On the Sounds Rendered by the s-, š-

and ´/z-Series in Elamite*

Jan Tavernier
Louvain-la-Neuve

1. Introduction
The study of Elamite phonology is a difficult task. Four reasons make it
extremely hard to get a hold on how the Elamites pronounced their
language (Reiner 1969:71; Grillot-Susini–Roche 1987:10; Khačikyan
1995:105; Krebernik 2005:161). Firstly, Elamite is written by means of
the Sumero-Akkadian cuneiform writing system, a script which was not
devised for expressing Elamite. As a result of this, the Sumero-Akkadi-
an cuneiform signs cannot render in a clear way all Elamite phonemes,
and as a consequence of this, some signs had to render more than one
phoneme. Secondly, given the isolated status of Elamite, there is no
comparative linguistic material for the study of Elamite phonology. Al-
though McAlpin (1981) reconstructed a Proto-Elamo-Dravidian sub-
strate language and thus proposed a connection between Elamite and
the Dravidian languages, the link between them is still too weak to allow
far-going conclusions on Elamite phonology. Thirdly, one should keep
in mind the possibility of Elamite dialects, which may touch on phono-
logical issues. Finally the phonological system also had its own dia-
chronic development.1

* Abbreviations follow the system used in Northern Akkad Project Reports 8

(1993):49–77. Other abbreviations are: AE = Achaemenid Elamite; AE (AHam) =
Achaemenid Elamite—texts dated to Atta-¶amiti-Inšušinak; MBHT = Middle
Babylonian texts from Haft Tepe; MBSu = Middle Babylonian texts from Susa; ME
= Middle Elamite; ME TTM = Middle Elamite texts from Tal-i Malyān; NA =
Neo-Assyrian; NB = Neo-Babylonian; NE = Neo-Elamite; OAkk. = Old Akkadian;
OBMa = Old Babylonian texts from Mari; OBSu = Old Babylonian texts from
Susa; OBTM = Old Babylonian texts from Tal-i Malyān; OE = Old Elamite.
1
Basically there are five diachronic stages in the development of the Elamite
language. The first one is Old Elamite (ca. 2300–1500 BC), followed by Middle
Elamite (ca. 1500–1000 BC), Early Neo-Elamite (NE I; ca. 1000–650 BC), Late
Neo-Elamite (NE II; ca. 650–550 BC) and Achaemenid Elamite (ca. 550–330 BC).
This framework is relevant not only for Elamite phonology, but also for mor-
1060 Limited-corpus Languages of the Ancient Near East

Possibly because of these reasons, the recently published Elamite

grammars (Grillot-Susini–Roche 1987; Khačikyan 1998; Stolper
2004; Krebernik 2005) devote relatively little attention to Elamite
phonology.
Nevertheless, these difficulties should not deter the scholar from try-
ing to solve the mysteries of Elamite phonology given that there are three
types of sources which enable him / her to conduct such a study.
(1) Transcriptions of non-Elamite words in Elamite texts (Iranian, Akkadian
and West-Semitic proper names and loanwords).
(2) Transcriptions of Elamite proper names and words in non-Elamite texts.
(3) Spelling variants within Elamite texts themselves.

These sources can easily be divided into two groups: internal data
have no connection whatsoever with a non-Elamite language (No. 3),
whereas external data (nos. 1–2) do have a connection with text attesta-
tions in languages other than Elamite.
This article focuses on a particular phonological issue, the sounds
rendered by the s-, š- and ´/z-series in Elamite. It is not certain how
many sibilants and/or affricates Elamite possessed (Foy 1898:129;
Reiner 1969:72–73) and therefore several scholars have published vari-
ous ideas on this topic. With Old Persian phonology in mind, Wester-
gaard (1845:343, 348–349, 355–356) believed in the existence of El. /č/,
/s/, /š/ and /z/. Holtzmann (1851:147, 154, 168–169; also Weissbach
1890:47 and Grillot-Susini–Roche 1987:10) accepted the existence of /s/,
/š/ and /z/. Four years later Norris (1855:39, 44 and 50) mentioned /č/,
/ç/, /s/ and /θ/, whereas according to Mordtmann (1862:31) Elamite pos-
sessed /ç/, /s/, /θ/ and /z/. Other scholars assume the existence of only /s/,
/š/ and /č/ (Hüsing 1898:15; Bork 1910:569–571 and 1925:74; Weidner
1917:32; Paper 1955:25–29; Reiner 1969:72; McAlpin 1981:65, 90–91).
In his study on Proto-Elamo-Dravidian (PED) McAlpin includes a his-
torical phonology of these three phonemes. PED /*š/ remained /š/ in
Elamite (whereas it disappeared in Dravidian). PED /*c/ became Proto-
Elamite /*c/, but subsequently had a more complicated development, as
the following table shows:

phology and syntax. Only with regard to the writing system it should be slightly
modified, in the sense that the Middle Elamite period is divided into two subdivi-
sions: Classical Middle Elamite (the royal inscriptions) and Late Middle Elamite
(the administrative tablets from Tall-i Malyān).
 J. Tavernier, On the Sounds Rendered by the s-, š-, ´/z-Series… 1061

Table 1: Development of PED and Proto-Elamite /*c/ (McAlpin 1981)

Phonetic context Middle Elamite Achaemenid Elamite
Before /a/ /*c / > s /*c / > s
/*c / > z (= /č /) (rarely)
Before /e/, /i/ /*c / > s /*c / > z (= /č /)
/*c / > z (= /č /) (rarely)
Before /u/ /*c / > z (= /č /) Not mentioned2
/*c / > s (rarely)

In reality, however, these rules do not withstand critical research. For

example, many times /č / precedes /a/ and this /č/ cannot have other ori-
gins than from PED /*c / in McAlpin’s view.
Khačikyan (1995:105–107; 1998:7–8) has two alveolar fricatives (/s/ and
/š/) and two affricates (/c / and /č /). Stolper (2004:71) postulates at least three
fricatives, transcribed as s, š and z, and an affricate /č /. Labat (1951:28) de-
nies the existence of /θ/ and has doubts on /č /, whereas Steve (1992:14) cau-
tiously mentions /ž/ and /ğ/. Recently Krebernik (2005:162) has mentioned
/s/, /š/, /z/ with their respectively geminated variants. Finally Henkelman
(2008:278, n. 635) assumes the existence of an affricated dental /ts/. All to-
gether eight fricatives (/ç /, /s/, /ss/, /š/, /šš/, /θ/, /z/ and /zz/) and five affricates
(/c /, /č /, /ğ/ and /ts/ and /ž/) have been mentioned in previous publications.
Some phonemes are accepted by almost all scholars: /č/, /s/, /š/ and, to
a lesser extent, /z/, whereas others (/c / /ç /, /ğ/, /ts/, /θ/ and /ž/) are only
mentioned in the works of one or two authors.
Three types of signs may render sibilants or affricates: the s-signs,
the š-signs and the ´/z-signs. A problem that emerges when studying
the sibilants is that the VC-signs (e. g. ÁŠ) may render various sibilants
(áš = ás; is = iz, etc.). Also CV-signs, which generally have a clearer
distinction, may be confusing, e. g. zu may be read as sú. Another
problem is that the cuneiform writing system is not apt to express af-
fricates. Finally it should be noted that CVC-signs will not be used in
this discussion because of their ambiguity as to the precise character of
the consonants they express.

