ELAM v. Elamite language



v. Elamite language

The Elamite language is known from texts in cuneiform script, most of them found at Susa but some from other sites in western and southwestern Iran and, in the east, in Fārs and ranging in date from the 24th to the 4th century B.C.E. This span can be divided into four main periods, Old, Middle, and Neo-Elamite and Late Elamite or Achaemenid; most of the surviving material comes from the last three periods, during which the language changed considerably, especially in syntactic structure. There are also indications that the language did not represent one single dialect. The genetic relationship of Elamite to other languages has not so far been established with certainty, though attempts have been made to connect it with Dravidian (McAlpin). A few lexical items are known from non-Elamite texts, chiefly from Mesopotamia.

Script and phonology.

Beginning at the end of the 4th millennium B.C.E. two other writing systems are known from ancient Iran: Proto-Elamite (see iii, above) and Linear Elamite (see iv, above), but they have not yet been deciphered. The Akkadian syllabic system, probably adopted in the 23rd century B.C.E., was used until the 4th century B.C.E. It had signs for syllables consisting of vowels (V) and various combinations of consonants and vowels CV, VC, and CVC, as well as signs for whole words (logograms) and signs serving as semantic classifiers (markers of logograms and indicators of such noun categories as divine names, male or female names, place names, and grain or wood products). Elamite phonology, however, was quite different from that of Akkadian; for instance, it may have included vowels other than the Akkadian a, i, u, and e and consonant groups unknown to Akkadian, including combinations of three consonants or two consonants at the ends of words. To express groups of two or three consonants scribes therefore had to use some VC or CV signs simply for C (e.g., kušihš, written ku-ši-ih-š(i), and kušihšta, written ku-ši-ih-(i)š-ta/ku-ši-ih-š(i)-ta). The additional vowel was generally i.

Because the syllabary was so unsuited for the Elamite language, it is almost impossible to establish the phonemic system. Nevertheless, the ways in which the scribes used the signs at their disposal reveal numerous features of Elamite phonology, including sandhi and vowel harmony, as well as, in the later period, the loss of h in initial and final position, before consonants, and sometimes also between vowels. The shift of final u to i, which was particularly common in the late period, was sometimes accompanied by vowel assimilation (e.g., puktu > pukti > pikti “help,” turu > turi > tiri “to take”). There may have been a stress accent on the first syllable, which would account for the frequent elision of vowels in the second syllable (e.g., kuti-ka, written ku-ti-ka/ku-ut-ka, pulu-hu, written pu-ul-hu) and the occasional writing of first syllables as if they had long vowels or double consonants. The elongation of the vowel of the first syllable was marked by a supplementary vowel, an h, or reduplication of the following consonant (e.g., mari/mauri/marri “to take,” muša/muhša/mušša “to account (for).”


Word formation. Elamite was an agglutinative language. A word usually consisted of a “base” (stem) plus one or more suffixes. The base ended in a vowel and was either identical with a root ending in a vowel or was formed by adding a vowel to a root ending in a consonant. The added vowel could cause the final consonant of the root to be doubled (see below). The root could be monosyllabic or disyllabic. Roots without suffixes and bases with only vocalic extensions could function as words. Word classes included nouns, pronouns, and verbs.

Root words were very common and were of the forms V, VC, VCV, VCC, VCCV, CV, CVC, CVCV, CVCC, CVCCV, and CVCVC. Most of them were nouns, either animate or inanimate (e.g., el “look,” me “rear,” te “favor,” hih “power,” hiš “name,” gil “order,” kik “sky,” kuk “protection,” nap “god,” igi “brother,” uhi “stone,” ulhi “dwelling,” husa “wood,” kiri “goddess,” lani “silver,” halti “gate,” kassu “horn,” hinap “door socket”). Others were pronouns (e.g., i “this,” ap “they”). Some roots may have expressed a general range of meaning, however.

