of five (or six) vowels, i e a (o?) u, and word-final ə (< e, i). Length distinction was probably phonemic. The extent of consonantal clusters in initial, medial, and final position is not certain. Accent appears to have been on the penult syllable. Assimilation and contraction are widely found: ereli-ne-lə > erelelə “kings.” Morphology. Word classes are defined by distinct sets of derivational and inflectional suffixes, each of which expresses a single grammatical function.

Urartian has no gender distinction, but has a highly developed case system. Typologically it is an ergative language that defines both the nominal and verbal system. That is, the morphological marking of the participants is determined both syntactically and semantically, and both the intransitive subject and transitive object are in the absolute case, while the transitive subject-agent is in the marked ergative case: Chart 2.

In Urartian the ergative is found in both the indicative past tense and in all modal forms. Therefore, even though no present forms are found in the texts, the system is probably similar to that of the Caucasian languages and not confined to past tenses (tense-split) as is the case in Old Persian and Young Avestan (based on the participle in -), and in the majority of the modern non-southwestern Iranian languages. Nominal System. Derivation. There are numerous suffixes to derive nouns from other nominals and verbs, including: (a) -šə, abstracts (ulgu-šə “life,” ušma-šə “might,” arniu-šə “deed, exploit”); (b) -, nouns of profession (ere-, “king”); (c) -usə, suitability (pul-usə “stela”); (d) -, basic meaning unknown (eba- “land”); (e) -adə, unknown (hur-adi “warrior,” possibly a loanword from Hurrian).

Cases and specificity. Nine cases have so far been distinguished. The absolutive case suffixes are sg. -ø, pl. -. The plural marker *-(Hurrian -) is superseded by the case endings, except in the directive and ablative. In addition, there are relational suffixes attached to the root, sg. -ne, pl. abs. -ne, else –na, which mark the definiteness/specificity of a noun phrase (discussed further below): Chart 3.

Pronouns. Both independent pronouns and suffixed pronouns are documented only for the first and third singular. The two suffixed pronouns are 1st sing. -ukə/a “my” and 3rd sing. -iye/a ~ -i(-) “his” (e.g., ulguši-ya-nə ed-i-nə “for his life”). The 1st sing. and 3rd sing. independent pronouns are documented in various case forms. They are unique by having two distinct forms for the absolute, one for the intransitive subject, the other for the transitive object, resulting in five forms. (By coincidence, a similar five-way split is found in the Tati dialect of Hazārrud of Ṭārom in eastern Azerbaijan; Yarshater, 1970, pp. 451-67; discussed by Windfuhr, 1990). In addition is found the 1st sing. dative/patient - (unattested in Hurrian, and reminiscent of the Indo-Iranian 1st sing. clitic pronoun acc. *-, dat.-gen. *-mai). The possessive adjectives distinguish both number and case: Chart 4.

The personal suffixes - and are movable and may attach to the clause-initial component or to the verb: patient, alu-š-nə . . . “who-erg.-it/him . . .”; zad-it-u-mə “they (-it) built me (the road)”; dative, iu Ḫaldi-š-mə LUGAL-tuẖə ar-u-nə “when Ḫaldi-erg.-me kingship gave-it (-); ẖaš-i-a-lə-mə “they (-) granted me.” Deixis is binary, and distinguishes demonstrative i “this” from anaphoric (far) ina “that.” Both pronouns are documented in various case forms (shown with relational suffix -ne/a): Chart 5.

Verbal system. The verbal system distinguishes the following categories morphologically: transitivity-agency (ergative system), voice, and mood. Aspect and tense are unmarked. All indicative forms are contextually (translated as) preterits. This contrasts with Hurrian, which had the markers preterit -ōž and future -ēt .Verbal roots are either transitive or intransitive; some may be both. Overt morphological so-called class markers are: intransitive -a, transitive-causative -u, passive -ul-a (e.g., - “to enter,” aš-u- “to enter, occupy,” aš-ul-a- “to be occupied”). A small intransitive set of verbs of motion has -št-i/a-, possibly implying intensive or reflexive function (e.g., sul-išt-i-bə “I prostrated myself”).