2. Non-Elamite expressions in Elamite texts

The first step in the study of Elamite sibilants and affricates is an analysis
of Elamite renderings of non-Elamite expressions, especially of Akkadian

2
As the sign ZU had disappeared by the Achaemenid period, a clear distinc-
tion between /cu/ and su cannot be made.
1062 Limited-corpus Languages of the Ancient Near East

and Old Iranian ones (cf. Foy 1898:128–129; Cameron 1948:40–45 and
1954–1959:471; Paper 1955:29–30;3 Khačikyan 1995:106–107 and 1998:
7–8). There are not many Akkadian expressions rendered in Elamite, in
contrast to the numerous Iranian proper names and loanwords appear-
ing in Achaemenid Elamite texts. Nevertheless the Akkadian expressions
are important, since they are, contrary to the Iranian proper names and
loanwords, not limited to the Neo-Elamite and Achaemenid period. Al-
though CVC-signs are not valuable for this study, one example of an Ak-
kadian sibilant (ziqquratu) rendered by CVC-signs, is here included, since
it is the only attestation of Akkadian /z/ in Elamite.
Table 2: Akkadian and Aramaic names and words in Elamite
Akkadian Period Elamite
ÚAL
Attar-sūrī- (Aramaic), PN AE At-tur-r[u-i]š-šu-ri-iš
ÚAL
Bēlšunu, PN AE Be-ul-šu-un
AŠ
ešertu ‘chapel, shrine’ ME i-ši-ir-tu
kiššu ‘bundle’ (pl. kišati) ME ki-ša-a-ti
d
Išme-karāb, a theonym ME Iš-ni-ka4-ra-ab
Man-ištūšu, a royal name4 ME Ma-an-iš-du-uz-zu
misarru ‘(metal) band’ ME mi-za-ru-um
mīšaru ‘justice’ ME mi-ša-ri
nikassu ‘account’ AE nu-ik-kás-su-um-me
nisannu month name AE nu-šá-an
paspasu ‘duck’ ME ba-as/z-ba-as/z
AE ba-is/z.KIMIN
pašīšu anointing priest NE I–II ba-ši-šu
qištu ‘wood, forest’ ME ki-iš-tu4-um
ÚAL
Sîn-qatēni, PN AE Ši-in-ka4-tan-na
´almu ‘statue’ ME sa-al-mu-um (1 time)
ME za-al-mu (14 times)
NE I za-al-mi (1 time)
NE I za-al-mu (11 times)
NE II za-al-mi (2 times)
NE II za-al-mu (14 times)
AE (AHam)5 za-al-mu (3 times)

3
Paper made the mistake of including Babylonian renderings of Old Persian
names and loanwords in his table of sibilant correspondence-sets. These names
and loanwords, however, do not yield reliable information on Elamite phonology,
but rather on Old Iranian phonology.
4
The later spelling of this name is Ma-ni-iš-ti-iš-šu (Steinkeller 1987–1990:334).
5
The inscription on a stela of Atta-¶amiti-Inšušinak (EKI 86–89) was formerly
dated to the second half of the 7th century BC (Vallat 1996:391) but recently the
idea was expressed that this king could very well be identical with the rebel
 J. Tavernier, On the Sounds Rendered by the s-, š-, ´/z-Series… 1063

Akkadian Period Elamite

´almu ‘statue’ (contd.) AE za-ul-man (3 times)
´imittu ‘bundle of silver scrap’ ME si-mi-it-tum4, si-mi-it-tu8-um
´īt šamši ‘sunrise’ ME si-it-ša-am-ši
ÚAL
Šamaš-gerra, PN AE Ša-ma-iš-ki-la/ra
šaqû ‘high’ ME ša-qu-tu4
ziqquratu ‘ziggurat’ ME sig-ra-tu4-me (1 time)
ME zak-ra-tu4-me (3 times)
NE II zik-kur-ti-um (1 time)

As can be seen in this overview Akkadian (and Aramaic) /s/ is rendered

in Elamite by the three types of signs indicating sibilants: s (nikassu), š (At-
tar-sūrī-, nisannu, Sîn-qatēni) and ´/z (misarru), albeit more frequently by š.
Akkadian /´/ is rendered by s in the Middle Elamite texts from Tal-i
Malyān (´imittu and ´īt šamši) and by ´/z in other texts (´almu); only once is
it written with s (´almu). Elamite š-signs are the only ones used to render
Akkadian /š/ (Bēlšunu, ešertu, kiššu, Išme-karāb, mīšaru, pašīšu, qištu,
Šamaš-gerra, šaqû), except for the rendering of Man-ištūšu. The render-
ing of Akkadian /z/ by CVC-signs makes it difficult to draw any conclu-
sions about it.
Numerous Iranian proper names and words are transcribed in Neo-
and Achaemenid Elamite texts. Accordingly they are a very important
source for the study of late Elamite sibilants. The Elamo-Iranian transpo-
sitions may be divided into four categories: (1) directly transmitted
Iranica, (2) semi-directly transmitted Iranica, (3) non-Iranian proper
names and loanwords in Old Persian and (4) indirectly transmitted
Iranica. The first two categories are the basis for the Elamo-Iranian
transpositional system and are also most important for the study of Elam-