The base might be nominal (taking only noun suffixes), verbal (taking verb suffixes and certain noun suffixes; cf. English agent nouns), or nomino-verbal (taking both noun and verb suffixes). The base normally ended in a vowel, being either a root ending in a vowel (examples of nominal bases derived from root words: me- < me “rear,” ukku- < ukku “top,” akti- < akti “enamel,” amma- < amma “mother”; examples of verbal bases: na- “to say,” ni- “to be,” sa- “to walk,” ta- “to place,” uzzu- “to arrive,” pera- “to read,” sinnu- “to come,” tingi- “to bring”; example of a nomino-verbal base derived from a root word: giri- [a nominal stem yielding the derived noun and the verbal stem] “to show gratitude, pay homage” < giri “gratitude, homage”) or a root ending in a consonant with a vocalic extension. It may be that many roots ending in consonants have so far been found only with vocalic extensions and not as single root words. Bases containing the extension -i were nominal (e.g., stems yielding only derived nouns: kiki- < kik “sky,” muri- or muru- [with vowel harmony] < mur “ground”) or nomino-verbal (e.g., gili- [nominal stem and verbal stem] “to give orders” < gil “order,” meni- [nominal stem and verbal stem] “to rule” < men “crown,” mer(r)i- [nominal stem and verbal stem] “to govern” < mer “scepter,” peti- [nominal stem and verbal stem] “to [make] fight, rebellion” < pet “fight,” ruhi- or ruhu- [with vowel harmony; nominal stem and verbal stem] “to engender, sire” < ruh “man”). Bases with extension -a, sometimes -u, were always verbal (e.g., kaz(z)a- “to strike” < kaz “blow,” tah(h) a- “to agree” < tah “pact,” tik(k)a- “to plan” < tik “project, plan,” situ- “to make happy” < sit “happiness”). The same root could receive different vocalic extensions, for instance, -i or -a/u-, thus yielding either a nominal or a verbal base (e.g., kaz “blow” > nominal stem kaz(z)i- or verbal stem kaz(z)a-, sit “happiness” > nominal stem siti- or verbal stem situ-).

Both nouns and verbs could be formed by reduplication (e.g., the noun patpat < pat “low”; verbal bases [iterative] lili- < li- “to give,” tatallu- < tallu- “to write,” hahpu- < hapu- “to understand”). Compound nouns were used (e.g., kikmurun “world” < kik “sky + murun “earth”), and verbal bases could also be formed by attaching a verb to a substantive (e.g., hunsa “to equalize, exchange, reward” < hun “equal/same” + sa “to go”).

Nominal suffixes. So-called “classifiers” were added to a nominal base to form words. Such nominal suffixes defined the noun as the speaker (first person), the person spoken to (second person), or the person or thing spoken of (third person). The suffixes were, for the first person, -k; for the second, -t; for the third, animate singular, -r, animate plural, -p, and inanimate -me (including abstract and collective nouns), -t, and -n (neutral). Inanimate nouns of the -me class could be identical with nominal and pronominal bases or verbal bases and participial forms (verbal nouns). Nominal suffixes were also attached to the locative particle aha- “there,” to the negative particle in-, to the numerals, and to country names to form ethnic nouns; attached to the pronominal base i- (< i, invariable demonstrative pronoun), they also served to form anaphoric and personal pronouns. Indeed, anaphoric pronouns i-r, i-p (respectively third-person singular and plural), i (me class), and i-n (-n class), as well as personal pronouns, both animate u, nuku (first-person singular and plural), nu, num (second-person singular and plural), and i-r, ap (third-person singular and plural) and inanimate i (me class) and i-n (n class), included forms with nominal suffixes. There was an accusative form for personal pronouns, marked with nominal suffix -n. Personal pronouns (archaizing or dialectal forms) added to words were used as possessive adjectives.