Mood includes the indicative and a number of modal formations: imperative, jussive -in, and a set based on the modal marker -l: optative -; conditional -l-(e)yə; desiderative -i-l-anə, with class marker -i. The function of a number of other markers is little understood. Examples: Chart 6.

The negative particles are preposed ui and prohibitive mi(kui). Person marking and ergative. The texts contain no examples for 2nd person forms, except the 2nd singular imperative, and no examples of 1st plural forms, with one possible exception. All other forms are 1st singular, 3rd singular, and 3rd plural.

Person marking is complex. Basically, there are two distinct sets of suffixes: (1) the subjects of intransitives (including passive forms), and (2) the patients of transitive-causatives. Of these, the patient suffixes have the remarkable characteristic that they specify not only the patient, but at the same time the ergative agent; for example, the 3rd sing. - in its function of patient simultaneously implies 1st sing. agent. In addition, a small number of intransitive verbs of motion, marked by -št (possibly implying intensive or reflexive function), constitutes an intermediate set, with 1st sing. - instead of intr. 1st sing. -. Additionally, the implied 3rd plural agent may be expressed by the marker -t-. Synoptically the pattern may be shown as follows: Chart 7.

Examples: (1) existential verb (3s -ø), man-u-də, -ø, -lə “I, he, they was/were (there), stayed”; (2) intransitive, nun-a-də, -bə, -lə “I, he, they came”; (3) passive, aš-ul-a-bə, -lə “it/they was/were occupied”; (4) ergative, 1st sing. agent/3rd sing. patient: aš-u- lit. “I put it in”; 3rd plur. agent/3rd sing. patient: tur-ut-in-e- “they may destroy him” (jussive); (5) motion, -: 1st sing. sul-ušt-i- “I prostrated myself,” 3rd sing. a(y)-išt-i- “it jumped (horse).” The binary ergative suffix - implying 1st singular agent/3rd singular patient appears to be sometimes replaced by either the 1st intransative - or 3rd singular ergative -: ab-a- "(whatever) I requested,” qapq-ar-u-l-i- “I wanted to besiege it [the city] (optative).” It is possible that the binary function of some suffixes results from contraction: reflexive/motional 1st sing. - < *d-mə; agential 1st sing./3rd sing. - < *m-bə, 1st sing./3rd plur. - < *-m-lə. Syntax. Noun phrase. The deictic pronouns precede the noun (phrase). Modifiers usually follow their head: LUGÁL Bia-i-na-wə “king of the Bia-anians” ([-i gen. sing.], -na 3rd plur. oblique relational suffix, - 3rd plur. genitive); LUGÁL tar-y-ə “(is) the great king.” However, names in the genitive precede the head: Minua-i patarə "Minua-town.”

Dependent nouns and relational suffixes. Definite noun phrases insert relational suffixes before the case ending, sing. -ne, plur. absolute -ne, oblique -na: 3rd sing. absolute, Minua- LUGÁL tar-y-ə “M. (is) the great king,” 3rd plur. absolute, ẖuradi-ne-lə “the warriors,” 3rd plur. ergative ẖuradi-na-šə.

Typological distinct is the so-called Suffixaufnahme in genitive constructions. Both the number of the dependent noun and the case of the noun phrase as a whole are anticipated (copied) by a lexically empty grammatical dummy relational suffix (relative) + case suffix (see Table 5).

Demonstrative pronouns show the same agreement by dummy copying: ina-ne- arniuši-ne- “these exploits” (‘these, the exploits’).

There are a good number of postpositions. They are based on locational nominals, have the possessive marker, but do not enter into Suffixaufnahme. As adverbial markers, their case suffixes are mostly ablative-instrumental -, but also ablative and dative: ulguši-ya-nə “for (-) his (-(i)ya) life,” ka-i-ukə “before-me (poss.).” There is one preposition: parə “towards.”

Word order. The usual order is subject-object-verb: agent-patient-verb, but initial patient is also found. The dative and directive often follow the verb. Datives and instrumental with a god’s name are topicalized in initial position; the verb may be topicalized in vivid description: ušt-a-də Mana-idə / ebanə at-u-bə “forth I (-) marched towards Mana / the land I consumed” (- 1st sing. agent/3rd sing patient); Ḫaldi-ə M.-š i-ni pulusə kuγ-u-nə “to (-ə) Ḫaldi M. (-šə ergative) set up this stela” (- 3rd sing. agent/3rd sing. patient).” Possession is expressed by the genitive: Tariria-i i-nə uldə “T.’s (is) this vineyard.”