Aθamaita- mentioned in Darius’ Bīsītūn Inscription (Waters 2000:85; Tavernier

2004:22–29; Henkelman 2008:291). In a rather haughty tone Vallat has criticized
this idea, which he has consigned to the “poubelles de l’Histoire” (Vallat 2006:61).
Vallat bases his criticism on one sentence in an unpublished text from Atta-¶amiti-
Inšušinak. It goes as follows: “Comme Šutur-Na¶¶unte, fils de Úumban-immena,
manœuvre à Ayapir, j’ai mâté son armée” (Vallat 2006:59). According to Vallat this
sentence proves that Šutur-Na¶¶unte and Atta-¶amiti-Inšušinak must have been
contemporaries. Leaving aside the methodological error of citing unpublished sen-
tences exclusively in translation and without their context, a remark must still be
made: it is by no means sure that the Atta-¶amiti-Inšušinak cited by Vallat is the
same individual as the Atta-¶amiti-Inšušinak of the stela. As long as Vallat presents
his arguments in such a way, they cannot be convincing and therefore I see no rea-
son to change my point of view concerning this issue.
1064 Limited-corpus Languages of the Ancient Near East

ite phonology. In the first category, the names and words whose Iranian
original is attested in Old Persian are collected. As there are no Old Ira-
nian texts dating from the Neo-Elamite period, there are no Neo-Elamite
Iranica belonging to this category. The second group includes Neo-
Elamite and Achaemenid Elamite Iranica that differ only slightly from an
attested Old Persian name or word, for example because a proper name
or loanword belongs to another Iranian dialect or because it is a younger
equivalent of an Old Persian name or word. The third category contains
non-Iranian names and loanwords that were transmitted to Elamite
through Old Persian. The fourth category is composed of Iranica whose
Iranian original is not preserved, but which can be reconstructed accord-
ing to the transpositional system constructed on the basis of the expres-
sions of categories one and two (Tavernier 2007:3–4).
Old Iranian /č/ is mostly rendered by ´/z-signs. More seldom it is ex-
pressed by s- or t-signs. The sequence /ču/ had to be rendered by SU,
because ZU had disappeared from the syllabary and was no longer
available for the scribes. Old Iranian /ç/ was always written by means of
š-signs. There is only one exception to this: *dauçaka- ‘sacrifice,’ appears
in Elamite as da-u-si-ka4 (Hinz 1973:108–109 and 1975:91; Koch
1977:127; Henkelman 2008:212–213; Tavernier 2007:462, No. 4.4.22.8).
Old Iranian /j/ is almost exclusively expressed by El. ´/z-signs. /Ju/ is ex-
pressed by SU because of the lack of ZU and the name *Jīča- is once
spelled Si-iz-za (Tavernier 2007:220, No. 4.2.895). Old Iranian /s/ and
/š/ are rendered by š-signs, with two exceptions: *patisēka-, a kind of
payment, is twice written bat-ti-zé-kaš (Hallock 1969:135, 676; Tavern-
ier 2007:444, No. 4.4.10.15) and *Pēšiyāhvādiya- appears mostly as Be-
zí-ia-ma-ti-ia (Tavernier 2007:74, No. 2.3.36). Old Iranian /z/ is mostly
represented by El. ´/z-signs, but El. š-signs could also be used to render
this phoneme. Because of the already mentioned problem with ZU, the
sequence /zu/ was rendered by SU, which could also be used in order to
write /za/ before /w/. Finally Old Iranian /ž/ was rendered by El. s-, š- or
´/z-signs.
The Elamite s- and š-series were, next to the d/t-series, also used for
the transcription of Iranian /θ/.

3. Elamite expressions in Akkadian texts

The second step is to study Elamite expressions in Akkadian texts. It
should be noted that reconstructed Elamite forms are not included in the
general conclusions derived from this table.
 J. Tavernier, On the Sounds Rendered by the s-, š-, ´/z-Series… 1065

Table 3: Elamite names and words in Akkadian (see also Krebernik 2006:84–90)
No. Elam. spelling Elam. period Akk. spelling Akk. period
1 ak-sir6 ME, NE II ak-si-ir7 OBSu
ak-sír7 ME ak-še-er OB
ak-še-ir ME, AE ak-sìr7 MB
ak-ši-ri7 NE II
ak-ši-ra AE
2 Az-za-za NE II A-za-za OAkk.
BE/ÚAL 7
3 Úal-lu-iš NE II, AE el-ta-aš OBSu
mÚa-lu-si
NA
mÚa-lu-su
NA
mÚal-lu-si
NA
mÚal-lu-ši
NA
mÚal-lu-šú
NB
4 ¶al-taš7 NE II ¶al-taš7 OBSu
ÚAL
Úal-da-iš AE al-da-a-še NA
al-da-še NA
al-da-si NA
al-da-su NA
¶al-da-a-šú NA
al-da-šú NB
ìl-da-šú NB
5 ¶a-né-eš7 OE ¶a-né-eš Ur III, OB
¶a-ni-iš ME, NE II ¶a-ni-iš7 OBSu
6 *hašša ¶a-aš-ša OBSu
7 *hupirririša ¶u-pír-ri-ri-ša OBSu
¶u-ši- ¶u-si- 7
8 ° NE II, AE ° OBSu
9 *hutliš ¶u-ut-li-iš OBSu
10 dKi-ri-ri-ša ME, NE II7 Ki-ri-ri-ša OBSu, MBSu
d
Ki-ri-ri-šá AE
11 dKir-wa-si-ir OE Ki-ma¶-si-ir OAkk.
d d
Kir-ma-sir ME Ki-ir-me-si-ir MBHT
d
Ki-ir-wa-si-ir MBHT
d
Ki-ir-<<sa>> NA
-ma-as
12 *Kutuzuluš Ku-du-šu-lu-uš OBMa
Ku-du-zu-lu-uš OBSu
13 *kuššuku Ku-uš-šu-ki OBSu
14 *Kušum Gu-šum OBSu
Gu-ú-šum OBSu

6
As an element in anthroponyms and toponyms.
7
Attested as an abbreviated form Kiriša (dKi-ri-iš-ša) in EKI 76:34.
1066 Limited-corpus Languages of the Ancient Near East