Verbal suffixes. Personal suffixes of the verbal conjugation, two participial suffixes, and animate third-person singular and plural suffixes could be added to the verbal base. The infinitive could be derived from the verbal base without suffix. The personal suffixes of the verbal conjugation expressed the perfective aspect and indicated person and number (first-person singular -h, second-person singular -t, third-person singular ; first-person plural -hu, second-person plural -ht, third-person plural -hš).

The verbal bases were used to produce the participial form with suffix -n (passive or reflexive, imperfective or durative aspect) and the participial form with suffix -k (passive, perfective aspect). To express the subject or agent the two participles and the agent noun received animate nominal suffixes (e.g., hutta-n “(is/was being/will be) done,” hutta-n-r “he on whom the process hutta-n depends” = “he is/was/will be doing”; hutta-k “(it was) done,” hutta-k-r “he on whom the process hutta-k depended” = he did/has done”). The verbal base was also used to produce agent nouns with nominal third-person suffixes -r and -p (active, interminate aspect; e.g., hutta-r/p “he is/they are a doer/doers”).

The optative particle -ni followed the perfective forms, whereas the prohibitive particle ani/u was followed by the imperfective forms.


The syntax of written Elamite clearly contained elements from various dialects and was therefore unstable. Furthermore, in the course of time sentence structure shifted from being based on nouns and the use of anaphoric pronouns with animate classifiers indicating person to a system centered on verbs.

Generally speaking, the sentence consisted of a noun followed by various complements (qualifiers, determiners). The verb was always placed at the end of a clause, and in simple statements the most common word order was subject-object-verb/agent-patient-verb. The relations among the elements of nominal chains were indicated by nominal suffixes and word order. For example, when a noun qualified a preceding noun or pronoun (cf. English apposition) its formative suffix, if present, agreed with the nominal class of the word it qualified (e.g., u sunki-k “I king(-k),” the animate first-person singular suffix -k showing that “king” refers to “I”; nap pahi-r “god protector(-r),” the animate third-person singular suffix -r showing that “protector” refers to an animate third-person singular noun; sunki-r peti-r ak tari-r “king(-r) enemy(-r) and/or ally(-r)” = “king, enemy, and/or ally”).

When a noun or pronoun (whether or not formed with a suffix) was the determining complement of a preceding noun, it was marked with the suffix of the nominal class of the word it completed (e.g., pahi-r sunki-p-r “protector(-r) king(-p)-r” = “protector of the kings,” the first -r showing that the word is animate third-person singular, the suffix -p indicating the animate third-person plural, the second -r showing that the second word is dependent upon pahi-r; takki-me u-me “life(-me) I-me” = “life of me” or “my life,” the second -me showing that “I” is dependent upon takki-me; igi-šutu u-p “brother-sister I-p” = “brothers and sisters of me” or “my brothers and sisters,” the third-person plural suffix -p indicating a relationship between “I” and “brothers and sisters”; cf. English genitive).

In more complex nominal chains a determining noun could itself be determined by its own determining noun (e.g., takki-me igi-šutu u-p-me “life(-me) my brothers-and-sisters-me” = “life of my brothers and sisters,” the suffix -me forming the abstract takki-me and the second -me indicating that “my brothers and sisters” is dependent upon takki-me).