Complex sentences. The system of clause subordination is well developed and marked, with finite verbs in various moods.

Temporal clauses are introduced by iu and a-šə: Chart 8.

Clauses with a-šə regularly use the optative, probably containing ergative -šə as the clausal agent marker. The optative appears to imply wish or intention: a-šə teš-u-l“when they harvest the vineyard”; ašə . . . ašẖ-ašt-u-l “when they make an offering” (i.e., when they are to make). Relative clauses are introduced by al-: al-ə ab-a-də ẖaš-i(y)-a-l(ə)-mə DINGIRmeš “what I requested (-), the gods granted (-) to me (-)”; with conditional mood, al-u-šə i-nə DUB-te tur-u-l-eyə “whoever (erg. -šə) might destroy this inscription.”

E) Quotations. Quotations are introduced by al-ə “says”: LUGÁL-šal-ə [+ direct speech] “thus says the king.”

TABLE 5. Suffixaufnahme in Urartian.


Bibliography of Urartian:

R. M. Boehmer, “Volkstum und Städte der Mannäer,” BaM 3, pp. 11-24.

Igor Michailovich Diakonoff, Hurritisch und Urartäisch, tr. from Russian by Karl Sdrembek, Münchener Studien zur Sprachwissenschaft, N.S. 6, Munich, 1971.

Volkert Haas, ed., Das Reich Urartu. Ein altorientalischer Staat im 1. Jahrtausend v. Chr., Konstanzer Altorientalische Symposien, Band 1, XENIA. Konstanzer Altorientalische Vorträge und Forschungen, Heft 17, Konstanz, 1986.

Margarita L. Khachikyan [Margareta L. Xac’ik’yan], Khurritskiĭ i urartskiĭ yazyki/Khurieren ev urarteren lezunere (The Hurrian and Urartian languages), Yerevan, 1985.

Margaret L. Khačikjan, “On the Typology of the Hurro-Urartian Verb,” in Muhammad Dandamayev et al., eds., Societies and Languages of the Ancient Near East. Studies in Honour of I. M. Diakonoff, Warminster, UK, 1982, pp. 165-68.

Solmaz M. Kashkaĭ (Solmaz Mir-Taġi Gyzy Gashgaj), “O gorodakh-krepostyakh na territorii Manny” (On the city-fortresses in the land of Mannea), Drevniĭ Vostok 2, 1976, pp. 89-97.

Idem, Iz istorii Manneĭskogo tsarstva (Of the history of the Mannean kingdom), Baku, 1977.

G. A. Melikishvili, Urartskiĭ yazyk (The Urartian language), Moscow, 1964; tr. Karl Sdrembek as Die urartäische Sprache, Rome, 1971.

Erich Neu, Das Hurritische: Eine altorientalische Sprache in neuem Licht, Wiesbaden, 1988.

M. Salvini, Geschichte und Kultur der Urartäer, Darmstadt, 1995.

Gernot Wilhelm, Grundzüge der Geschichte und Kultur der Hurriter, Darmstadt, 1982; tr. Jennifer Barnes as The Hurrians, Warminster, UK, 1989.

Idem, “Suffixaufnahme in Hurrian and Urartian,” in Frans Plank, ed., Double Case: Agreement by Suffixaufnahme, New York, 1995, pp. 113-35; online at 104242401691.

Gernot Wilhelm, “Hurrian,” and “Urartian,” in Roger D. Woodard, ed., 2004, pp. 60-94 and pp. 95-118.

Ran Zadok, The Ethno-Linguistic Character of Northwestern Iran and Kurdistan in the Neo-Assyrian Period, Jaffa, Israel, 2002; also published in Iran 40, 2002, pp. 89-151.

Paul E. Zimansky, Ancient Ararat: A Handbook of Urartian Studies, Delmar, N.Y., 1998.

(Gernot Windfuhr)

Originally Published: December 15, 2006

Last Updated: March 29, 2012

This article is available in print.
Vol. XIII, Fasc. 4, pp. 390-393