No. Elam. spelling Elam. period Akk. spelling Akk. period

15 *kutuhtaš Ku-du-u¶-ta-aš OBSu
16 *liriša li-ri-šà OBSu
17 *lurs/šu lu-ur-si OBSu
lu-ur-ši OBSu
lu-ur-šu OBSu
d d
18 Ma-zi-at OE Ma-an-zi-at Ur III
d d
Ma-an-za-at ME Ma-an-zi-it Ur III
d d
Ma-za-at ME Ma-an-za-at OA, OB, MB,
MBHT, NA,
NB
d d
19 Na-pi-ri-ša ME Na-pi-ri-ša OBSu
d
Na-ap-ri-si NA
20 pa-¶aš8 ME Pa-¶a-aš OBSu,
MBHT
21 *šarnup šá-ár-nu-up-pu NB
22 dŠá-iz-zí NE II d
Šà-zi OBSu
23 dŠi-da-nu NE II d
Šu-da-a-nu NA
d d
Šu-d[a-nu] NE II Šu-da-nu NA
24 si-il-¶a-ak OE, ME si-il-¶a-ak7 Ur III Susa,
OBSu, OBTM
ši-il-¶a-° ME, NE II še-el-¶a-ak Ur III
šil-¶a-° ME, NE II, ši-il-¶a Ur III
AE ši-il-¶a-° OB, OBSu9
25 Si-[ir-uk-du-u¶] OE Si-ir-uk-du-u¶ OBSu
Ši-ir-uk-du-¶ ME Si-ir-uk-tu¶ OBSu
m
26 Ši-mu-mu AE Si-mu-mu OBSu
d d
27 Si-mu-ut OE, ME Si-mu-ut OB, OBSu,
MB, MBHT
d d
Ši-mut ME, NE I–II Ši-mu-ut OB
d d
Ši-mu-ut ME, AE Ši-mut MB
d
Šu-mu-du NA
28 si-ia-an ME, NE I–II zi-a-na-am OBSu
si-a-na NE II zi-a-nam OBSu
zi-ia-an NE II, AE zi-a-ni OBSu
zi-ia-an OBSu
zi-ia-na-[am] OBTM

8
Although this is written with a CVC-sign, it is certain that š is meant, since
the verbal endings of the third person singular of conjugation I are exclusively
written with š in Elamite.
9
Dossin (1962:157) reads Ši-il-¶a-¶a on an inscribed bronze axe from Luris-
tān. Nevertheless the sign traces on his plate rather point to Si-il-¶a-¶a than to a
reading Ši-il-¶a-¶a.
 J. Tavernier, On the Sounds Rendered by the s-, š-, ´/z-Series… 1067

No. Elam. spelling Elam. period Akk. spelling Akk. period

zi-ia-an (contd.) zi-ia-nam OBSu
zi-i-a-ni MBSu
29 su-u¶-mu-tú ME zu-mi-tum OBSu
zu-u¶-mu-tú ME
m
30 Šu-tur-dUTU NE I m
Šu-túr-dNa-¶u- NA
un-de
31 taš ME, NE II, ta-aš OBSu
AE
da-iš AE da-aš MBHT
32 du-ni-iš ME tu-ni-iš7 OB
du-nu-iš NE II, AE
33 Zí-we-pa-la-ar-¶u-u¶- OE Si-we-pa-la-a[r- OBSu
pa-ak ¶u-u¶-pa-ak]
Si-me-pá-la-ar-¶u-u¶- ME Še-ep-la-ar-pa-ak OBMa
pa-ak
Si-me-pá-la-ar-¶u-u¶- ME Še-ep-[l]a-ra-pa- OBMa
pá-ak ak
Si-me-pá-la-ar-¶u-up- ME Zí-we-pa-la-ar- OBSu
pá-ak ¶u-uh-pa-ak
Si-me-pá-la-ar-¶u-up- ME Zí-we-pa-la-ar- OB Babylon
pak ¶u-ú¶-p[a-ak]

34 Zí-it OE Si-it OB
Ši-ti UrIII or OB
35 zu-kir OE su-kir OAkk. Susa
su-gìr ME, NE I zu-uk-ki-ir OBSu
su-un-gìr ME
su-un-ki-ir ME
zu-uk-ki-ir ME
zu-un-ki-ir ME

The following table (Table 4) summarizes the various transpositions

between Akkadian and Elamite. In a strict sense, only transpositions be-
longing to the same chronological period are listed.
Table 4
Period Elamite Akk. Akk. Examples
(Elam) (Mes.)
OE—Old Akk. s s Kirwasir
´/z s zunkir
OE—Ur III š š haneš
´/z ´/z Manzat
OE—Old Bab. s s s Simut, Siruktu¶
š š š haneš
1068 Limited-corpus Languages of the Ancient Near East

Period Elamite Akk. Akk. Examples

(Elam) (Mes.)
OE—Old Bab. ´/z s Ziwepalarhuhpak
(contd.) ´/z s Zit
´/z š Ziwepalarhuhpak
´/z ´/z ´/z Ziwepalarhuhpak, zunkir
ME—Middle s š Simut
Bab. s ´/z siyan
š s Simut
š s š Kiririša
š s š Simut
´/z ´/z ´/z Manzat10
NE—Neo-Ass. š s Halluš
š š Halluš, Šitanu, Šutur-
Nahhunte
AE—Neo-Ass. š š haltaš

Elamite š-signs are mostly rendered by Akk. š-signs as well. The (at
first sight) exceptional Akkadian writing Simut for Šimut is explained by
the occurrence of Old and Middle Elamite Simut. Neo-Assyrian has a real
variation of s and š in its renderings of Elamite š, but this is probably due
to Assyrian phonology itself, where an s/š-variation exists (Parpola 1974:
1–2; Fales 1986:61–63; GAG § 37). If the strict diachronic approach is set
aside, i. e. if transpositions of different periods are included, it becomes
apparent that the transposition El. š—Akk. š is maintained, with one ex-
ception: El. akšer (ME—AE) is once rendered by ak-si-ir (in a personal
name) in an Old Babylonian text from Susa.
Elamite s-signs are rendered by signs belonging to the s- and ´/z-series. This
pattern remains unchanged, when including non-diachronic transpositions.
Finally Elamite ´/z-signs are mostly written by means of ´/z-signs (Old
Elamite—Ur III and Old Babylonian, Middle Elamite—Middle Babylo-
nian; see, however, below, No. 1). In two cases they are rendered by
s-signs (Old Elamite—Old Akkadian, Old Elamite—Old Babylonian), but
each of these examples is attested only once. The first example may be
due to Old Akkadian orthography while the usual spelling of Ziwepalar-
huhpak in the Old Babylonian texts from Susa is with z.
If the strict diachronic approach is set aside, one can see that Old
Elamite ´/z is exclusively rendered by Akkadian ´/z-signs, both in Mesopo-
tamia and in Elam.