Words could be subordinated to one another by means of nominal constructions, simple or complex. In these constructions the anticipated subordinating word was in turn represented by an anaphoric pronoun, followed by a determining word indicating the nature of the subordination (manner, place, or time). The anaphoric pronoun and the determining word formed a syntactic unit. In the complex construction a personal pronoun was subordinated to the determined anaphoric pronoun. This personal pronoun was marked with the suffix of the nominal class of the subordinating word. For example, in the sentence peti-r i-r pat-r u-r “enemy(-r) he(-r) below-r I-r” = “(should thou place) the enemy under me” the suffix -r of the anaphoric pronoun shows that the anticipated subordinating word is animate third-person singular, and the last -r shows that “I” is dependent upon the determined anaphoric pronoun “he.” In the simple construction the word immediately preceding the syntactic unit was subordinated to the determined anaphoric pronoun referring to it (e.g., kiti-ŋzalmu u-me i-n ukku-n “the divine rule(-n) … statue I-me it(-n) above-n” = “the divine rule … (should be placed) above my statue,” the suffix -n of the anaphoric pronoun indicating that the subordinating word is inanimate of the n class. The same syntactic unit could complete an anticipated word (e.g., suhmutui giri-me tah “stele (me class) it gratitude-me I have placed” = “I have placed … the stele in gratitude,” the anaphoric prounoun i of the syntactic unit indicating that the completed word is inanimate of the me class. As the focus of Elamite syntax gradually shifted from nouns to verbs, anaphoric pronouns were used less to introduce syntactic units, and in the late period they came to be understood as simple postpositions or adverbs, used with or without nominal suffixes.

A verb could be used to modify a following verb, in which case the personal suffix could be omitted from the first verb (e.g., pepši(-h) hutta-h “(I-)renewed I-made” = “I did again”).

A complex statement could contain several clauses, either coordinated or linked by subordination. Subordination was hierarchical and obeyed the word order sub-subordinate, subordinate, main clause. The mark of subordination (last in a chain of suffixes) could be a nominal suffix, the suffix -a, or both for relative clauses, whether or not introduced by relative pronouns (animate akka, inanimate appa), and the suffix -a (as, whereas, because) after other subordinate clauses. It was attached to what would be the verb of the subordinate clause in English. In addition, the suffix -a was attached to suffixes of determining words placed at the ends of subordinated noun clauses.

When the main verb at the end of a clause was preceded by pronouns referring to nouns in an earlier clause the syntactic functions of those nouns were reflected in the order of the pronouns: beneficiary/indirect object, subject/agent, direct object/patient (usually followed by -n).

Direct and indirect speech were identified by the verb ma- placed at the end of the quotation.

Examples. siyan d.Upurkupak-me “temple (siyan, me class) GOD.Upurkupak-me" sunki-p uri-p u-p “kings(-p) predecessors(-p) I-pŠušun in-me (> im-me) “(in) Susa not(-me) (i.e., siyan, me class) kuši-hš-me-a (> ma) “they-have-built-me-that” u kuši-h “I I-have-built” = “The temple of Upurkupak that the kings my predecessors had not built in Susa I have built.”

d.Nahhunte “GOD.Nahhunte" kulla-n-k-a “(I am) implor-ing(-k)-which” kula-i (> kula-a) “prayer-this” u i-r (> ur) “for me he” tumpa-n-r-a “fulfill-ing(-r)-he-whereas” ak “and” turu-n-k-a "(I am) express-ing(-r)-which” hutta-n-r-a “realiz-ing(-r)-whereas” siyan-kuk siyan "siyankuk temple” i-me “he-me” (i.e., siyan, me class) upat hussi-p-me “bricks (upat, plural animate) colored(-p)-me" kuših “I-have-built” = “For Nahhunte, (who) fulfills for me this prayer that I implore and (who) realizes (it) as I express (it), I have built his siyan-kuk temple of colored bricks.”

tetin-i “column-this” hiš u-me-ni ak hiš “name of-mine and the names” appa aha tallu-h-a “that there I-have-written-that” akka melka-n-r-a “who(ever) (will be) destroy-ing(-r)-who” ak suku-n-r-a “and will be suppress-ing(-r)-who” ak hiš duhi-e “and name own-his” aha-r tatallu-n-r-a “there(-r) reinscrib-ing(-r)-who” d.Inšušinak i-r si-r-a "GOD.Inšušinak he(-r) before-r-who” ani uzzu-n “let-not be walk-ing” = “Whoever would destroy and suppress name of mine and names that I have written on this column and (then) would reinscribe his own name, may he not walk before Inšušinak!”