10
Possibly the following nasal had influence on the real character of this sibi-
lant.
 J. Tavernier, On the Sounds Rendered by the s-, š-, ´/z-Series… 1069

Two more interesting aspects should be mentioned:

1) Some Elamite expressions showing an initial variation s/z are rendered in
Elamite Akkadian by equivalents with initial z. El. S/Ziwepalarhuhpak,
s/ziyan, s/zuhmutu and s/zunki appear in Akkadian as Ziwepalarhuhpak, ziyan,
zuhmutum and zukkir. In Mari this variation was expressed with š-signs:
S/Ziwepalarhuhpak became Šeplarpak.
2) Elamite words with an initial variation of s and š (cf. infra) are rendered in
Elamite Akkadian by s, whereas in Mesopotamian Akkadian š is the exclu-
sive way to render these lexemes (sil¶a-, Simut and Siruktu¶).11

The following tables incorporate all the data in a clear overview. The
italics in Table 7 indicate that these transpositions are not very frequent.
Table 5: Akkado-Elamite transpositions
Akkadian Elamite
Voiceless alveolar fricative (/s/) s (rare)
š
´/z (rare)
Emphatic alveolar fricative (/´/) s
´/z
Voiceless palato-alveolar fricative (/š/) š
Voiced alveolar fricative (/z/) z
Table 6: Irano-Elamite transpositions
Old Iranian Elamite
Voiceless palato-alveolar affricate (/č /) s (seldom)
t (rare)
´/z (frequent)
Voiceless alveolar fricative (/ç /) s (once)
š (frequently)
Voiced palato-alveolar affricate (/j/) s (rare)
´/z (frequent)
Voiceless alveolar affricate (/s/) š (frequent)
´/z (once)
Voiceless palato-alveolar fricative (/š/) š (frequent)
´/z (once)
Voiceless interdental fricative (/θ/) d/t
s (frequent)
š (rare)
Voiced alveolar fricative (/z/) š (rare)
´/z (frequent)

11
This study is based on material from ElW. Additional Old Babylonian ex-
amples are Ri-ib-Ši-mu-ut (BIN 10, 157:7), Si-el-¶a, Še-el-¶a, Šim-še-il-¶a (Whit-
ing 1987:29, n. 90) and dŠi-mu-ut-a-bi (OECT 15, 95:22).
1070 Limited-corpus Languages of the Ancient Near East

Old Iranian Elamite

Voiced palato-alveolar fricative (/ž/) s (once; SU for ZU)
š (three times)
´/z (twice)

Table 7: Elamo-Akkadian transpositions

Elamite El. Akkadian Mes. Akkadian
s voiceless alveolar voiceless alveolar fricative (/s/)
fricative (/s/)
s/š (ini- voiceless alveolar voiceless palato-alveolar
tial) fricative (/s/) fricative (/š/)
voiceless palato-alveolar
fricative (/š/)
s/z (ini- voiced alveolar voiced alveolar fricative (/z/)
tial) fricative (/z/) voiceless palato-alveolar fricative (/š/)
š voiceless alveolar fricative voiceless alveolar fricative (/s/)13
12
(/s/)
voiceless palato-alveolar voiceless palato-alveolar fricative
fricative (/š/) (/š/)
´/z voiced alveolar voiced alveolar fricative (/z/)
fricative (/z/) voiceless alveolar fricative (/s/)
voiceless palato-alveolar fricative (/š/)

4. Spelling variants in Elamite

The third source for studying Elamite phonology is the corpus of spelling
variants within Elamite. First of all, it must be emphasized that some lex-
emes are spelled exclusively with only one grapheme, s, š or z.
(1) Only s: halsa- ‘to banish’; hasur ‘anointer’ (?); husa- ‘tree’; kars- ‘to paint,
dye’; sa- ‘to travel’; sari- ‘to destroy’; sati- ‘to ration, portion’; sira-
‘to weigh, hang.’
(2) Only š: 3rša- ‘great’; hušu- ‘to retaliate’; meša ‘later’; niški- ‘to guard’; šalhu-
‘to command’; šara- ‘to enforce’; šera- ‘to command’; šinni- ‘to
come’; šišna- ‘beautiful.’
(3) Only z: haz- ‘to be big’; izzi- ‘to go’; Manzat, a theonym (both in Akkadian
and Elamite texts); zak- ‘to spend; value’; zau̯m- ‘to labour’; zikka-
‘to heap up.’

It is unfortunately not possible to determine to what extent scribal

habits inspired the exclusiveness of the writings above. There is no dia-
chronic aspect involved here since most of the lexemes are attested from

12
Examples: akšir, ¶uši- and Šimumu.
13
In Neo-Assyrian only.
 J. Tavernier, On the Sounds Rendered by the s-, š-, ´/z-Series… 1071

the Old or Middle Elamite down to the Achaemenid period (halsa-, izzi-,
niški-, sari-, sira-, šalhu-, šera-, etc.).
More information comes from the spelling variations, of which there
are various types. The first one is s/š. Foy (1898:129; also Khačikyan
1995:106) mentions a shift from Old Elamite s to Middle and later Elam-
ite š, with reference to Old Elamite sutet > Middle El. šutme and to Old
and Middle El. Simut > Middle and Neo-Elamite Šimut. That this shift
should be considered a “rule,” as Khačikyan puts it, is, however, exag-
gerated. A closer look at the various examples yields a modified result.
There is certainty on the alternation s/š. Examples are (1) hipis (AE) ~
ipiš (NE II) ‘axe,’ (2) Insušinak (OE, ME, NE I) ~ Inšušinak (ME, NE I),
a theonym, (3) musika (AE) ~ mušika (AE) ‘it is counted,’ (4) pepsi- (ME) ~
pepši- (ME, NE II, AE) ‘to renew,’ (5) sil¶a- (ME, NE II14) ~ šilha- (NE II,
AE) ‘strong,’ (6) s[ip]ari (ME TTM) ~ šipari (AE), a month name, (7)
Simut (OE, ME) ~ Šimut (ME, NE I–II, AE), a theonym, (8) Si[ruktu¶]
(OE) ~ Širuktu¶ (ME), a personal name, (9) suhter (ME) ~ šuhter (ME;
only once) ‘altar,’ (10) sut- (OE) ~ šut- (ME) and šit- (AE) ‘night’15 and (11)
šasika (AE) ~ šašika (AE) ‘left over.’
In some cases the variation seems to have a diachronic nature. In oth-
ers, however, both variants occur in the same time period (e. g. nos. 3, 4
and 8). Remarkably, the diachronic examples all concern an initial varia-
tion: silha, sipari, Simut, Siruktuh, suhter (su-u¶-te-er, su-u¶-ter) and sutet
(su-dè-et). Four of these forms concern a sequence rendered by the sign
SI, which can easily be read ší (as Glassner–Herrero 1990:12 and Stolper
1982:60 transcribe).16 In that sense, this could be used to argue against
any shift or variation. Fortunately the fifth and sixth form differ from the
four other ones and corroborate the variation. It is, however, still possible
that the variation only applies to the sequence /su/.
The general pattern seems to be that Old Elamite and Middle Elamite
s (in the cases of silha-, sipari and Simut) became š during the Middle
Elamite period. This is, however, not a general rule and may thus point

14
In the adjective si-ul-¶i-te-ek-ra, attested three times in the inscriptions of
Hanne (ca. 625–600 BC). These attestations are rather exceptional, appearing at
a time when all other attestations of the lexeme šilha ‘strong,’ are written with š.
Perhaps a dialect aspect of Hanne’s inscriptions could be seen here.
15
S/šut- is the stem and appears in such forms as sutet, šutkume, šutme and
šitmana.
16
One should also keep in mind that i may have a palatalizing influence on
sibilants.
1072 Limited-corpus Languages of the Ancient Near East

to a specific phoneme which in the Middle Elamite period merged with a

phoneme rendered by š (at least in initial position).
It is also illustrative that sometimes the Akkadian scribes in OBSu
used s (at least in non-initial position) where later Elamite has š: aksir ~
akšir ‘guide, leader,’ husi- ~ huši- ‘to hurdle.’
Another type of variation is s/z. Examples are (1) hasi (NE II) ~ hazi
(NE II, AE) ‘hair (?),’ (2) kass- (ME) ~ kazz- (ME, NE II, AE) ‘to forge,
smith,’ (3) kusi- (ME) ~ kuši- (ME, NE, AE) ‘to build,’ (4) mass- (ME, NE I–
II) ~ mazz- (NE II, AE) ‘to cut off,’ (5) Pessitme (AE) ~ Pezzatme (AE), a
toponym, (6) siya- (ME, NE II, AE) ~ ziya- (NE II, AE) ‘to see,’ (7) siyan
(ME, NE II) ~ ziyan (NE II, AE) ‘temple,’ (8) sila (ME) ~ zila (OE, NE II,
AE) ‘statue,’ (9) Silili (OBSu) ~ Zilili (OBSu), a personal name, (10)
Simepalarhuhpak (ME) vs. Zimepalarhuhpak (OE), (11) sizzim (NE II) ~
zizzim (NE II), a type of shoes (?), (12) Sinini (NE II) ~ Zinini (AE), a per-
sonal name, (13) sip (ME) ~ zip (AE) ‘door, gate,’ (14) sukuka (AE) ~ zi-
kuka (AE) ‘demanded, requested,’ (15) Sunkiki (ME) ~ Zunkiki (ME), a
personal name.
Some examples of a variation š/t/z are attested: (1) huršubum (OBSu) ~
hurtebum (OBSu), a month name, (2) kit- ‘to pour out’ (with forms ki-iz ~
ki-ti-iš and ki-iz-za17 ~ ki-ti-iš-da) and (3) kuti- ‘to carry’ (with forms ku-iz ~
ku-ti-iš, ku-iz-da ~ ku-iz-iš-da, ku-iz-da-ti-iš-da, ku-iz-za-iš-da and ku-ti(-iš)-
šá ~ ku-(-iz)-za, ku-iz-za-iš). Tempt ‘lord’ (with forms semm, semt, šemm, šemt,
temm, temp, temt and tept) is a special case. The forms beginning with s and š
are attested in Mesopotamian sources dating from the Ur-III period to the
Old Babylonian period. In all probability these spellings are the result of
popular etymology, as Akkadian has a lexeme simtu ‘appropriate, proper
(symbol)’ (Zadok 1984:43), although Henkelman (2008:278, n. 635) con-
siders the various spellings as possible proof for an affricated dental /ts/ in
Elamite.
Another interesting case could be the divine name Zit, spelled dZí-it in
the Narām-Sîn Treaty (EKI 2 i 6). This name was connected by Vallat
(2000:1068; 2002–2003:530, 537; see Henkelman 2008:278, n. 635) with
Ši-ti (Iraq 38, 62:6), occurring in a Mesopotamian magical spell. More-
over, it is very likely that both spellings are related to Si-it (MHEO 2,
75:6), as Šit(i) and Sit appear in very similar contexts. This would, how-
ever, be the only example of such a variation.

17
According to ElW 472 this form was pronounced /kitsa/.
 J. Tavernier, On the Sounds Rendered by the s-, š-, ´/z-Series… 1073

Finally there is the interchangeability of š and z. Stève (1992:17; also Kha-

čikyan 1995; Stolper 2004:71 and Vallat 2004:136, n. 6) argues that š and z
often alternated in Elamite, but only three examples of such an alternation
are attested. The first one is the Middle Elamite divine name Hutekašan (in
f
Ú-tuk-dÚu-te-ka4-ša-an), which appears as Hutekazan on an Old Elamite
cylinder seal (Úu-te-kà-za-an; Vallat 2004:137). The second one is the per-
sonal name Kutuš/zuluš. Ku-du-šu-lu-uš is attested in an Old Babylonian
text from Mari (Durand 1986:119, n. 26), while Ku-du-zu-lu-uš occurs vari-
ous times in Old Babylonian texts from Susa. The third example is the place
name Anšan (OE, ME)/Anzan (ME, NE I–II, AE). It should be mentioned,
however, that toponyms are usually very difficult to explain and that the
name Anšan may not even be Elamite at all. Accordingly, the phonological
conclusions that can be drawn from the latter example are minimal, as a
consequence of which there are only two examples for this variation. This, of
course, does not support a variation š/z in Elamite phonology.18
It seems that each of these variations (s/š, s/z, s/š/z and š/t/z; perhaps š/z)
points to a specific phoneme. Diachronic aspects (e. g. phoneme shifts)
are not likely to have played a great role here, since all spellings occur in
all periods.

4. Analysis
What conclusions can be drawn from this data? According to Khačikyan
(1995:106; 1998:7; also Stolper 2004:71) the s-series must render an af-
fricate for three reasons: (1) the connection between s and OP /θ/, (2) the
variation s/š and (3) the variation s/š/t. This is not probable. As the ´/z-
series also indicates a fricative, the interchangeability of s and z must be
seen in this connection. Moreover, a variation s/š/t/z does not exist, since
the spellings with s all come from Mesopotamia and are probably the re-
sult of popular etymology. Moreover, there is one example of a variation
s/š/z, signs which again can all render fricatives.