kukunnum “the kukunnumsunki-p uri-p u-p “kings(-p) predecessors(-p) I(-p)” in-me (> im-me) “not(-me)” (i.e., kukunnum, me class) kuši-hš-a “they-have-built-which” u kuši-h “I I-have-built” d.Napiriša ak d.Inšušinak "GOD.Napiriša and GOD.Inšušinak” siyan-kuk-p-a "siyan-kuk-p-who" duni-h “I-have-attributed” hutta-k hali-k u-me “done labored I-me” (i-n) li-n-a "(it) present-n-that” ap u i-n (> apun) “to-them by-me it” tela-k-ni “address-ed-let-be” = “The kukunnum that the kings my predecessors had not built I built. I have attributed (it) to Napiriša and Inšušinak, (who are) of siyan-kuk. May my ‘work’ be addressed as a present by me to them!”

m.Kutir-d.Nahhunte “PERSON.Kutir-GOD.-Nahhunte” zalmu erintum-ia “statues baked-bricks-of” huhta-š “he-has-built” ak siyan d.Inšušinak-me “and the temple (siyan, me class) GOD.Inšušinak-meaha-n “there(-n) (n class = neutral) kuši-n-k “(I am) build-ing(-k) ma-r “(is) speaker” ak in-me (> im-me) “and not(-me) (i.e., siyan, me class) kuši-š “he-has-built” = “Kutir-Nahhunte has made statues of baked bricks and says ‘I have built the temple of Inšušinak there,’ and he has not built (it).”



F. Bork, “Die elamische Klam-mer,” Archiv für Orientforschung 9, 1934, pp. 292-300.

I. M. Diakonoff, “Die elamische Sprache,” Die Sprachen des alten Vorderasien, Moscow, 1967, pp. 85-112.

J. Friedrich, “Altpersisches und Elamisches,” Orientalia, N.S. 17, 1949, pp. 1-29.

G. B. Gragg, “Less-Understood Languages of Ancient Western Asia,” in J. Sasson, ed., Civilizations of the Ancient Near East IV, New York, 1995, pp. 2161-79.

F. Grillot-Susini, Éléments de grammaire élamite, Paris, 1987.

Idem, “Une nouvelle approche de la morphologie élamite. Racines, bases et familles de mots,” JA 282, 1994, pp. 1-18.

R. T. Hallock, Persepolis Fortification Tablets, Oriental Institute Publications 92, Chicago, 1969.

W. Hinz, “Problems of Linear Elamite,” JRAS, 1975, pp. 106-15.

Idem and H. Koch, Elamisches Wörter-buch, 2 vols., AMI N.S. 17, 1987.

W. F. König, Die elamischen Königsinschriften, Archiv für Orient-forschung 16, 1965.

R. Labat, “Structure de la langue élamite,” Conférences de l’Institut de Linguistique de l’Université de Paris 10, 1957, pp. 23-42.

D. W. McAlpin, Proto-Elamite-Dravidian. The Evidence and Its Implications, Philadelphia, 1981.

P. Meriggi, La scrittura proto-elamica, 3 vols., Rome, 1971-74.

H. H. Paper, The Phonology and Morphology of Royal Achaemenid Elamite, Ann Arbor, Mich., 1955.

E. Reiner, The Elamite Language, HO I2/1-2, pp. 54-118.

Idem, “Elamite,” in International Encyclopaedia of Linguistics I, New York, 1992, pp. 406-09.


Originally Published: December 15, 1998

Last Updated: December 13, 2011

This article is available in print.
Vol. VIII, Fasc. 3, pp. 332-335

Cite this entry:

FRANÇOISE GRILLOT-SUSINI, “ELAM iii. Proto-Elamite,” Encyclopaedia Iranica,  VIII/3, pp. 332-335, available online at http://www.iranicaonline.org/articles/elam-v (accessed on 30 December 2012).