18
The variations between Hutekašan and Hutekazan on the one hand and
Kutušuluš and Kutuzuluš on the other hand might be explained by assuming
that the former ones are palatalized pronunciations of the latter ones. In fact
there are both diachronic and synchronic (geographic) differences between the
various spellings. Diachronically, in later Elamite the palatalized equivalent of /z/,
i. e. /ž/, is mostly rendered by a š-sign. Synchronically, the Mari texts seem to have
a preference for š. The two names attested in these texts are Kutušuluš and Šep-
larpak, which appear with z (Kutuzuluš and Ziwepalarhuhpak) in the Old Baby-
lonian texts from Babylon and Susa.
1074 Limited-corpus Languages of the Ancient Near East

Accordingly, these reasons are rather indications for the s-series de-
noting a fricative. In addition to this the external data favour a connec-
tion between the s-series and fricatives: El. s = Old Ir. /θ/. It must be ac-
cepted that the s-series expresses a fricative. The character of this frica-
tive is, however, not known with certainty (see below).
It is also possible that the s-series can express an affricate. This suspi-
cion is raised by the variation between s- and z-signs (Stolper 2004:71).
This (and not the three reasons mentioned by Khačikyan) is the main in-
dication for this idea. Again the character of the affricate is far from cer-
tain. The spelling variation s/z shows that the affricates behind s and z are
not very different. The affricate closest to /č / is its non-palatalized equiva-
lent /c /. Khačikyan (1995:107; 1998:6) observes that the external evi-
dence (El. s = OP /θ/ and El. ´/z = OP /č / and /j /) suggests that the s-series
renders /c /, whereas the ´/z-series was used to express /č /.
With regard to the š-series the situation seems quite clear (Khačikyan
1995:106 and 1998:7). Both the external data (e. g. Akk. and Old Ir. /s/
and /š/ are mostly rendered by š) and the variation s/š indicate that this
series renders one or more fricatives. Mesopotamian orthography sup-
ports this, since the š-series is used for the notation of historical non-
affricates in Akkadian (Diakonoff 1988:37). In all likelihood the š-series
rendered both /s/ and /š/.
The ´/z-series only rendered one phoneme according to Khačikyan
(1995:106). Nevertheless the external data favours a connection of the ´/z-
series and affricates on the one hand and a connection of the ´/z-series
and the voiced alveolar fricative /z/ on the other hand. The latter is not
abnormal as this series was also used to render this phoneme in Ak-
kadian. It can be assumed that this series is the usual notation for an af-
fricate, but it is not sure which affricate is meant. Here the internal varia-
tion š/t/z may be important and may point the researcher in the direction
of a phoneme /č / (Paper 1955:29–30; Khačikyan 1995:106; 1998:7; Stol-
per 2004:71),19 although Labat (1951:28) is reluctant to accept the exis-
tence of such a phoneme in Elamite. The forms ku-iz and ku-ti-iš then
represent spoken /kuč /, while ku-ti-iš-da and ku-iz-da render /kučta/.

19
Stolper’s idea that spellings like ku-iz-iš-da and ku-iz-da-ti-iš-da served to
clarify a cluster /tšt/ does not pose a problem for this assumption since this cluster
contains an affricate. ElW 308 believe that this variation (s/š/t) indicates the inter-
dental fricative /θ/ (probably because s, š and t are the three sign series that can
render OP /θ/).
 J. Tavernier, On the Sounds Rendered by the s-, š-, ´/z-Series… 1075

It is likely that Elamite also had a phoneme or allophone /s’/, a lax

version of /s/, that was written by means of the ´/z-series. This would par-
allel the existence of lax and tense variants in the Elamite plosive series.
Proof for this phoneme is the frequent transposition of Akkadian and
Old Iranian /z/ by El. z. If Elamite had not known /s’/, then Ir. /z/ would
have been rendered more often by š-signs.
With historical Dravidian phonology in mind, one could postulate a
retroflex approximant /À/ for Elamite (Proto-Dravidian has a such a pho-
neme, see McAlpin 1981:24–25). This phoneme would be expressed by
the ´/z-series. Yet, the existence of such a phoneme remains very uncer-
tain, because of the weak ties between Dravidian and Elamite.
Based on an alleged variation š/z, Steve has postulated the existence of
palato-alveolar affricates /č / (voiceless) or /j / (voiced) on the one hand or
of a voiced fricative /ž/ on the other hand. It has been shown, however,
that this variation should not be automatically accepted within the Elam-
ite graphemic and phonological system.
The only question left open is the character of the fricative rendered
by the s-series. This question is very hard to answer. First of all, it is not
impossible that one phoneme could be rendered by more than one sign
(e. g. /t/ by d and t; cf. Reiner 1969:72–73). Accordingly, s may also repre-
sent /s/. Nevertheless, the variation s/š and the fact that in initial position
this variation became š during the middle Elamite period point to a sepa-
rate phoneme, which merged with /s/ in initial position. Possible candi-
dates are an interdental fricative /θ/ (based on the external evidence: El. s
= OP /θ/) or a lateral fricative /ś/.
To sum up, Elamite probably had six (or seven, if one accepts the ret-
roflex approximant) alveolar fricatives, palato-alveolar fricatives and af-
fricates: the alveolar fricatives /s/ and /s’/, a palato-alveolar fricative (/š/),
two affricates (/c / and /č /) and one yet unknown fricative (/θ/ or /ś/). The
table below connects these phonemes with graphemes.

Sign series Phoneme

s Interdental or lateral fricative /ś/ or /θ/
Alveolar affricate (?) /c /
š Alveolar fricative /s/
Palato-alveolar fricative /š/
´/z Alveolar fricative /s’/
Palato-alveolar affricate /č /
Retroflex approximant (?) /À/
0000
1076 Limited-corpus Languages of the Ancient Near East

Variations Phoneme
s/š Alveolar, interdental or lateral frica- /s/, /ś/, /š/
tive; or Palato-alveolar fricative or /θ/
s/z Alveolar affricate /c / or /s’/
š/t/z Palato-alveolar affricate /č /

In the reverse order, this gives

Elamite phoneme Graphic rendering
/c / s-series, s/z-variation
/č / ´/z-series, š/t/z-variation
/s/ š-series, s/š-variation
/ś/ or /θ/ s-series, s/š-variation
/s’/ ´/z-series, s/z-variation
/š/ š-series, s/š-variation
/À/ ´/z-series

References
Bork 1910 Bork, F. Nochmals das Alter der altpersischen Keilschrift.
ZDMG 64:569–580.
Bork 1925 Bork, F. Elam. B. Sprache. Reallexikon der Vorgeschichte.
Vol. 3. Berlin. Pp. 70–83.
Cameron 1948 Cameron, G. G. Persepolis Treasury Tablets (OIP 65). Chi-
cago.
Cameron 1954–1959 Cameron, G. G. The “Daiva” Inscription of Xerxes: In
Elamite. WO 2:470–476.
Diakonoff 1988 Diakonoff, I. M. Afrasian Languages. Moscow.
Dossin 1962 Dossin, G. Bronzes inscrits du Luristan de la collection
Foroughi. IrAnt 2:149–164.
Durand 1986 Durand, J.-M. Fragments rejoints pour une histoire éla-
mite. Fragmenta Historiae Elamicae. Mélanges offerts à M.-J.
Steve. Paris. Pp. 111–128.
Fales 1986 Fales, F. M. Aramaic Epigraphs on Clay Tablets of the Neo-
Assyrian Period (SS NS 2). Roma.
Foy 1898 Foy, W. Beiträge zur Erklärung der susischen Achaeme-
nideninschriften. ZDMG 52:564–605.
Glassner–Herrero 1990 Glassner, J. J.; Herrero, P. Haft Tépé: choix de textes. I.
IrAnt 25:1–45.
Grillot-Susini–Roche
1987 Grillot-Susini, F.; Roche, C. Eléments de grammaire élamite
(Editions Recherche sur les Civilisations 29). Paris.
Hallock 1969 Hallock, R. T. Persepolis Fortification Tablets (OIP 92). Chi-
cago.
Henkelman 2008 Henkelman, W. F. M. The Other Gods Who are: Studies in
Elamite-Iranian Acculturation Based on the Persepolis Fortifi-
cation Texts (Achaemenid History 14). Leiden.
 J. Tavernier, On the Sounds Rendered by the s-, š-, ´/z-Series… 1077

Hinz 1973 Hinz, W. Neue Wege im Altpersischen (GOF 3/Ir 1). Wies-
baden.
Hinz 1975 Hinz, W. Altiranisches Sprachgut der Nebenüberlieferungen
(GOF 3/Ir 3). Wiesbaden.
Holtzmann 1851 Holtzmann, H. Ueber die zweite Art der Achämenidi-
schen Keilschrift. ZDMG 5:145–178.
Hüsing 1898 Hüsing, G. Elamische Studien. I (MVAG 3/7). Berlin.
Khačikyan 1995 Khačikyan, M. Notes on Elamite Phonology. SMEA 35:
105–109.
Khačikyan 1998 Khačikyan, M. The Elamite Language (Documenta Asiana
4). Roma.
Koch 1977 Koch, H. Die religiösen Verhältnisse der Dareioszeit: Untersu-
chungen an Hand der elamischen Persepolistäfelchen (GOF 3/
Ir 4). Wiesbaden.
Krebernik 2005 Krebernik, M. Elamisch. Streck, M. P. (ed.). Sprachen des
alten Orients. Darmstadt. Pp. 159–182.
Krebernik 2006 Krebernik, M. Philologische Aspekte elamisch-mesopo-
tamischer Beziehungen im Überblick. B&B 3:61–99.
Labat 1951 Labat, R. Structure de la langue élamite (état présent de
la question). Conférences de l’Institut de Linguistique de Paris
9:23–42.
McAlpin 1981 McAlpin, D. Proto-Elamo-Dravidian: The Evidence and Its
Implications (Transactions of the American Philosophical
Society 71/3). Philadelphia.
Mordtmann 1862 Mordtmann, A. D. Erklärung der Keilinschriften zweiter
Gattung. ZDMG 16:1–126.
Norris 1855 Norris, E. Memoir on the Scythic Version of the Behistun
Inscription. JRAS 15:1–213.
Paper 1955 Paper, H. H. The Phonology and Morphology of Royal
Achaemenid Elamite. Ann Arbor.
Parpola 1974 Parpola, S. The alleged Middle/Neo-Assyrian Irregular
Verb *na´´ and the Assyrian Sound Change š > s. Assur
1/1:1–10.
Reiner 1969 Reiner, E. The Elamite Language. Friedrich, J. (ed.). Alt-
kleinasiatische Sprachen (HdO I 2/1–2/2). Leiden. Pp. 54–118.
Steinkeller 1987–1990 Steinkeller, P. Man-ištūšu. A. Philologisch. RlA 7:334–
335.
Stève 1992 Stève, M.-J. Syllabaire élamite: histoire et paléographie (CPOP
1). Neuchâtel–Paris.
Stolper 1982 Stolper, M. W. On the Dynasty of Šimaški and the Early
Sukkalma¶s. ZA 72:42–67.
Stolper 2004 Stolper, M. W. Elamite. Woodard, R. (ed.). The Cambridge
Encyclopedia of the World’s Ancient Languages. Cambridge.
Pp. 60–94.
Tavernier 2004 Tavernier, J. Some Thoughts on Neo-Elamite Chronology.
Arta: Achaemenid Research on Texts and Archaeology (http://www.
achemenet.com/ressources/enligne/arta/table.htm).
1078 Limited-corpus Languages of the Ancient Near East

Tavernier 2007 Tavernier, J. Iranica in the Achaemenid Period (ca. 550–330

B. C.). Lexicon of Old Iranian Proper Names and Loanwords,
Attested in Non-Iranian Texts (OLA 158). Leuven.
Vallat 1996 Vallat, F. Nouvelle analyse des inscriptions néo-élamites.
Gasche, H.; Hrouda, B. (eds.). Collectanea Orientalia. His-
toire, Arts de l’espace et industrie de la terre: Études offertes en
hommage à Agnès Spycket (CPOA 1). Neuchâtel. Pp. 385–
395.
Vallat 2000 Vallat, F. Le ‘clergé’ élamite. Graziani, S. (ed.). Studi sul
vicino oriente antico, dedicata alla memoria di Luigi Cagni.
Napoli. Pp. 1065–1074.
Vallat 2002–2003 Vallat, F. Suse G. La religion suso-élamite. SDB 74:529–
553.
Vallat 2004 Vallat, F. Le cylindre de Úute-kazan et la chronologie des
premiers sukkalma¶. Akkadica 125:135–140.
Vallat 2006 Vallat, F. Atta-¶amiti-Inšušinak, Šutur-Na¶¶unte et la
chronologie néo-élamite. Akkadica 127:59–62.
Waters 2000 Waters, M. W. A Survey of Neo-Elamite History (SAAS 12).
Helsinki.
Weidner 1917 Weidner, E. F. Studien zur hethitischen Sprachwissenschaft
(LSS 7/1–2). Leipzig.
Weissbach 1890 Weissbach, F. H. Die Achämenideninschriften zweiter Art.
Leipzig.
Westergaard 1845 Westergaard, N. L. Zur Entzifferung der Achämenidi-
schen Keilschrift zweiter Gattung. ZKM 6:337–466.
Whiting 1987 Whiting, R. Old Babylonian Letters from Tell Asmar (AS 22).
Chicago.
Zadok 1984 Zadok, R. The Elamite Onomasticon (AION Sup 40). Na-
poli.